2024
Qiang, Dan; McKenzie, Grant
Navigating the Post-Pandemic Urban Landscape: Disparities in Transportation Recovery & Regional Insights from New York City Journal Article
In: Computers, Environment and Urban Systems, vol. 110, pp. 102111, 2024.
@article{nokey,
title = {Navigating the Post-Pandemic Urban Landscape: Disparities in Transportation Recovery & Regional Insights from New York City},
author = {Dan Qiang and Grant McKenzie},
url = {https://grantmckenzie.com/academics/Qiang_2024.pdf},
doi = {10.1016/j.compenvurbsys.2024.102111},
year = {2024},
date = {2024-04-06},
journal = {Computers, Environment and Urban Systems},
volume = {110},
pages = {102111},
abstract = {The onset of the global Covid-19 pandemic in early 2020 brought many transportation systems in North America to a standstill. As life returned to normal, various modes of transportation exhibited differing rates of recovery, with disparities across regions. Limited research has delved into the regional variations in the recovery of these modes of transit over the past years. Such analysis is crucial for gaining insights into urban recovery and resilience, as well as understanding the factors influencing such recovery. In this work, we investigate the usage recovery of taxis, ride-hailing services, and subway ridership following the Covid-19 pandemic. We focus on New York City as our case study, employing clustering techniques to identify neighborhoods with similar recovery patterns. Furthermore, we examine the socio-economic, demographic, and built-environment factors contributing to regional variations in this recovery. Our research findings reveal that different modes of transportation responded differently to the pandemic, and these responses exhibited regional disparities. These findings hold significance for future health-related emergency response strategies and the regulation of existing transportation infrastructure.
},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
The onset of the global Covid-19 pandemic in early 2020 brought many transportation systems in North America to a standstill. As life returned to normal, various modes of transportation exhibited differing rates of recovery, with disparities across regions. Limited research has delved into the regional variations in the recovery of these modes of transit over the past years. Such analysis is crucial for gaining insights into urban recovery and resilience, as well as understanding the factors influencing such recovery. In this work, we investigate the usage recovery of taxis, ride-hailing services, and subway ridership following the Covid-19 pandemic. We focus on New York City as our case study, employing clustering techniques to identify neighborhoods with similar recovery patterns. Furthermore, we examine the socio-economic, demographic, and built-environment factors contributing to regional variations in this recovery. Our research findings reveal that different modes of transportation responded differently to the pandemic, and these responses exhibited regional disparities. These findings hold significance for future health-related emergency response strategies and the regulation of existing transportation infrastructure.
Verma, Priyanka; McKenzie, Grant
Regional Comparison of Socio-Demographic Variation in Urban E-scooter Usage Journal Article
In: Environment and Planning B: Urban Analytics and City Science, 2024.
@article{Verma2024,
title = {Regional Comparison of Socio-Demographic Variation in Urban E-scooter Usage},
author = {Priyanka Verma and Grant McKenzie},
url = {https://grantmckenzie.com/academics/Verma_2024.pdf},
year = {2024},
date = {2024-03-26},
urldate = {2024-03-09},
journal = {Environment and Planning B: Urban Analytics and City Science},
abstract = {In recent years we have witnessed explosive growth in the shared, free-floating, electric scooter industry. While still controversial in many North American cities, a number of large e-scooter operators have managed to carve out a piece of the urban transportation landscape. As these vehicles shift from novelty services to increasingly reliable modes of short personal travel, the discussion has turned to investigating who exactly benefits from these micromobility services and who are being left behind. Though population surveys have been administered to identify the socio-demographic characteristics of e-scooter riders in the past, little work has linked these characteristics through trips, or investigated the regional variation in these demographic factors. In this work we explore the variability and similarities in e-scooter rider characteristics across three major U.S. cities. To accomplish this, we apply a Moran’s Eigenvector Spatial Filtering linear regression model and compare our results to more commonly used spatial regression approaches. Our results indicate that the spatial filtering approach outperforms other methods in identifying socio-demographic characteristics of e-scooter users, across multiple regions. We find that many socio-demographics associated with e-scooter usage are regionally variant, despite younger users making up the core user base in all cities. There are variations in usage based on gender, income, and race across cities with Black and Hispanic populations remaining underserved. The implications of these findings are discussed.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
In recent years we have witnessed explosive growth in the shared, free-floating, electric scooter industry. While still controversial in many North American cities, a number of large e-scooter operators have managed to carve out a piece of the urban transportation landscape. As these vehicles shift from novelty services to increasingly reliable modes of short personal travel, the discussion has turned to investigating who exactly benefits from these micromobility services and who are being left behind. Though population surveys have been administered to identify the socio-demographic characteristics of e-scooter riders in the past, little work has linked these characteristics through trips, or investigated the regional variation in these demographic factors. In this work we explore the variability and similarities in e-scooter rider characteristics across three major U.S. cities. To accomplish this, we apply a Moran’s Eigenvector Spatial Filtering linear regression model and compare our results to more commonly used spatial regression approaches. Our results indicate that the spatial filtering approach outperforms other methods in identifying socio-demographic characteristics of e-scooter users, across multiple regions. We find that many socio-demographics associated with e-scooter usage are regionally variant, despite younger users making up the core user base in all cities. There are variations in usage based on gender, income, and race across cities with Black and Hispanic populations remaining underserved. The implications of these findings are discussed.
2023
McKenzie, Grant; Zhang, Hongyu; Gambs, Sebastien
Privacy and Ethics in GeoAI Book Chapter
In: Gao, Song; Hu, Yingjie; Li, Wenwen (Ed.): Handbook of Geospatial Artificial Intelligence, Chapter 19, CRC Press, 1, 2023, ISBN: 9781003308423.
@inbook{McKenzie2023c,
title = {Privacy and Ethics in GeoAI},
author = {Grant McKenzie and Hongyu Zhang and Sebastien Gambs},
editor = {Song Gao and Yingjie Hu and Wenwen Li},
url = {https://grantmckenzie.com/academics/Privacy_GeoAI.pdf},
doi = {https://doi.org/10.1201/9781003308423},
isbn = {9781003308423},
year = {2023},
date = {2023-12-28},
urldate = {2023-12-28},
booktitle = {Handbook of Geospatial Artificial Intelligence},
publisher = {CRC Press},
edition = {1},
chapter = {19},
abstract = {Any advancement in technology is accompanied by new concerns over its ethical use and impacts on privacy. While a notoriously difficult term to define, privacy as it relates to technology usage, can be described as the ability of an individual or group to control their personal information. Like many ethical concepts, this definition evolves with changes in societal and technical norms. The emergence of machine learning and related artificial intelligence techniques has again shifted societal concerns about the privacy of our persons, socio-demographic group membership, and personal data. Location data are particularly sensitive as they link information across sources and can be used to infer a wide variety of personal information. This makes data privacy one of the most important ethical discussions within the field of geographic artificial intelligence (GeoAI). The main objective of this chapter is to explore the unique privacy concerns associated with AI techniques used for analyzing geospatial information. After providing an overview of the topic, we describe some of the most common techniques and leading application areas through which data privacy and GeoAI are converging. Finally, we suggest a number of ways that privacy within GeoAI can improve and highlight emerging topics within the field.},
keywords = {},
pubstate = {published},
tppubtype = {inbook}
}
Any advancement in technology is accompanied by new concerns over its ethical use and impacts on privacy. While a notoriously difficult term to define, privacy as it relates to technology usage, can be described as the ability of an individual or group to control their personal information. Like many ethical concepts, this definition evolves with changes in societal and technical norms. The emergence of machine learning and related artificial intelligence techniques has again shifted societal concerns about the privacy of our persons, socio-demographic group membership, and personal data. Location data are particularly sensitive as they link information across sources and can be used to infer a wide variety of personal information. This makes data privacy one of the most important ethical discussions within the field of geographic artificial intelligence (GeoAI). The main objective of this chapter is to explore the unique privacy concerns associated with AI techniques used for analyzing geospatial information. After providing an overview of the topic, we describe some of the most common techniques and leading application areas through which data privacy and GeoAI are converging. Finally, we suggest a number of ways that privacy within GeoAI can improve and highlight emerging topics within the field.
Brunila, Mikael; LaViolette, Jack; CH-Wang, Sky; Verma, Priyanka; Féré, Clara; McKenzie, Grant
Toward a Critical Toponymy Framework for Named Entity Recognition: A Case Study of Airbnb in New York City Best Paper Proceedings Article
In: Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing (EMNLP2023), 2023.
@inproceedings{Brunila2023,
title = {Toward a Critical Toponymy Framework for Named Entity Recognition: A Case Study of Airbnb in New York City},
author = {Mikael Brunila and Jack LaViolette and Sky CH-Wang and Priyanka Verma and Clara Féré and Grant McKenzie},
doi = {https://doi.org/10.48550/arXiv.2310.15302},
year = {2023},
date = {2023-12-10},
urldate = {2023-10-25},
booktitle = {Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing (EMNLP2023)},
abstract = {Critical toponymy examines the dynamics of power, capital, and resistance through place names and the sites to which they refer. Studies here have traditionally focused on the semantic content of toponyms and the top-down institutional processes that produce them. However, they have generally ignored the ways in which toponyms are used by ordinary people in everyday discourse, as well as the other strategies of geospatial description that accompany and contextualize toponymic reference. Here, we develop computational methods to measure how cultural and economic capital shape the ways in which people refer to places, through a novel annotated dataset of 47,440 New York City Airbnb listings from the 2010s. Building on this dataset, we introduce a new named entity recognition (NER) model able to identify important discourse categories integral to the characterization of place. Our findings point toward new directions for critical toponymy and to a range of previously understudied linguistic signals relevant to research on neighborhood status, housing and tourism markets, and gentrification.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Critical toponymy examines the dynamics of power, capital, and resistance through place names and the sites to which they refer. Studies here have traditionally focused on the semantic content of toponyms and the top-down institutional processes that produce them. However, they have generally ignored the ways in which toponyms are used by ordinary people in everyday discourse, as well as the other strategies of geospatial description that accompany and contextualize toponymic reference. Here, we develop computational methods to measure how cultural and economic capital shape the ways in which people refer to places, through a novel annotated dataset of 47,440 New York City Airbnb listings from the 2010s. Building on this dataset, we introduce a new named entity recognition (NER) model able to identify important discourse categories integral to the characterization of place. Our findings point toward new directions for critical toponymy and to a range of previously understudied linguistic signals relevant to research on neighborhood status, housing and tourism markets, and gentrification.
Brunila, Mikael; Verma, Priyanka; McKenzie, Grant
When Everything Is "Nearby": How Airbnb Listings in New York City Exaggerate Proximity Proceedings Article
In: Proceedings of the 12th International Conference on Geographic Information Science, Schloss Dagstuhl -- Leibniz-Zentrum fur Informatik, 2023.
@inproceedings{Brunila2023b,
title = {When Everything Is "Nearby": How Airbnb Listings in New York City Exaggerate Proximity},
author = {Mikael Brunila and Priyanka Verma and Grant McKenzie},
url = {https://drops.dagstuhl.de/opus/volltexte/2023/18911/},
doi = {10.4230/LIPIcs.GIScience.2023.16},
year = {2023},
date = {2023-08-31},
booktitle = {Proceedings of the 12th International Conference on Geographic Information Science},
volume = {277},
publisher = {Schloss Dagstuhl -- Leibniz-Zentrum fur Informatik},
series = {Leibniz International Proceedings in Informatics (LIPIcs)},
abstract = {In recent years, the emergence and rapid growth of short-term rental (STR) markets has exerted considerable influence on real estate in most large cities across the world. Central location and transit access are two primary factors associated with the prevalence and expansion of STRs, including Airbnbs. Nevertheless, perhaps due to methodological challenges, no research has addressed how location and proximity are represented in the titles and descriptions of STRs. In this paper, we introduce a new methodological pipeline to extract spatial relations from text and show that expressions of distance in STR listings can indeed be quantified and measured against real-world distances. We then comparatively analyze Airbnb reviews (written by guests) and listings (written by hosts) from New York City in order to demonstrate systematically how listings exaggerate proximity compared to reviews. Moreover, we discover spatial patterns to these differences that warrant further investigation.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
In recent years, the emergence and rapid growth of short-term rental (STR) markets has exerted considerable influence on real estate in most large cities across the world. Central location and transit access are two primary factors associated with the prevalence and expansion of STRs, including Airbnbs. Nevertheless, perhaps due to methodological challenges, no research has addressed how location and proximity are represented in the titles and descriptions of STRs. In this paper, we introduce a new methodological pipeline to extract spatial relations from text and show that expressions of distance in STR listings can indeed be quantified and measured against real-world distances. We then comparatively analyze Airbnb reviews (written by guests) and listings (written by hosts) from New York City in order to demonstrate systematically how listings exaggerate proximity compared to reviews. Moreover, we discover spatial patterns to these differences that warrant further investigation.
Qiang, Dan; McKenzie, Grant
Mobility Vitality: Assessing Neighborhood Similarity Through Transportation Patterns In New York City Proceedings Article
In: Proceedings of the 12th International Conference on Geographic Information Science, Schloss Dagstuhl -- Leibniz-Zentrum fur Informatik, 2023.
@inproceedings{Qiang2023,
title = {Mobility Vitality: Assessing Neighborhood Similarity Through Transportation Patterns In New York City},
author = {Dan Qiang and Grant McKenzie},
year = {2023},
date = {2023-08-31},
booktitle = {Proceedings of the 12th International Conference on Geographic Information Science},
volume = {277},
publisher = {Schloss Dagstuhl -- Leibniz-Zentrum fur Informatik},
series = {Leibniz International Proceedings in Informatics (LIPIcs)},
abstract = {Though numerous studies have examined human mobility within an urban environment, few have explored the concept of urban vitality purely through the lens of urban transportation. Given the importance of different modes of transportation within a city, such analysis is necessary. In this short paper, we introduce the novel concept of mobility vitality by integrating human mobility and urban vitality, offering a multilayered framework to assess the degree of transportation and mobility within and between regions. The mobility patterns of three transportation modes, namely subway, taxicab, and bike-share, are first examined independently. These patterns are then aggregated to form the composite measure of static mobility vitality. Through this measure, we evaluate similarities between neighborhoods. Our results observed significant spatial differences in the travel patterns of three transportation modes on weekdays and weekends. Moreover, neighborhoods with high static mobility vitality have relatively similar mobility patterns. Ultimately, this approach aims to find neighborhoods with imbalanced transportation infrastructure or inadequate public.
},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Though numerous studies have examined human mobility within an urban environment, few have explored the concept of urban vitality purely through the lens of urban transportation. Given the importance of different modes of transportation within a city, such analysis is necessary. In this short paper, we introduce the novel concept of mobility vitality by integrating human mobility and urban vitality, offering a multilayered framework to assess the degree of transportation and mobility within and between regions. The mobility patterns of three transportation modes, namely subway, taxicab, and bike-share, are first examined independently. These patterns are then aggregated to form the composite measure of static mobility vitality. Through this measure, we evaluate similarities between neighborhoods. Our results observed significant spatial differences in the travel patterns of three transportation modes on weekdays and weekends. Moreover, neighborhoods with high static mobility vitality have relatively similar mobility patterns. Ultimately, this approach aims to find neighborhoods with imbalanced transportation infrastructure or inadequate public.
McKenzie, Grant; Zhang, Hongyu
Platial k-anonymity: Improving location anonymity through temporal popularity signatures Best Paper Proceedings Article
In: Proceedings of the 12th International Conference on Geographic Information Science, Schloss Dagstuhl -- Leibniz-Zentrum fur Informatik, 2023, ISSN: 1868-8969.
@inproceedings{McKenzie2023b,
title = {Platial k-anonymity: Improving location anonymity through temporal popularity signatures},
author = {Grant McKenzie and Hongyu Zhang },
url = {https://drops.dagstuhl.de/opus/volltexte/2023/18904/},
issn = {1868-8969},
year = {2023},
date = {2023-08-31},
urldate = {2023-08-31},
booktitle = {Proceedings of the 12th International Conference on Geographic Information Science},
volume = {277},
publisher = {Schloss Dagstuhl -- Leibniz-Zentrum fur Informatik},
series = {Leibniz International Proceedings in Informatics (LIPIcs)},
abstract = {While it is increasingly necessary in today's digital society, sharing personal location information comes at a cost. Sharing one's precise place of interest, e.g., Compass Coffee, enables a range of location-based services, but substantially reduces the individual's privacy. Methods have been developed to obfuscate and anonymize location data while still maintaining a degree of utility. One such approach, spatial k-anonymity, aims to ensure an individual's level of anonymity by reporting their location as a set of k potential locations rather than their actual location alone. Larger values of k increase spatial anonymity while decreasing the utility of the location information. Typical examples of spatial k-anonymized datasets present elements as simple geographic points with no attributes or contextual information. In this work, we demonstrate that the addition of publicly available contextual data can significantly reduce the anonymity of a k-anonymized dataset. Through the analysis of place type temporal visitation patterns, hours of operation, and popularity values, one's anonymity can decreased by more than 50 percent. We propose a platial k-anonymity approach that leverages a combination of temporal popularity signatures and report the amount that k must increase in order to maintain a certain level of anonymity. Finally, a method for reporting platial k-anonymous regions is presented and the implications of our methods are discussed.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
While it is increasingly necessary in today's digital society, sharing personal location information comes at a cost. Sharing one's precise place of interest, e.g., Compass Coffee, enables a range of location-based services, but substantially reduces the individual's privacy. Methods have been developed to obfuscate and anonymize location data while still maintaining a degree of utility. One such approach, spatial k-anonymity, aims to ensure an individual's level of anonymity by reporting their location as a set of k potential locations rather than their actual location alone. Larger values of k increase spatial anonymity while decreasing the utility of the location information. Typical examples of spatial k-anonymized datasets present elements as simple geographic points with no attributes or contextual information. In this work, we demonstrate that the addition of publicly available contextual data can significantly reduce the anonymity of a k-anonymized dataset. Through the analysis of place type temporal visitation patterns, hours of operation, and popularity values, one's anonymity can decreased by more than 50 percent. We propose a platial k-anonymity approach that leverages a combination of temporal popularity signatures and report the amount that k must increase in order to maintain a certain level of anonymity. Finally, a method for reporting platial k-anonymous regions is presented and the implications of our methods are discussed.
McKenzie, Grant; Battersby, Sarah; Setlur, Vidya
MixMap: A user-driven approach to place-based semantic similarity Journal Article
In: Cartography and Geographic Information Science, pp. 1-16, 2023.
@article{McKenzie2023,
title = {MixMap: A user-driven approach to place-based semantic similarity},
author = {Grant McKenzie and Sarah Battersby and Vidya Setlur},
doi = {10.1080/15230406.2023.2176930},
year = {2023},
date = {2023-03-02},
urldate = {2023-03-02},
journal = {Cartography and Geographic Information Science},
pages = {1-16},
publisher = {Taylor & Francis},
abstract = {What other locations are like my neighborhood? How? Why? The heart of many spatial analyses is in finding similarities or dissimilarities between locations. Discovering patterns and interpreting similarity is a complicated process that is based on both the spatial characteristics and the semantics or meaning that we assign to place. Human conceptualization of similarity in locations is multi-faceted and cannot be captured with a simple assessment of single numeric attributes like population density or median income; however, these quantifiable attributes are the basis for an initial pass of sense-making. MixMap facilitates the incorporation of similarity measures and spatial analytics to provide an information reduction (or semantic generalization) that brings the user closer to actionable insights. Through a preliminary evaluation of MixMap, we found that the tool supports the geospatial inquiry of determining similarity between regions, where participants can manipulate individual weights of the various attributes describing these locations. Based on feedback and observations from the study, we discuss potential implications and considerations for exploring the role of context and additional place-specific parameters for computing similarity, as well as understanding the nuances of semantics for place similarity in geospatial analysis tools.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
What other locations are like my neighborhood? How? Why? The heart of many spatial analyses is in finding similarities or dissimilarities between locations. Discovering patterns and interpreting similarity is a complicated process that is based on both the spatial characteristics and the semantics or meaning that we assign to place. Human conceptualization of similarity in locations is multi-faceted and cannot be captured with a simple assessment of single numeric attributes like population density or median income; however, these quantifiable attributes are the basis for an initial pass of sense-making. MixMap facilitates the incorporation of similarity measures and spatial analytics to provide an information reduction (or semantic generalization) that brings the user closer to actionable insights. Through a preliminary evaluation of MixMap, we found that the tool supports the geospatial inquiry of determining similarity between regions, where participants can manipulate individual weights of the various attributes describing these locations. Based on feedback and observations from the study, we discuss potential implications and considerations for exploring the role of context and additional place-specific parameters for computing similarity, as well as understanding the nuances of semantics for place similarity in geospatial analysis tools.
2022
Zhang, Hongyu; McKenzie, Grant
Towards place-based privacy: Challenges and opportunities in the “smart” world Proceedings Article
In: IEEE International Symposium on Technology and Society 2022, IEEE, 2022.
@inproceedings{Zhang2022b,
title = {Towards place-based privacy: Challenges and opportunities in the “smart” world},
author = {Hongyu Zhang and Grant McKenzie },
url = {https://grantmckenzie.com/academics/HZhang_ISTAS_2022.pdf},
year = {2022},
date = {2022-11-10},
urldate = {2022-11-10},
booktitle = {IEEE International Symposium on Technology and Society 2022},
publisher = {IEEE},
abstract = {The emergence of “smart” technologies has given rise to new interaction models merging our physical realities with our digital environments. As a result, new privacy threats have emerged, substantially impacting both individuals and groups. In this short paper, we summarize many of the privacy challenges we face in the smart and connected world, and identify opportunities for further research. Drawing from the recent literature on geoprivacy, user-tailored privacy, and group privacy, we explore this topic through the lens of contextually aware, place-based, or platial, information analysis.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
The emergence of “smart” technologies has given rise to new interaction models merging our physical realities with our digital environments. As a result, new privacy threats have emerged, substantially impacting both individuals and groups. In this short paper, we summarize many of the privacy challenges we face in the smart and connected world, and identify opportunities for further research. Drawing from the recent literature on geoprivacy, user-tailored privacy, and group privacy, we explore this topic through the lens of contextually aware, place-based, or platial, information analysis.
Brunila, Mikael; McConnell, Michael; Grigg, Stalgia
DRIFT: End-to-end encrypted spatial feature sharing & instant messaging Proceedings Article
In: Proceedings of the 6th ACM SIGSPATIAL Workshop on Location-based Recommendations, Geosocial Networks and Geoadvertising, ACM, 2022.
@inproceedings{Brunila2022b,
title = {DRIFT: End-to-end encrypted spatial feature sharing & instant messaging},
author = {Mikael Brunila and Michael McConnell and Stalgia Grigg},
year = {2022},
date = {2022-11-01},
booktitle = {Proceedings of the 6th ACM SIGSPATIAL Workshop on Location-based Recommendations, Geosocial Networks and Geoadvertising},
publisher = {ACM},
abstract = {Most online communication today is inherently temporal and aspatial. Instant messaging services are structured around a timeline inter- face which prioritizes a linear succession of events and guides our attention towards the novel. In this way, the different textures of so- cial life are lost in linear reduction. In this paper, we present DRIFT, a novel and open-source instant messaging application framework, based on a different paradigm of communication that preserves temporality but organizes it around space. Instead of the timeline, our application grounds messaging in the map and its pins, offering users a tool that encourages spatio-temporal communication and the sharing of spatial features. Given increasing concerns about the safety and privacy of online user interaction, we integrate state-of- the art encryption as a core feature of our application. Firstly, to protect user messages and map pins, we implement end-to-end en- cryption with the Double Ratchet key management algorithm and the open standard Matrix protocol. Secondly, to maintain location privacy, we allow users to batch download map tilesets and machine learning models to perform operations such as search entirely on device, avoiding compromising API calls to cloud services. With these combined features, DRIFT aims to introduce a new model for online interaction that upends the short attention span imposed by the narrow timeline and replace it with a spatio-temporally rich and secure instant messaging tool for both laymen and more vulnerable users such as journalists, human rights activists, and whistleblowers.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Most online communication today is inherently temporal and aspatial. Instant messaging services are structured around a timeline inter- face which prioritizes a linear succession of events and guides our attention towards the novel. In this way, the different textures of so- cial life are lost in linear reduction. In this paper, we present DRIFT, a novel and open-source instant messaging application framework, based on a different paradigm of communication that preserves temporality but organizes it around space. Instead of the timeline, our application grounds messaging in the map and its pins, offering users a tool that encourages spatio-temporal communication and the sharing of spatial features. Given increasing concerns about the safety and privacy of online user interaction, we integrate state-of- the art encryption as a core feature of our application. Firstly, to protect user messages and map pins, we implement end-to-end en- cryption with the Double Ratchet key management algorithm and the open standard Matrix protocol. Secondly, to maintain location privacy, we allow users to batch download map tilesets and machine learning models to perform operations such as search entirely on device, avoiding compromising API calls to cloud services. With these combined features, DRIFT aims to introduce a new model for online interaction that upends the short attention span imposed by the narrow timeline and replace it with a spatio-temporally rich and secure instant messaging tool for both laymen and more vulnerable users such as journalists, human rights activists, and whistleblowers.
Brunila, Mikael; LaViolette, Jack
What kind of company do words keep? Revisiting the distributional semantics of J.R. Firth & Zellig Harris Proceedings Article
In: Proceedings of the 2022 Annual Conference of the North American Chapter of the Association for Computational Linguistics, 2022.
@inproceedings{Brunila2022,
title = {What kind of company do words keep? Revisiting the distributional semantics of J.R. Firth & Zellig Harris},
author = {Mikael Brunila and Jack LaViolette},
url = {https://arxiv.org/abs/2205.07750},
year = {2022},
date = {2022-07-10},
urldate = {2022-07-10},
booktitle = {Proceedings of the 2022 Annual Conference of the North American Chapter of the Association for Computational Linguistics},
abstract = {The power of word embeddings is attributed to the linguistic theory that similar words will appear in similar contexts. This idea is specifically invoked by noting that "you shall know a word by the company it keeps," a quote from British linguist J.R. Firth who, along with his American colleague Zellig Harris, is often credited with the invention of "distributional semantics." While both Firth and Harris are cited in all major NLP textbooks and many foundational papers, the content and differences between their theories is seldom discussed. Engaging in a close reading of their work, we discover two distinct and in many ways divergent theories of meaning. One focuses exclusively on the internal workings of linguistic forms, while the other invites us to consider words in new company - not just with other linguistic elements, but also in a broader cultural and situational context. Contrasting these theories from the perspective of current debates in NLP, we discover in Firth a figure who could guide the field towards a more culturally grounded notion of semantics. We consider how an expanded notion of "context" might be modeled in practice through two different strategies: comparative stratification and syntagmatic extension.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
The power of word embeddings is attributed to the linguistic theory that similar words will appear in similar contexts. This idea is specifically invoked by noting that "you shall know a word by the company it keeps," a quote from British linguist J.R. Firth who, along with his American colleague Zellig Harris, is often credited with the invention of "distributional semantics." While both Firth and Harris are cited in all major NLP textbooks and many foundational papers, the content and differences between their theories is seldom discussed. Engaging in a close reading of their work, we discover two distinct and in many ways divergent theories of meaning. One focuses exclusively on the internal workings of linguistic forms, while the other invites us to consider words in new company - not just with other linguistic elements, but also in a broader cultural and situational context. Contrasting these theories from the perspective of current debates in NLP, we discover in Firth a figure who could guide the field towards a more culturally grounded notion of semantics. We consider how an expanded notion of "context" might be modeled in practice through two different strategies: comparative stratification and syntagmatic extension.
Romm, Daniel; Verma, Priyanka; Karpinski, Elizabeth; Sanders, Tracy; McKenzie, Grant
Differences in First-Mile and Last-Mile Behaviour in Candidate Multi-Modal Boston Bike-share Micromobility Trips Journal Article
In: Journal of Transport Geography, vol. 102, iss. June, 2022.
@article{Romm2022,
title = {Differences in First-Mile and Last-Mile Behaviour in Candidate Multi-Modal Boston Bike-share Micromobility Trips},
author = {Daniel Romm and Priyanka Verma and Elizabeth Karpinski and Tracy Sanders and Grant McKenzie},
url = {https://doi.org/10.1016/j.jtrangeo.2022.103370},
year = {2022},
date = {2022-06-07},
urldate = {2022-06-07},
journal = {Journal of Transport Geography},
volume = {102},
issue = {June},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
McKenzie, Grant; Romm, Daniel; Zhang, Hongyu; Brunila, Mikael
PrivyTo: A privacy preserving location sharing platform Journal Article
In: Transactions in GIS, 2022.
@article{McKenzie2022,
title = {PrivyTo: A privacy preserving location sharing platform},
author = {Grant McKenzie and Daniel Romm and Hongyu Zhang and Mikael Brunila},
url = {https://grantmckenzie.com/academics/McKenzie2022_PrivyTo.pdf
https://onlinelibrary.wiley.com/doi/full/10.1111/tgis.12924},
doi = {10.1111/tgis.12924},
year = {2022},
date = {2022-05-12},
urldate = {2022-03-11},
journal = {Transactions in GIS},
publisher = {Wiley Press},
abstract = {Concern over the privacy of one's personal location is at an all-time high, yet the desire to share our lives with friends, family, and the public persists. Current methods and applications for sharing location content with the range of people in our lives are sorely lacking. Application users are often limited to sharing a single spatial resolution with all individuals, regardless of relation, and with little control over how this content is shared. Processes for sharing typically involve allowing a for-profit company access to your location before it can be transmitted to the intended recipient. In this work we propose a set of design goals and a design pattern for sharing personal location information that are realized through a prototype mobile web application. Our approach is built on the novel idea of obfuscated and encrypted location views, and promotes a uniquely open method for sharing. The intention of this paper is to demonstrate that location sharing need not require one to expose private location information to third parties, and that methods exist to put an individual back in control of their content.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Concern over the privacy of one's personal location is at an all-time high, yet the desire to share our lives with friends, family, and the public persists. Current methods and applications for sharing location content with the range of people in our lives are sorely lacking. Application users are often limited to sharing a single spatial resolution with all individuals, regardless of relation, and with little control over how this content is shared. Processes for sharing typically involve allowing a for-profit company access to your location before it can be transmitted to the intended recipient. In this work we propose a set of design goals and a design pattern for sharing personal location information that are realized through a prototype mobile web application. Our approach is built on the novel idea of obfuscated and encrypted location views, and promotes a uniquely open method for sharing. The intention of this paper is to demonstrate that location sharing need not require one to expose private location information to third parties, and that methods exist to put an individual back in control of their content.
Zhang, Hongyu; McKenzie, Grant
Rehumanize geoprivacy: From disclosure control to human perception Journal Article
In: GeoJournal, 2022.
@article{Zhang2022,
title = {Rehumanize geoprivacy: From disclosure control to human perception},
author = {Hongyu Zhang and Grant McKenzie},
doi = {https://doi.org/10.1007/s10708-022-10598-4},
year = {2022},
date = {2022-02-16},
urldate = {2022-01-31},
journal = {GeoJournal},
publisher = {Springer},
abstract = {Traditional boundaries between people are vanishing due to the rise of Internet of Things technology. Our smart devices keep us connected to the world, but also monitor our daily lives through an unprecedented amount data collection. As a result, defining privacy has become more complicated. Individuals want to leverage new technology (e.g., making friends through sharing private experiences) and also avoid unwanted consequences (e.g., targeted advertising). In the age of ubiquitous digital content, geoprivacy is unique because concerns in this area are constantly changing and context-dependent. Multiple factors influence people’s location disclosure decisions, including time, culture, demographics, spatial granularity, and trust. Existing research primarily focuses on the computational efforts of protecting geoprivacy, while the variation of geoprivacy perceptions has yet to receive adequate attention in the data science literature. In this work, we explore geoprivacy from a cognate-based perspective and tackle our changing perception of the concept from multiple angles. Our objectives are to rehumanize this field from contextual, cultural, and economic dimensions and highlight the uniqueness of geodata under the broad topic of privacy. It is essential that we understand the spatial variations of geoprivacy perceptions in the era of big data. Masking geographic coordinates can no longer fully anonymize spatial data, and targeted geoprivacy protection needs to be further investigated to improve user experience.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Traditional boundaries between people are vanishing due to the rise of Internet of Things technology. Our smart devices keep us connected to the world, but also monitor our daily lives through an unprecedented amount data collection. As a result, defining privacy has become more complicated. Individuals want to leverage new technology (e.g., making friends through sharing private experiences) and also avoid unwanted consequences (e.g., targeted advertising). In the age of ubiquitous digital content, geoprivacy is unique because concerns in this area are constantly changing and context-dependent. Multiple factors influence people’s location disclosure decisions, including time, culture, demographics, spatial granularity, and trust. Existing research primarily focuses on the computational efforts of protecting geoprivacy, while the variation of geoprivacy perceptions has yet to receive adequate attention in the data science literature. In this work, we explore geoprivacy from a cognate-based perspective and tackle our changing perception of the concept from multiple angles. Our objectives are to rehumanize this field from contextual, cultural, and economic dimensions and highlight the uniqueness of geodata under the broad topic of privacy. It is essential that we understand the spatial variations of geoprivacy perceptions in the era of big data. Masking geographic coordinates can no longer fully anonymize spatial data, and targeted geoprivacy protection needs to be further investigated to improve user experience.
2021
McKenzie, Grant; Mwenda, Kevin
Identifying regional variation in place visit behavior during a global pandemic Journal Article
In: Journal of Spatial Information Science, vol. 2021, no. 23, pp. 95-124, 2021.
@article{McKenzie2021c,
title = {Identifying regional variation in place visit behavior during a global pandemic},
author = {Grant McKenzie and Kevin Mwenda},
doi = {10.5311/JOSIS.2021.23.170},
year = {2021},
date = {2021-12-24},
urldate = {2021-11-11},
journal = {Journal of Spatial Information Science},
volume = {2021},
number = {23},
pages = {95-124},
abstract = {The emergence of the SARS-CoV-2 virus in 2019 lead to a global pandemic that altered the activity behavior of most people on our planet. While government regulations and public concern modified visitation patterns to places of interest, little research has examined the nuanced changes in the length of time someone spends at a place, nor the regional variability of these changes. In this work, we examine place visit duration in four major U.S. cities, identify which place types saw the largest and smallest changes, and quantify variation between cities. Furthermore, we identify socio-economic and demographic factors that contribute to changes in visit duration and demonstrate the varying influence of these factors by region. The results of our analysis indicate that the pandemic's impact on visiting behavior varies between cities, though there are commonalities found in certain types of places. Our findings suggest that places of interest within lower income communities experienced less change in visit duration than others. An increase in the percentage of younger, Black or Hispanic populations within a community also resulted in a smaller decrease in visit duration than in other communities. These findings offer insight into the factors that contribute to changes in visiting behavior and the resilience of communities to a global pandemic.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
The emergence of the SARS-CoV-2 virus in 2019 lead to a global pandemic that altered the activity behavior of most people on our planet. While government regulations and public concern modified visitation patterns to places of interest, little research has examined the nuanced changes in the length of time someone spends at a place, nor the regional variability of these changes. In this work, we examine place visit duration in four major U.S. cities, identify which place types saw the largest and smallest changes, and quantify variation between cities. Furthermore, we identify socio-economic and demographic factors that contribute to changes in visit duration and demonstrate the varying influence of these factors by region. The results of our analysis indicate that the pandemic's impact on visiting behavior varies between cities, though there are commonalities found in certain types of places. Our findings suggest that places of interest within lower income communities experienced less change in visit duration than others. An increase in the percentage of younger, Black or Hispanic populations within a community also resulted in a smaller decrease in visit duration than in other communities. These findings offer insight into the factors that contribute to changes in visiting behavior and the resilience of communities to a global pandemic.
McKenzie, Grant
Leveraging Place Reviews to Identify the Effects of COVID-19 on Canadian Tourism Proceedings Article
In: Proceedings of the third International Symposium on Platial Information Science (PLATIAL'21), Enschede, the Netherlands, 2021.
@inproceedings{McKenzie2021e,
title = {Leveraging Place Reviews to Identify the Effects of COVID-19 on Canadian Tourism},
author = {Grant McKenzie},
url = {https://grantmckenzie.com/academics/McKenzie_Platial2021.pdf},
doi = {10.5281/zenodo.5767190},
year = {2021},
date = {2021-12-08},
urldate = {2021-12-08},
booktitle = {Proceedings of the third International Symposium on Platial Information Science (PLATIAL'21)},
address = {Enschede, the Netherlands},
abstract = {The emergence of the COVID-19 pandemic disrupted travel world-wide and substantially impacted tourism in most countries. Though many governmental agencies and tourism boards have published data on the impact of the pandemic, in Canada, the vast majority of these data are reported at the national level or sparsely within individual regions. In this preliminary work, we leverage user-contributed tourist attraction reviews to better understand the nuanced changes in travel behavior resulting from the COVID-19 pandemic. We examine the regional impacts as well as the affects on different categories of tourism within Canada. The purpose of this short paper is to demonstrate the value of analyzing place-based user-generated tourism data and highlight some of the ways it can be leveraged by policy experts and tourism agencies.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
The emergence of the COVID-19 pandemic disrupted travel world-wide and substantially impacted tourism in most countries. Though many governmental agencies and tourism boards have published data on the impact of the pandemic, in Canada, the vast majority of these data are reported at the national level or sparsely within individual regions. In this preliminary work, we leverage user-contributed tourist attraction reviews to better understand the nuanced changes in travel behavior resulting from the COVID-19 pandemic. We examine the regional impacts as well as the affects on different categories of tourism within Canada. The purpose of this short paper is to demonstrate the value of analyzing place-based user-generated tourism data and highlight some of the ways it can be leveraged by policy experts and tourism agencies.
Romm, Daniel; Zhang, Hongyu; Verma, Priyanka; McKenzie, Grant; Chen, Emily
"Data Horror": Mapping (Spatial) Data Privacy Violations onto a Cognitive Account of Horror Proceedings Article
In: Proceedings of the Second Spatial Data Science Symposium , 2021.
@inproceedings{Romm2021,
title = {"Data Horror": Mapping (Spatial) Data Privacy Violations onto a Cognitive Account of Horror},
author = {Daniel Romm and Hongyu Zhang and Priyanka Verma and Grant McKenzie and Emily Chen},
url = {https://escholarship.org/uc/item/7902g5hh},
doi = {10.25436/E23S3T},
year = {2021},
date = {2021-12-01},
urldate = {2021-12-13},
booktitle = {Proceedings of the Second Spatial Data Science Symposium },
series = {Spatial Data Science Symposium 2021 Short Paper Proceedings},
abstract = {While spatial data privacy is not a new concern, recent information technology developments that allow for the increased collection and alternative use of spatial data have brought the discussion about geoprivacy back in focus. In this work, we draw a parallel between a conceptualization of horror based on work from cognitive scientists and philosophers, and the intrusiveness of current data collection methods, the unauthorized use of this data, and the transgressions made by data stewards. By drawing this connection, we discuss the familiar topic of data privacy through a novel lens that clarifies the importance of data privacy and elucidates the particular importance of geoprivacy.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
While spatial data privacy is not a new concern, recent information technology developments that allow for the increased collection and alternative use of spatial data have brought the discussion about geoprivacy back in focus. In this work, we draw a parallel between a conceptualization of horror based on work from cognitive scientists and philosophers, and the intrusiveness of current data collection methods, the unauthorized use of this data, and the transgressions made by data stewards. By drawing this connection, we discuss the familiar topic of data privacy through a novel lens that clarifies the importance of data privacy and elucidates the particular importance of geoprivacy.
Brunila, Mikael; LaViolette, Jack
WMDecompose: A Framework for Leveraging the Interpretable Properties of Word Mover's Distance in Sociocultural Analysis Proceedings Article
In: The 5th Joint SIGHUM Workshop on Computational Linguistics for Cultural Heritage, Social Sciences, Humanities and Literature, 2021.
@inproceedings{nokey,
title = {WMDecompose: A Framework for Leveraging the Interpretable Properties of Word Mover's Distance in Sociocultural Analysis},
author = {Mikael Brunila and Jack LaViolette},
url = {https://arxiv.org/abs/2110.07330},
year = {2021},
date = {2021-11-11},
urldate = {2021-11-11},
booktitle = {The 5th Joint SIGHUM Workshop on Computational Linguistics for Cultural Heritage, Social Sciences, Humanities and Literature},
abstract = {Despite the increasing popularity of NLP in the humanities and social sciences, advances in model performance and complexity have been accompanied by concerns about interpretability and explanatory power for sociocultural analysis. One popular model that balances complexity and legibility is Word Mover's Distance (WMD). Ostensibly adapted for its interpretability, WMD has nonetheless been used and further developed in ways which frequently discard its most interpretable aspect: namely, the word-level distances required for translating a set of words into another set of words. To address this apparent gap, we introduce WMDecompose: a model and Python library that 1) decomposes document-level distances into their constituent word-level distances, and 2) subsequently clusters words to induce thematic elements, such that useful lexical information is retained and summarized for analysis. To illustrate its potential in a social scientific context, we apply it to a longitudinal social media corpus to explore the interrelationship between conspiracy theories and conservative American discourses. Finally, because of the full WMD model's high time-complexity, we additionally suggest a method of sampling document pairs from large datasets in a reproducible way, with tight bounds that prevent extrapolation of unreliable results due to poor sampling practices. },
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Despite the increasing popularity of NLP in the humanities and social sciences, advances in model performance and complexity have been accompanied by concerns about interpretability and explanatory power for sociocultural analysis. One popular model that balances complexity and legibility is Word Mover's Distance (WMD). Ostensibly adapted for its interpretability, WMD has nonetheless been used and further developed in ways which frequently discard its most interpretable aspect: namely, the word-level distances required for translating a set of words into another set of words. To address this apparent gap, we introduce WMDecompose: a model and Python library that 1) decomposes document-level distances into their constituent word-level distances, and 2) subsequently clusters words to induce thematic elements, such that useful lexical information is retained and summarized for analysis. To illustrate its potential in a social scientific context, we apply it to a longitudinal social media corpus to explore the interrelationship between conspiracy theories and conservative American discourses. Finally, because of the full WMD model's high time-complexity, we additionally suggest a method of sampling document pairs from large datasets in a reproducible way, with tight bounds that prevent extrapolation of unreliable results due to poor sampling practices.
McKenzie, Grant; Adams, Benjamin
Natural Language Processing in GIScience Applications Book Chapter
In: Wilson, John (Ed.): The GIS&T Body of Knowledge, University Consortium of Geographic Information Science, 2021.
@inbook{McKenzie2021b,
title = {Natural Language Processing in GIScience Applications},
author = {Grant McKenzie and Benjamin Adams},
editor = {John Wilson},
url = {https://grantmckenzie.com/academics/McKenzieUCGISBOK2021.pdf},
year = {2021},
date = {2021-11-08},
urldate = {2021-11-08},
booktitle = {The GIS&T Body of Knowledge},
publisher = {University Consortium of Geographic Information Science},
keywords = {},
pubstate = {published},
tppubtype = {inbook}
}
Chen, Emily; McKenzie, Grant
Mobility Response to COVID-19-related Restrictions in New York City Proceedings Article
In: The 2nd ACM SIGSPATIAL International Workshop on Spatial Computing for Epidemiology (SpatialEpi'21), ACM, 2021, ISSN: 98-1-4503-9119-1/21/11.
@inproceedings{ChenMcKenzie2021,
title = {Mobility Response to COVID-19-related Restrictions in New York City},
author = {Emily Chen and Grant McKenzie},
url = {https://grantmckenzie.com/academics/ChenMcKenzie2021.pdf},
doi = {10.1145/3486633.3491094},
issn = {98-1-4503-9119-1/21/11},
year = {2021},
date = {2021-11-02},
urldate = {2021-11-02},
booktitle = {The 2nd ACM SIGSPATIAL International Workshop on Spatial Computing for Epidemiology (SpatialEpi'21)},
publisher = {ACM},
abstract = {The first case of the 2019 novel coronavirus was detected in the United States in January 2020, and since then, efforts to contain the virus, such as stay-at-home policies, have greatly restricted human mobility. While stay-at-home policies and concern over the virus contributed to an increase in time spent at home, little is known as to how a change in home dwell time varied by population. The work presented in this paper seeks to understand the relationships between levels of mobility and socioeconomic and demographics characteristics of communities within New York City from February to April 2020. Through analyzing the factors that contributed to changes in home dwell time, this work aims to support policymakers and inform future strategies for infection mitigation. Findings from this research reinforce the need for physical distancing policies that acknowledge the existence of socioeconomic and demographic diversity between not only geographic regions in the U.S. but also within a single city. },
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
The first case of the 2019 novel coronavirus was detected in the United States in January 2020, and since then, efforts to contain the virus, such as stay-at-home policies, have greatly restricted human mobility. While stay-at-home policies and concern over the virus contributed to an increase in time spent at home, little is known as to how a change in home dwell time varied by population. The work presented in this paper seeks to understand the relationships between levels of mobility and socioeconomic and demographics characteristics of communities within New York City from February to April 2020. Through analyzing the factors that contributed to changes in home dwell time, this work aims to support policymakers and inform future strategies for infection mitigation. Findings from this research reinforce the need for physical distancing policies that acknowledge the existence of socioeconomic and demographic diversity between not only geographic regions in the U.S. but also within a single city.
McKenzie, Grant; Romm, Daniel
Measuring urban regional similarity through mobility signatures Journal Article
In: Computers, Environment and Urban Systems, vol. 89, pp. 101684, 2021.
@article{McKenzie2021,
title = {Measuring urban regional similarity through mobility signatures},
author = {Grant McKenzie and Daniel Romm},
url = {https://grantmckenzie.com/academics/CitySim_2021.pdf
https://platial.science/citysim/
https://www.sciencedirect.com/science/article/pii/S0198971521000910},
doi = {10.1016/j.compenvurbsys.2021.101684},
year = {2021},
date = {2021-09-01},
journal = {Computers, Environment and Urban Systems},
volume = {89},
pages = {101684},
abstract = {The task of identifying similar regions within and between cities is an important aspect of urban data science as well as applied domains such as real estate, tourism, and urban planning. Regional similarity is typically assessed through comparing socio-demographic variables, resource availability, or urban infrastructure. An essential dimension, often overlooked for this task, is the spatiotemporal mobility patterns of people within a city. In this work we present a novel approach to identifying regional similarity based on human mobility as proxied through micromobility trips. We use a dataset consisting of e-scooter trip origins and destinations for two major European cities that differ in population size and urban structure. Three dimensions of these data are used in modeling the spatial and temporal variability in movement between regions in cities, allowing us to compare regions through a mobility lens. The result is a parameterized similarity model and interactive web platform for comparing regions across different urban environments. The application of this model suggests that human mobility patterns are a quantifiable, unique, and appropriate characteristic through which to measure urban similarity.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
The task of identifying similar regions within and between cities is an important aspect of urban data science as well as applied domains such as real estate, tourism, and urban planning. Regional similarity is typically assessed through comparing socio-demographic variables, resource availability, or urban infrastructure. An essential dimension, often overlooked for this task, is the spatiotemporal mobility patterns of people within a city. In this work we present a novel approach to identifying regional similarity based on human mobility as proxied through micromobility trips. We use a dataset consisting of e-scooter trip origins and destinations for two major European cities that differ in population size and urban structure. Three dimensions of these data are used in modeling the spatial and temporal variability in movement between regions in cities, allowing us to compare regions through a mobility lens. The result is a parameterized similarity model and interactive web platform for comparing regions across different urban environments. The application of this model suggests that human mobility patterns are a quantifiable, unique, and appropriate characteristic through which to measure urban similarity.
McKenzie, Grant; Baez, Carlos
A Spatiotemporal approach to micromobility Book Chapter
In: Sigler, T.; Corcoran, J. (Ed.): A Modern Guide to the Urban Sharing Economy, pp. 195-208, Elgar Publishing, 2021, ISBN: 978 1 78990 955 5.
@inbook{McKenzie2021d,
title = {A Spatiotemporal approach to micromobility},
author = {Grant McKenzie and Carlos Baez},
editor = {T. Sigler and J. Corcoran},
url = {https://grantmckenzie.com/academics/mckenziebaez2021.pdf},
isbn = {978 1 78990 955 5},
year = {2021},
date = {2021-09-01},
urldate = {2021-09-01},
booktitle = {A Modern Guide to the Urban Sharing Economy},
pages = {195-208},
publisher = {Elgar Publishing},
keywords = {},
pubstate = {published},
tppubtype = {inbook}
}
2020
McKenzie, Grant; Adams, Benjamin
A country comparison of place-based activity response to COVID-19 policies Journal Article
In: Applied Geography, vol. 125, no. 2020, pp. 102363, 2020.
@article{McKenzie2020,
title = {A country comparison of place-based activity response to COVID-19 policies},
author = {Grant McKenzie and Benjamin Adams},
url = {https://grantmckenzie.com/academics/McKenzieAdams2020.pdf},
doi = {10.1016/j.apgeog.2020.102363},
year = {2020},
date = {2020-05-19},
journal = {Applied Geography},
volume = {125},
number = {2020},
pages = {102363},
organization = {arXiv.org},
abstract = {The emergence of the novel Coronavirus Disease in late 2019 (COVID-19) and subsequent pandemic led to an immense disruption in the daily lives of almost everyone on the planet. Faced with the consequences of inaction, most national governments responded with policies that restricted the activities conducted by their inhabitants. As schools and businesses shuttered, the mobility of these people decreased. This reduction in mobility, and related activities, was recorded through ubiquitous location-enabled personal mobile devices. Patterns emerged that varied by place-based activity. In this work the differences in these place-based activity patterns are investigated across nations, specifically focusing on the relationship between government enacted policies and changes in community activity patterns. We show that people's activity response to government action varies widely both across nations as well as regionally within them. Three assessment measures are devised and the results correlate with a number of global indices. We discuss these findings and the relationship between government action and residents' response.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
The emergence of the novel Coronavirus Disease in late 2019 (COVID-19) and subsequent pandemic led to an immense disruption in the daily lives of almost everyone on the planet. Faced with the consequences of inaction, most national governments responded with policies that restricted the activities conducted by their inhabitants. As schools and businesses shuttered, the mobility of these people decreased. This reduction in mobility, and related activities, was recorded through ubiquitous location-enabled personal mobile devices. Patterns emerged that varied by place-based activity. In this work the differences in these place-based activity patterns are investigated across nations, specifically focusing on the relationship between government enacted policies and changes in community activity patterns. We show that people's activity response to government action varies widely both across nations as well as regionally within them. Three assessment measures are devised and the results correlate with a number of global indices. We discuss these findings and the relationship between government action and residents' response.
McKenzie, Grant
Urban mobility in the sharing economy: A spatiotemporal comparison of shared mobility services Journal Article
In: Computers, Environment and Urban Systems, vol. 79, pp. 101418, 2020.
@article{McKenzieCEUS2019,
title = {Urban mobility in the sharing economy: A spatiotemporal comparison of shared mobility services},
author = {Grant McKenzie },
url = {https://www.grantmckenzie.com/academics/McKenzie_CEUS2019.pdf},
year = {2020},
date = {2020-01-01},
journal = {Computers, Environment and Urban Systems},
volume = {79},
pages = {101418},
abstract = {The influx of micro-mobility services, such as dockless scooter-share and e-bikes, in many cities are contributing to a substantial change in urban transportation with adoption rates reminiscent of other shared-mobility services, such as ride-hailing, years prior. Touted as a solution to the last mile problem, a multitude of micro-mobility companies have situated themselves in urban centers promising low cost alternative transportation options for short, urban travel. The rapid arrival of these companies, however, has left little time for city officials, transportation planners, and citizens to assess the demand for these services and compare them to existing transportation options. In this work, we investigate two key aspects of these micro-mobility services. First, we identify the spatial and temporal differences between these mobility companies and highlight the nuanced differences in usage patterns. Second, we compare these new services to an existing mode of transportation, namely automobile-based ride-hailing, with regards to differences in travel time within a city. The results of these analyses indicate that while many micro-mobility companies are spatiotemporally similar, there are notable differences in where and when these services are used. Similarly, we find that automobile travel is not always the fastest means of transportation within an urban setting. During periods of heavy traffic congestion, e.g., rush hour, micro-mobility services offer a faster means of travel within the city. The findings presented in this work offer evidence on which to inform urban planning and transportation policy with respect to shared mobility services, free floating vehicles, and alternative urban transportation.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
The influx of micro-mobility services, such as dockless scooter-share and e-bikes, in many cities are contributing to a substantial change in urban transportation with adoption rates reminiscent of other shared-mobility services, such as ride-hailing, years prior. Touted as a solution to the last mile problem, a multitude of micro-mobility companies have situated themselves in urban centers promising low cost alternative transportation options for short, urban travel. The rapid arrival of these companies, however, has left little time for city officials, transportation planners, and citizens to assess the demand for these services and compare them to existing transportation options. In this work, we investigate two key aspects of these micro-mobility services. First, we identify the spatial and temporal differences between these mobility companies and highlight the nuanced differences in usage patterns. Second, we compare these new services to an existing mode of transportation, namely automobile-based ride-hailing, with regards to differences in travel time within a city. The results of these analyses indicate that while many micro-mobility companies are spatiotemporally similar, there are notable differences in where and when these services are used. Similarly, we find that automobile travel is not always the fastest means of transportation within an urban setting. During periods of heavy traffic congestion, e.g., rush hour, micro-mobility services offer a faster means of travel within the city. The findings presented in this work offer evidence on which to inform urban planning and transportation policy with respect to shared mobility services, free floating vehicles, and alternative urban transportation.
2019
McKenzie, Grant
Shared micro-mobility patterns as measures of city similarity Proceedings Article
In: van Kreveld, Marc; Speckmann, Bettina; Stroila, Matei; Trajcevski, Goce (Ed.): 1st ACM SIGSPATIAL International Workshop on Computing with Multifaceted Movement Data (MOVE++ 2019) in conjunction with the 27th ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems., ACM, 2019, ISBN: 78-1-4503-6951-0/19/11.
@inproceedings{Move2019,
title = {Shared micro-mobility patterns as measures of city similarity},
author = {Grant McKenzie },
editor = {Marc van Kreveld and Bettina Speckmann and Matei Stroila and Goce Trajcevski},
url = {https://grantmckenzie.com/academics/McKenzie_Move2019.pdf},
doi = {https://doi.org/10.1145/3356392.3365221},
isbn = {78-1-4503-6951-0/19/11},
year = {2019},
date = {2019-11-05},
booktitle = {1st ACM SIGSPATIAL International Workshop on Computing with Multifaceted Movement Data (MOVE++ 2019) in conjunction with the 27th ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems.},
publisher = {ACM},
abstract = {Micro-mobility services, such as dockless e-scooters and e-bikes, are inundating urban centers around the world. The mass adoption of these services, and ubiquity of the companies operating them, offer a unique opportunity through which to compare cities. In this position paper, a series of spatiotemporal measures are proposed based on activity data collected from shared micro-mobility services. The purpose of this paper is to identify a number of ways that these new mobility services can serve to augment existing city similarity approaches.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Micro-mobility services, such as dockless e-scooters and e-bikes, are inundating urban centers around the world. The mass adoption of these services, and ubiquity of the companies operating them, offer a unique opportunity through which to compare cities. In this position paper, a series of spatiotemporal measures are proposed based on activity data collected from shared micro-mobility services. The purpose of this paper is to identify a number of ways that these new mobility services can serve to augment existing city similarity approaches.
McKenzie, Grant
Spatiotemporal comparative analysis of scooter-share and bike-share usage patterns in Washington, D.C. Journal Article
In: Journal of Transport Geography, vol. 78, pp. 19-28, 2019.
@article{80,
title = {Spatiotemporal comparative analysis of scooter-share and bike-share usage patterns in Washington, D.C.},
author = {Grant McKenzie},
url = {https://www.grantmckenzie.com/academics/McKenzie_JTG2019.pdf},
doi = {10.1016/j.jtrangeo.2019.05.007},
year = {2019},
date = {2019-06-01},
journal = {Journal of Transport Geography},
volume = {78},
pages = {19-28},
chapter = {19},
abstract = {The United States is currently in the midst of a micro-mobility revolution of sorts. Almost overnight, U.S. cities have been inundated with short-term rental scooters owned and operated by start-up companies promising a disruption to the urban transportation status-quo. These scooter-share services are presented as a dockless alternative to traditionally government-funded, docking station-based bike-sharing programs. Given the rapid rise of electric scooter companies, and how little is known about their operations, there is pressing public interest in understanding the impact of these transportation-sharing platforms. By exploring the nuanced spatial and temporal activity patterns of each of these platforms, this research identifies differences and similarities between dockless e-scooters and existing bike-sharing services. The findings from this research contribute to our understanding of urban transportation behavior and differences within mobility platforms.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
The United States is currently in the midst of a micro-mobility revolution of sorts. Almost overnight, U.S. cities have been inundated with short-term rental scooters owned and operated by start-up companies promising a disruption to the urban transportation status-quo. These scooter-share services are presented as a dockless alternative to traditionally government-funded, docking station-based bike-sharing programs. Given the rapid rise of electric scooter companies, and how little is known about their operations, there is pressing public interest in understanding the impact of these transportation-sharing platforms. By exploring the nuanced spatial and temporal activity patterns of each of these platforms, this research identifies differences and similarities between dockless e-scooters and existing bike-sharing services. The findings from this research contribute to our understanding of urban transportation behavior and differences within mobility platforms.
McKenzie, Grant; Slind, Todd R
A user-generated data based approach to enhancing location prediction of financial services in sub-Saharan Africa Journal Article
In: Applied Geography, 2019.
@article{71,
title = {A user-generated data based approach to enhancing location prediction of financial services in sub-Saharan Africa},
author = {Grant McKenzie and Todd R Slind},
url = {https://www.sciencedirect.com/science/article/pii/S0143622818302078},
doi = {10.1016/j.apgeog.2019.02.005},
year = {2019},
date = {2019-01-01},
journal = {Applied Geography},
abstract = {The recent increase in user-generated content and social media adoption in developing countries offers an unprecedented opportunity to better understand the accessibility and spatial distribution of financial services in sub-Saharan Africa. Financial inclusion has been identified as a priority by multiple agencies in the region and on-the-ground efforts are currently underway to identify previously unknown financial access points in numerous developing African countries. Existing techniques for estimating the location of these access points rely on spatial analysis of often outdated or unsuitable publicly available datasets such as population density, road networks, etc. as well as expensive and time consuming surveys of locals in the region. In this work we propose an approach to augment existing spatial data analysis techniques through the inclusion of user-generated geo-content and geo-social media data. Though a comparison of standard regression models and machine learning techniques, this work proposes the use of alternative data sources to build prediction models for identifying financial access locations in countries where current estimation models are insufficient. Through a better understanding of geospatial distribution patterns this work aims at reducing data acquisition costs and providing decision makers with critical data more quickly and efficiently. Finally, we present a mobile application build on the outcomes of this analysis that is currently being used to better inform on-the-ground data collection efforts.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
The recent increase in user-generated content and social media adoption in developing countries offers an unprecedented opportunity to better understand the accessibility and spatial distribution of financial services in sub-Saharan Africa. Financial inclusion has been identified as a priority by multiple agencies in the region and on-the-ground efforts are currently underway to identify previously unknown financial access points in numerous developing African countries. Existing techniques for estimating the location of these access points rely on spatial analysis of often outdated or unsuitable publicly available datasets such as population density, road networks, etc. as well as expensive and time consuming surveys of locals in the region. In this work we propose an approach to augment existing spatial data analysis techniques through the inclusion of user-generated geo-content and geo-social media data. Though a comparison of standard regression models and machine learning techniques, this work proposes the use of alternative data sources to build prediction models for identifying financial access locations in countries where current estimation models are insufficient. Through a better understanding of geospatial distribution patterns this work aims at reducing data acquisition costs and providing decision makers with critical data more quickly and efficiently. Finally, we present a mobile application build on the outcomes of this analysis that is currently being used to better inform on-the-ground data collection efforts.
2018
McKenzie, Grant; Liu, Zheng; Hu, Yingjie; Lee, Myeong
Identifying urban neighborhood names through user-contributed online property listings Journal Article
In: ISPRS International Journal of Geo-Information, 2018.
@article{65,
title = {Identifying urban neighborhood names through user-contributed online property listings},
author = {Grant McKenzie and Zheng Liu and Yingjie Hu and Myeong Lee},
url = {https://grantdmckenzie.com/academics/McKenzie_Neighborhoods2018.pdf},
year = {2018},
date = {2018-10-01},
journal = {ISPRS International Journal of Geo-Information},
abstract = {Neighborhoods are vaguely defined, localized regions that share similar characteristics. They are most often defined, delineated, and named by the citizens that inhabit them rather than municipal government or commercial agencies. The names of these neighborhoods play an important role as a basis for community and sociodemographic identity, geographic communication, and historical context. In this work we take a data-driven approach to identifying neighborhood names based on the geospatial properties of user-contributed rental listings. Through a random forest ensemble learning model applied to a set of spatial statistics for all n-grams in listing descriptions, we show that neighborhood names can be uniquely identified within urban settings. We train a model based on data from Washington, DC and test it on listings in Seattle, WA and Montreal, QC. The results indicate that a model trained on housing data from one city can successfully identify neighborhood names in another. In addition, our approach identifies less common neighborhood names and suggestions alternative or potentially new names in each city. These findings represent a first step in the process of urban neighborhood identification and delineation.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Neighborhoods are vaguely defined, localized regions that share similar characteristics. They are most often defined, delineated, and named by the citizens that inhabit them rather than municipal government or commercial agencies. The names of these neighborhoods play an important role as a basis for community and sociodemographic identity, geographic communication, and historical context. In this work we take a data-driven approach to identifying neighborhood names based on the geospatial properties of user-contributed rental listings. Through a random forest ensemble learning model applied to a set of spatial statistics for all n-grams in listing descriptions, we show that neighborhood names can be uniquely identified within urban settings. We train a model based on data from Washington, DC and test it on listings in Seattle, WA and Montreal, QC. The results indicate that a model trained on housing data from one city can successfully identify neighborhood names in another. In addition, our approach identifies less common neighborhood names and suggestions alternative or potentially new names in each city. These findings represent a first step in the process of urban neighborhood identification and delineation.
McKenzie, Grant
Docked vs. Dockless Bike-sharing: Contrasting Spatiotemporal Patterns Conference
The 10th International Conference on Geographic Information Science, Schloss Dagstuhl Schloss Dagstuhl, Melbourne, Australia, 2018.
@conference{57,
title = {Docked vs. Dockless Bike-sharing: Contrasting Spatiotemporal Patterns},
author = {Grant McKenzie},
url = {http://www.grantmckenzie.com/academics/Dockless2018.pdf},
doi = {10.4230/LIPIcs.GIScience.2018.64},
year = {2018},
date = {2018-08-01},
booktitle = {The 10th International Conference on Geographic Information Science},
publisher = {Schloss Dagstuhl},
address = {Melbourne, Australia},
organization = {Schloss Dagstuhl},
abstract = {U.S. urban centers are currently experiencing explosive growth in commercial dockless bike-sharing services. Tens of thousands of bikes have shown up across the country in recent months providing limited time for municipal governments to set regulations or assess their impact on government-funded dock-based bike-sharing programs. Washington, D.C. offers an unprecedented opportunity to examine the activity patterns of both docked and dockless bike-sharing services given the history of bike-sharing in the city and the recent availability of dockless bike data. This work presents an exploratory step in understanding how dockless bike-sharing services are being used within a city and the ways in which the activity patterns differ from traditional dock station-based programs.},
keywords = {},
pubstate = {published},
tppubtype = {conference}
}
U.S. urban centers are currently experiencing explosive growth in commercial dockless bike-sharing services. Tens of thousands of bikes have shown up across the country in recent months providing limited time for municipal governments to set regulations or assess their impact on government-funded dock-based bike-sharing programs. Washington, D.C. offers an unprecedented opportunity to examine the activity patterns of both docked and dockless bike-sharing services given the history of bike-sharing in the city and the recent availability of dockless bike data. This work presents an exploratory step in understanding how dockless bike-sharing services are being used within a city and the ways in which the activity patterns differ from traditional dock station-based programs.
McKenzie, Grant; Janowicz, Krzysztof
OpenPOI: An Open Place of Interest Platform Conference
The 10th International Conference on Geographic Information Science, Schloss Dagstuhl Schloss Dagstuhl, Melbourne, Australia, 2018.
@conference{56,
title = {OpenPOI: An Open Place of Interest Platform},
author = {Grant McKenzie and Krzysztof Janowicz},
url = {http://www.grantmckenzie.com/academics/OpenPOI2018.pdf},
doi = {10.4230/LIPIcs.GIScience.2018.65},
year = {2018},
date = {2018-08-01},
booktitle = {The 10th International Conference on Geographic Information Science},
publisher = {Schloss Dagstuhl},
address = {Melbourne, Australia},
organization = {Schloss Dagstuhl},
abstract = {Places of Interest (POI) are a principal component of how human behavior is captured in todaytextquoterights geographic information. Increasingly, access to POI datasets are being restricted -- even silo-ed -- for commercial use, with vendors often impeding access to the very users that contribute the data. Open mapping platforms such as OpenStreetMap (OSM) offer access to a plethora of geospatial data though they can be limited in the attribute resolution or range of information associated with the data. Nuanced descriptive information associated with POI, e.g., ambience, are not captured by such platforms. Furthermore, interactions with a POI, such as checking in, or recommending a menu item, are inherently place-based concepts. Many of these interactions occur with high temporal volatility that involves frequent interaction with a platform, arguably inappropriate for the textquotelefttextquoteleftchangesettextquoterighttextquoteright model adopted by OSM and related datasets. In this short paper we propose OpenPOI, an open platform for storing, serving, and interacting with places of interests and the activities they afford.},
keywords = {},
pubstate = {published},
tppubtype = {conference}
}
Places of Interest (POI) are a principal component of how human behavior is captured in todaytextquoterights geographic information. Increasingly, access to POI datasets are being restricted -- even silo-ed -- for commercial use, with vendors often impeding access to the very users that contribute the data. Open mapping platforms such as OpenStreetMap (OSM) offer access to a plethora of geospatial data though they can be limited in the attribute resolution or range of information associated with the data. Nuanced descriptive information associated with POI, e.g., ambience, are not captured by such platforms. Furthermore, interactions with a POI, such as checking in, or recommending a menu item, are inherently place-based concepts. Many of these interactions occur with high temporal volatility that involves frequent interaction with a platform, arguably inappropriate for the textquotelefttextquoteleftchangesettextquoterighttextquoteright model adopted by OSM and related datasets. In this short paper we propose OpenPOI, an open platform for storing, serving, and interacting with places of interests and the activities they afford.
McKenzie, Grant; Adams, Benjamin
A data-driven approach to exploring similarities of tourist attractions through online reviews Journal Article
In: Journal of Location Based Services, vol. 12, pp. 94-118, 2018.
@article{59,
title = {A data-driven approach to exploring similarities of tourist attractions through online reviews},
author = {Grant McKenzie and Benjamin Adams},
url = {http://www.grantmckenzie.com/academics/McKenzieAdams_Tourism2018.pdf},
year = {2018},
date = {2018-07-01},
journal = {Journal of Location Based Services},
volume = {12},
pages = {94-118},
chapter = {94},
abstract = {The motivation for tourists to visit a city is often driven by the uniqueness of the attractions accessible within the region. The draw to these locations varies by visitor as some travelers are interested in a single specific attraction while others prefer thematic travel. Tourists today have access to detailed experiences of other visitors to these locations in the form of user-contributed text reviews, opinions, photographs, and videos, all contributed through online tourism platforms. The data available through these platforms offer a unique opportunity to examine the similarities and difference between these attractions, their cities, and the visitors that contribute the reviews. In this work we take a data-driven approach to assessing similarity through textual analysis of user-contributed reviews, uncovering nuanced differences and similarities in the ways that reviewers write about attractions and cities.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
The motivation for tourists to visit a city is often driven by the uniqueness of the attractions accessible within the region. The draw to these locations varies by visitor as some travelers are interested in a single specific attraction while others prefer thematic travel. Tourists today have access to detailed experiences of other visitors to these locations in the form of user-contributed text reviews, opinions, photographs, and videos, all contributed through online tourism platforms. The data available through these platforms offer a unique opportunity to examine the similarities and difference between these attractions, their cities, and the visitors that contribute the reviews. In this work we take a data-driven approach to assessing similarity through textual analysis of user-contributed reviews, uncovering nuanced differences and similarities in the ways that reviewers write about attractions and cities.
Fu, Cheng; McKenzie, Grant; Frias-Martinez, Vanessa; Stewart, Kathleen
Identifying spatiotemporal urban activities through linguistic signatures Journal Article
In: Computers, Environment and Urban Systems, 2018, ISSN: 0198-9715.
@article{63,
title = {Identifying spatiotemporal urban activities through linguistic signatures},
author = {Cheng Fu and Grant McKenzie and Vanessa Frias-Martinez and Kathleen Stewart},
url = {https://www.sciencedirect.com/science/article/pii/S0198971517303472},
doi = {10.1016/j.compenvurbsys.2018.07.003},
issn = {0198-9715},
year = {2018},
date = {2018-07-01},
journal = {Computers, Environment and Urban Systems},
abstract = {Identifying the activities that individuals conduct in a city is key to understanding urban dynamics. It is difficult, however, to identify different human activities on a large scale without incurring significant costs. This study focuses on modeling the spatiotemporal patterns of different activity types within cities by employing user-contributed, geosocial content as a proxy for human activities. In this work, we use linguistic topic modeling to analyze georeferenced twitter data in order to differentiate different activity types. We then examine the spatial and temporal patterns of the derived activity types in three U.S. cities: Baltimore, MD., Washington, D.C., and New York City, NY. The linguistic patterns reflect the spatiotemporal context of the places where the social media content is posted. We further construct a method to link what people post online to the activities conducted within a city. We then use these derived activities to profile the characteristics of neighborhoods in the three cities, and apply the activity signatures to discover similar neighborhoods both within and between the cities. This approach represents a novel activity-based method for assessing similarity between neighborhoods.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Identifying the activities that individuals conduct in a city is key to understanding urban dynamics. It is difficult, however, to identify different human activities on a large scale without incurring significant costs. This study focuses on modeling the spatiotemporal patterns of different activity types within cities by employing user-contributed, geosocial content as a proxy for human activities. In this work, we use linguistic topic modeling to analyze georeferenced twitter data in order to differentiate different activity types. We then examine the spatial and temporal patterns of the derived activity types in three U.S. cities: Baltimore, MD., Washington, D.C., and New York City, NY. The linguistic patterns reflect the spatiotemporal context of the places where the social media content is posted. We further construct a method to link what people post online to the activities conducted within a city. We then use these derived activities to profile the characteristics of neighborhoods in the three cities, and apply the activity signatures to discover similar neighborhoods both within and between the cities. This approach represents a novel activity-based method for assessing similarity between neighborhoods.
Adams, Benjamin; McKenzie, Grant
Crowdsourcing the Character of a Place: Character-Level Convolutional Networks for Multilingual Geographic Text Classification Journal Article
In: Transactions in GIS, 2018.
@article{46,
title = {Crowdsourcing the Character of a Place: Character-Level Convolutional Networks for Multilingual Geographic Text Classification},
author = {Benjamin Adams and Grant McKenzie},
url = {http://www.grantmckenzie.com/academics/CharacterOfPlace_2017.pdf},
doi = {10.1111/tgis.12317},
year = {2018},
date = {2018-01-01},
journal = {Transactions in GIS},
abstract = {This article presents a new character-level convolutional neural network model that can classify multilingual text written using any character set that can be encoded with UTF-8, a standard and widely used 8-bit character encoding. For geographic classification of text, we demonstrate that this approach is competitive with state-of-the-art word-based text classification methods. The model was tested on four crowdsourced data sets made up of Wikipedia articles, online travel blogs, Geonames toponyms, and Twitter posts. Unlike word-based methods, which require data cleaning and pre-processing, the proposed model works for any language without modification and with classification accuracy comparable to existing methods. Using a synthetic data set with introduced character-level errors, we show it is more robust to noise than word-level classification algorithms. The results indicate that UTF-8 character-level convolutional neural networks are a promising technique for georeferencing noisy text, such as found in colloquial social media posts and texts scanned with optical character recognition. However, word-based methods currently require less computation time to train, so are currently preferable for classifying well-formatted and cleaned texts in single languages.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
This article presents a new character-level convolutional neural network model that can classify multilingual text written using any character set that can be encoded with UTF-8, a standard and widely used 8-bit character encoding. For geographic classification of text, we demonstrate that this approach is competitive with state-of-the-art word-based text classification methods. The model was tested on four crowdsourced data sets made up of Wikipedia articles, online travel blogs, Geonames toponyms, and Twitter posts. Unlike word-based methods, which require data cleaning and pre-processing, the proposed model works for any language without modification and with classification accuracy comparable to existing methods. Using a synthetic data set with introduced character-level errors, we show it is more robust to noise than word-level classification algorithms. The results indicate that UTF-8 character-level convolutional neural networks are a promising technique for georeferencing noisy text, such as found in colloquial social media posts and texts scanned with optical character recognition. However, word-based methods currently require less computation time to train, so are currently preferable for classifying well-formatted and cleaned texts in single languages.
Hu, Yingjie; Mao, Huina; McKenzie, Grant
A Natural Language Processing and Geospatial Clustering Framework for Harvesting Local Place Names from Geotagged Housing Advertisements Journal Article
In: International Journal of Geographical Information Science, 2018.
@article{50,
title = {A Natural Language Processing and Geospatial Clustering Framework for Harvesting Local Place Names from Geotagged Housing Advertisements},
author = {Yingjie Hu and Huina Mao and Grant McKenzie},
year = {2018},
date = {2018-01-01},
journal = {International Journal of Geographical Information Science},
abstract = {Local place names are frequently used by residents living in a geographic region. Such place names may not be recorded in existing gazetteers, due to their vernacular nature, relative insignificance to a gazetteer covering a large area (e.g., the entire world), recent establishment (e.g., the name of a newly-opened shopping center), or other reasons. While not always recorded, local place names play important roles in many applications, from supporting public participation in urban planning to locating victims in disaster response. In this paper, we propose a computational framework for harvesting local place names from geotagged housing advertisements. We make use of those advertisements posted on local-oriented websites, such as Craigslist, where local place names are often mentioned. The proposed framework consists of two stages: natural language processing (NLP) and geospatial clustering. The NLP stage examines the textual content of housing advertisements, and extracts place name candidates. The geospatial stage focuses on the coordinates associated with the extracted place name candidates, and performs multi-scale geospatial clustering to filter out the non-place names. We evaluate our framework by comparing its performance with those of six baselines. We also compare our result with four existing gazetteers to demonstrate the not-yet-recorded local place names discovered by our framework.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Local place names are frequently used by residents living in a geographic region. Such place names may not be recorded in existing gazetteers, due to their vernacular nature, relative insignificance to a gazetteer covering a large area (e.g., the entire world), recent establishment (e.g., the name of a newly-opened shopping center), or other reasons. While not always recorded, local place names play important roles in many applications, from supporting public participation in urban planning to locating victims in disaster response. In this paper, we propose a computational framework for harvesting local place names from geotagged housing advertisements. We make use of those advertisements posted on local-oriented websites, such as Craigslist, where local place names are often mentioned. The proposed framework consists of two stages: natural language processing (NLP) and geospatial clustering. The NLP stage examines the textual content of housing advertisements, and extracts place name candidates. The geospatial stage focuses on the coordinates associated with the extracted place name candidates, and performs multi-scale geospatial clustering to filter out the non-place names. We evaluate our framework by comparing its performance with those of six baselines. We also compare our result with four existing gazetteers to demonstrate the not-yet-recorded local place names discovered by our framework.
2017
Kessler, Carsten; McKenzie, Grant
A Geoprivacy Manifesto Journal Article
In: Transactions in GIS, vol. 22, pp. 3-19, 2017.
@article{40,
title = {A Geoprivacy Manifesto},
author = {Carsten Kessler and Grant McKenzie},
url = {http://grantmckenzie.com/academics/GeoprivacyManifesto2017.pdf},
doi = {10.1111/tgis.12305},
year = {2017},
date = {2017-12-01},
journal = {Transactions in GIS},
volume = {22},
pages = {3-19},
abstract = {As location-enabled technologies are becoming ubiquitous, our location is being shared with an ever-growing number of external services. Issues revolving around location privacy – or geoprivacy – therefore concern the vast majority of the population, largely without knowing how the underlying technologies work and what can be inferred from an individualtextquoterights location, especially if recorded over longer periods of time. Research, on the other hand, has largely treated this topic from isolated standpoints, most prominently from the technological and ethical point of view. This article therefore reflects upon the current state of geoprivacy from a broader perspective. It integrates technological, ethical, legal, and educational aspects and clarifies how they interact and shape how we deal with the corresponding technology, both individually and as a society. It does so in the form of a manifesto, consisting of 21 theses that summarize the main arguments made in the article. These theses argue that location information is different from other kinds of personal information and, in combination, show why geoprivacy (and privacy in general) needs to be protected and should not become a mere illusion. The fictional couple of Jane and Tom is used as a running example to illustrate how common it has become to share our location information, and how it can be used – both for good and for worse.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
As location-enabled technologies are becoming ubiquitous, our location is being shared with an ever-growing number of external services. Issues revolving around location privacy – or geoprivacy – therefore concern the vast majority of the population, largely without knowing how the underlying technologies work and what can be inferred from an individualtextquoterights location, especially if recorded over longer periods of time. Research, on the other hand, has largely treated this topic from isolated standpoints, most prominently from the technological and ethical point of view. This article therefore reflects upon the current state of geoprivacy from a broader perspective. It integrates technological, ethical, legal, and educational aspects and clarifies how they interact and shape how we deal with the corresponding technology, both individually and as a society. It does so in the form of a manifesto, consisting of 21 theses that summarize the main arguments made in the article. These theses argue that location information is different from other kinds of personal information and, in combination, show why geoprivacy (and privacy in general) needs to be protected and should not become a mere illusion. The fictional couple of Jane and Tom is used as a running example to illustrate how common it has become to share our location information, and how it can be used – both for good and for worse.
Janowicz, Krzysztof; McKenzie, Grant
How "Alternative" are Alternative Facts? Measuring Statement Coherence via Spatial Analysis Conference
25th ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems (ACM SIGSPATIAL 2017), ACM ACM, Redondo Beach, CA, USA, 2017.
@conference{36,
title = {How "Alternative" are Alternative Facts? Measuring Statement Coherence via Spatial Analysis},
author = {Krzysztof Janowicz and Grant McKenzie},
url = {http://grantmckenzie.com/academics/AlternativeFacts_ACM2017.pdf},
year = {2017},
date = {2017-11-01},
booktitle = {25th ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems (ACM SIGSPATIAL 2017)},
publisher = {ACM},
address = {Redondo Beach, CA, USA},
organization = {ACM},
abstract = {Following the AAA principle by which anybody can say anything about any topic, the Web is no stranger to alternative facts. Nonetheless, with the increasing volume and velocity at which content is being published and difficulties to assess the credibility of information and the trustworthiness of sources, alternative facts are becoming a major challenge and an instrument for spreading disinformation. Interestingly, the diversity of todaytextquoterights data sources can also help us to counter alternative facts by measuring their coherence, i.e., the degree to which data from one source confirms or contradict data from another source. While a single dataset can be biased towards supporting or discrediting a statement, the diverse sources of data across media types that are publicly accessible today offer unique perspectives on which to assess a given statement. To give an intuitive example, a statement about the comparison of crowd sizes should align with photos of said crowds. However, these photos could be taken at different times, from different viewpoints, and could lead to different, sample-based estimations. Adding further data from heterogeneous sources, such as metro ridership, can either further support a statement or contradict it. In this thought experiment we discuss the role of geographic data, knowledge graphs, and spatial analysis in approaching alternative facts from a novel angle, namely by studying their coherence, i.e., whether they align with other statements, instead of trying to falsify them. In doing so, we aim at increasing the costs for maintaining alternative facts.},
keywords = {},
pubstate = {published},
tppubtype = {conference}
}
Following the AAA principle by which anybody can say anything about any topic, the Web is no stranger to alternative facts. Nonetheless, with the increasing volume and velocity at which content is being published and difficulties to assess the credibility of information and the trustworthiness of sources, alternative facts are becoming a major challenge and an instrument for spreading disinformation. Interestingly, the diversity of todaytextquoterights data sources can also help us to counter alternative facts by measuring their coherence, i.e., the degree to which data from one source confirms or contradict data from another source. While a single dataset can be biased towards supporting or discrediting a statement, the diverse sources of data across media types that are publicly accessible today offer unique perspectives on which to assess a given statement. To give an intuitive example, a statement about the comparison of crowd sizes should align with photos of said crowds. However, these photos could be taken at different times, from different viewpoints, and could lead to different, sample-based estimations. Adding further data from heterogeneous sources, such as metro ridership, can either further support a statement or contradict it. In this thought experiment we discuss the role of geographic data, knowledge graphs, and spatial analysis in approaching alternative facts from a novel angle, namely by studying their coherence, i.e., whether they align with other statements, instead of trying to falsify them. In doing so, we aim at increasing the costs for maintaining alternative facts.
McKenzie, Grant; Adams, Benjamin
Juxtaposing thematic regions derived from spatial and platial user-generated content. Conference
Proceedings of the 13th International Conference on Spatial Information Theory (COSIT textquoteright17), Dagstuhl Dagstuhl, LtextquoterightAquila, Italy, 2017.
@conference{37,
title = {Juxtaposing thematic regions derived from spatial and platial user-generated content.},
author = {Grant McKenzie and Benjamin Adams},
url = {http://grantmckenzie.com/academics/ThematicRegions_2017.pdf},
year = {2017},
date = {2017-09-01},
booktitle = {Proceedings of the 13th International Conference on Spatial Information Theory (COSIT textquoteright17)},
publisher = {Dagstuhl},
address = {LtextquoterightAquila, Italy},
organization = {Dagstuhl},
abstract = {Typical approaches to defining regions, districts or neighborhoods within a city often focus on place instances of a similar type that are grouped together. For example, most cities have at least one bar district defined as such by the clustering of bars within a few city blocks. In reality, it is not the presence of spatial locations labeled as bars that contribute to a bar region, but rather the popularity of the bars themselves. Following the principle that places, and by extension, placetype regions exist via the people that have given space meaning, we explore user-contributed content as a way of extracting this meaning. Kernel density estimation models of place-based social check-ins are compared to spatially tagged social posts with the goal of identifying thematic regions within the city of Los Angeles, CA. Dynamic human activity patterns, represented as temporal signatures, are included in this analysis to demonstrate how regions change over time.},
keywords = {},
pubstate = {published},
tppubtype = {conference}
}
Typical approaches to defining regions, districts or neighborhoods within a city often focus on place instances of a similar type that are grouped together. For example, most cities have at least one bar district defined as such by the clustering of bars within a few city blocks. In reality, it is not the presence of spatial locations labeled as bars that contribute to a bar region, but rather the popularity of the bars themselves. Following the principle that places, and by extension, placetype regions exist via the people that have given space meaning, we explore user-contributed content as a way of extracting this meaning. Kernel density estimation models of place-based social check-ins are compared to spatially tagged social posts with the goal of identifying thematic regions within the city of Los Angeles, CA. Dynamic human activity patterns, represented as temporal signatures, are included in this analysis to demonstrate how regions change over time.
Lee, Myeong; McKenzie, Grant; Aghi, Rajat
Exploratory cluster analysis of urban mobility patterns to identify neighborhood boundaries Conference
International Symposium on Location-Based Social Media Data and Tracking Data., Washington, D.C., 2017.
@conference{52,
title = {Exploratory cluster analysis of urban mobility patterns to identify neighborhood boundaries},
author = {Myeong Lee and Grant McKenzie and Rajat Aghi},
url = {http://www.grantmckenzie.com/academics/UrbanMobilityPatterns_2017.pdf},
year = {2017},
date = {2017-07-01},
booktitle = {International Symposium on Location-Based Social Media Data and Tracking Data.},
address = {Washington, D.C.},
abstract = {Defining neighborhood boundaries within a city is a complex and often subjective task. Neighborhoods boundaries are defined by the people that visit and live in the region, and activities that occur within those boundaries. Depending on the individual or group activity being conducted, these boundaries can change substan- tially. Transportation and human mobility patterns offer a novel basis on which to explore and delineate neighborhoods. In this work we take a first, exploratory step in capturing dynamically changing neighborhoods based on two different types of urban mobility data. Through clustering temporal urban mobility signatures of alternative transportation users in Washington, D.C., this work provides implications about the characteristics of different types of mobility data and research directions.},
keywords = {},
pubstate = {published},
tppubtype = {conference}
}
Defining neighborhood boundaries within a city is a complex and often subjective task. Neighborhoods boundaries are defined by the people that visit and live in the region, and activities that occur within those boundaries. Depending on the individual or group activity being conducted, these boundaries can change substan- tially. Transportation and human mobility patterns offer a novel basis on which to explore and delineate neighborhoods. In this work we take a first, exploratory step in capturing dynamically changing neighborhoods based on two different types of urban mobility data. Through clustering temporal urban mobility signatures of alternative transportation users in Washington, D.C., this work provides implications about the characteristics of different types of mobility data and research directions.
McKenzie, Grant; Janowicz, Krzysztof
Societal Geo-innovation. AGILE 2017, Springer Springer, Wageningen, Netherlands, 2017.
@conference{38,
title = {The Effect of Regional Variation and Resolution on Geosocial Thematic Signatures for Points of Interest},
author = {Grant McKenzie and Krzysztof Janowicz},
url = {http://grantmckenzie.com/academics/McKenzie_AGILE2017.pdf},
year = {2017},
date = {2017-05-01},
booktitle = {Societal Geo-innovation. AGILE 2017},
publisher = {Springer},
address = {Wageningen, Netherlands},
organization = {Springer},
abstract = {Computational models of place are a key component of spatial information theory and play an increasing role in research ranging from spatial search to transportation studies. One method to arrive at such models is to extract knowledge from user-generated content e.g., from texts, tags, trajectories, pictures, and so forth. Over the last years, topic modeling techniques such as latent Dirichlet allocation (LDA) have been studied to reveal linguistic patterns that characterize places and their types. Intuitively, people are more likely to describe places such as Yosemite National Park in terms of hiking, nature, and camping than cocktail or dancing. The geo-indicativeness of non-georeferenced text does not only apply to place instances but also place types, e.g., state parks. While different parks will vary greatly with respect to their landscape and thus human descriptions, the distribution of topics common to all parks will differ significantly from other types of places, e.g., night clubs. This aggregation of topics to the type level creates thematic signatures that can be used for place categorization, data cleansing and conflation, semantic search, and so on. To make full use of these signatures, however, requires a better understanding of their intra-type variability as regional differences effect the predictive power of the signatures. Intuitively, the topic composition for place types such as store and office should be less effected by regional differences than the topic composition for types such as monument and mountain. In this work, we approach this regional variability hypothesis by attempting to prove that all place types are aspatial with respect to their thematic signatures. We reject this hypothesis by comparing the signature similarities of 316 place types between major cities in the U.S. We then select the most and least varying place types and compare them to thematic signatures from regions outside of the U.S. Finally, we explore the effects of LDA topic resolution on differences between and within place types},
keywords = {},
pubstate = {published},
tppubtype = {conference}
}
Computational models of place are a key component of spatial information theory and play an increasing role in research ranging from spatial search to transportation studies. One method to arrive at such models is to extract knowledge from user-generated content e.g., from texts, tags, trajectories, pictures, and so forth. Over the last years, topic modeling techniques such as latent Dirichlet allocation (LDA) have been studied to reveal linguistic patterns that characterize places and their types. Intuitively, people are more likely to describe places such as Yosemite National Park in terms of hiking, nature, and camping than cocktail or dancing. The geo-indicativeness of non-georeferenced text does not only apply to place instances but also place types, e.g., state parks. While different parks will vary greatly with respect to their landscape and thus human descriptions, the distribution of topics common to all parks will differ significantly from other types of places, e.g., night clubs. This aggregation of topics to the type level creates thematic signatures that can be used for place categorization, data cleansing and conflation, semantic search, and so on. To make full use of these signatures, however, requires a better understanding of their intra-type variability as regional differences effect the predictive power of the signatures. Intuitively, the topic composition for place types such as store and office should be less effected by regional differences than the topic composition for types such as monument and mountain. In this work, we approach this regional variability hypothesis by attempting to prove that all place types are aspatial with respect to their thematic signatures. We reject this hypothesis by comparing the signature similarities of 316 place types between major cities in the U.S. We then select the most and least varying place types and compare them to thematic signatures from regions outside of the U.S. Finally, we explore the effects of LDA topic resolution on differences between and within place types
Gao, Song; Janowicz, Krzysztof; Montello, Daniel R; Hu, Yingjie; Yang, Jiue-An; McKenzie, Grant; Ju, Yiting; Gong, Li; Adams, Benjamin; Yan, Bo
A data-synthesis-driven method for detecting and extracting vague cognitive regions Journal Article
In: International Journal of Geographical Information Science, vol. 31, pp. 1245–1271, 2017.
@article{39,
title = {A data-synthesis-driven method for detecting and extracting vague cognitive regions},
author = {Song Gao and Krzysztof Janowicz and Daniel R Montello and Yingjie Hu and Jiue-An Yang and Grant McKenzie and Yiting Ju and Li Gong and Benjamin Adams and Bo Yan},
url = {http://grantmckenzie.com/academics/VagueRegions2017.pdf},
doi = {10.1080/13658816.2016.1273357},
year = {2017},
date = {2017-01-01},
journal = {International Journal of Geographical Information Science},
volume = {31},
pages = {1245–1271},
abstract = {Cognitive regions and places are notoriously difficult to represent in geographic information science and systems. The exact delineation of cognitive regions is challenging insofar as borders are vague, membership within the regions varies non-monotonically, and raters cannot be assumed to assess membership consistently and homogeneously. In a recent study, Montello et al. (2014) devised a novel grid-based task in which participants rated the membership of individual cells in a given region and contrasted this approach to a standard boundarydrawing task. Specifically, the authors assessed the vague cognitive regions of Northern California and Southern California. The boundary between these cognitive regions was found to have variable width, and region membership peaked not at the most northern or southern cells but at substantially less extreme latitudes. The authors thus concluded that region membership is about attitude, not just latitude. In the present work, we reproduce this study by approaching it from a computational fourth-paradigm perspective, i.e., by the synthesis of high volumes of heterogeneous data from various sources. We compare the regions which we identify to those from Montello et al. (2014), identifying differences and commonalities. Our results show a significant positive correlation to those in the original study. Beyond the extracted regions themselves, we compare and contrast the empirical and analytical approaches of these two methods, one a conventional human-participants study and the other an application of increasingly popular data-synthesis-driven research methods in GIScience},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Cognitive regions and places are notoriously difficult to represent in geographic information science and systems. The exact delineation of cognitive regions is challenging insofar as borders are vague, membership within the regions varies non-monotonically, and raters cannot be assumed to assess membership consistently and homogeneously. In a recent study, Montello et al. (2014) devised a novel grid-based task in which participants rated the membership of individual cells in a given region and contrasted this approach to a standard boundarydrawing task. Specifically, the authors assessed the vague cognitive regions of Northern California and Southern California. The boundary between these cognitive regions was found to have variable width, and region membership peaked not at the most northern or southern cells but at substantially less extreme latitudes. The authors thus concluded that region membership is about attitude, not just latitude. In the present work, we reproduce this study by approaching it from a computational fourth-paradigm perspective, i.e., by the synthesis of high volumes of heterogeneous data from various sources. We compare the regions which we identify to those from Montello et al. (2014), identifying differences and commonalities. Our results show a significant positive correlation to those in the original study. Beyond the extracted regions themselves, we compare and contrast the empirical and analytical approaches of these two methods, one a conventional human-participants study and the other an application of increasingly popular data-synthesis-driven research methods in GIScience
McKenzie, Grant; Janowicz, Krzysztof
In: PLoS ONE, 2017.
@article{42,
title = {ISED: Constructing a high-resolution elevation road dataset from massive, low-quality in-situ observations derived from geosocial fitness tracking data},
author = {Grant McKenzie and Krzysztof Janowicz},
url = {http://grantmckenzie.com/academics/McKenzie_ISED_2017.pdf},
year = {2017},
date = {2017-01-01},
journal = {PLoS ONE},
abstract = {Gaining access to inexpensive, high-resolution, up-to-date, three-dimensional road network data is a top priority beyond research, as such data would fuel applications in industry, governments, and the broader public alike. Road network data are openly available via user-generated content such as OpenStreetMap (OSM) but lack the resolution required for many tasks, e.g., emergency management. More importantly, however, few publicly available data offer information on elevation and slope. For most parts of the world, up-to-date digital elevation products with a resolution of less than 10 meters are a distant dream and, if available, those datasets have to be matched to the road network through an error-prone process. In this paper we present a radically different approach by deriving road network elevation data from massive amounts of in-situ observations extracted from user-contributed data from an online social fitness tracking application. While each individual observation may be of low-quality in terms of resolution and accuracy, taken together they form an accurate, high-resolution, up-to-date, three-dimensional road network that excels where other technologies such as LiDAR fail, e.g., in case of overpasses, overhangs, and so forth. In fact, the 1m spatial resolution dataset created in this research based on 350 million individual 3D location fixes has an RMSE of approximately 3.11m compared to a LiDAR-based ground-truth and can be used to enhance existing road network datasets where individual elevation fixes differ by up to 60m. In contrast, using interpolated data from the National Elevation Dataset (NED) results in 4.75m RMSE compared to the base line. We utilize Linked Data technologies to integrate the proposed high-resolution dataset with OpenStreetMap road geometries without requiring any changes to the OSM data model.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Gaining access to inexpensive, high-resolution, up-to-date, three-dimensional road network data is a top priority beyond research, as such data would fuel applications in industry, governments, and the broader public alike. Road network data are openly available via user-generated content such as OpenStreetMap (OSM) but lack the resolution required for many tasks, e.g., emergency management. More importantly, however, few publicly available data offer information on elevation and slope. For most parts of the world, up-to-date digital elevation products with a resolution of less than 10 meters are a distant dream and, if available, those datasets have to be matched to the road network through an error-prone process. In this paper we present a radically different approach by deriving road network elevation data from massive amounts of in-situ observations extracted from user-contributed data from an online social fitness tracking application. While each individual observation may be of low-quality in terms of resolution and accuracy, taken together they form an accurate, high-resolution, up-to-date, three-dimensional road network that excels where other technologies such as LiDAR fail, e.g., in case of overpasses, overhangs, and so forth. In fact, the 1m spatial resolution dataset created in this research based on 350 million individual 3D location fixes has an RMSE of approximately 3.11m compared to a LiDAR-based ground-truth and can be used to enhance existing road network datasets where individual elevation fixes differ by up to 60m. In contrast, using interpolated data from the National Elevation Dataset (NED) results in 4.75m RMSE compared to the base line. We utilize Linked Data technologies to integrate the proposed high-resolution dataset with OpenStreetMap road geometries without requiring any changes to the OSM data model.
Liu, Zheng; McKenzie, Grant
Identifying Spatiotemporal Activity Patterns in Beijing Based on Geosocial Microblogs Booklet
2017.
@booklet{53,
title = {Identifying Spatiotemporal Activity Patterns in Beijing Based on Geosocial Microblogs},
author = {Zheng Liu and Grant McKenzie},
url = {http://www.grantmckenzie.com/academics/CGIS_Zheng2017.pdf},
year = {2017},
date = {2017-01-01},
abstract = {Poster},
month = {01},
keywords = {},
pubstate = {published},
tppubtype = {booklet}
}
Poster
2016
Janowicz, Krzysztof; Hu, Yingjie; McKenzie, Grant; Gao, Song; Regalia, Blake; Mai, Gengchen; Zhu, Rui; Adams, Benjamin; Taylor, Kerry
Moon Landing or Safari? A Study of Systematic Errors and their Causes in Geographic Linked Data Conference
Geographic Information Science, Montreal, Canada, 2016.
@conference{16,
title = {Moon Landing or Safari? A Study of Systematic Errors and their Causes in Geographic Linked Data},
author = {Krzysztof Janowicz and Yingjie Hu and Grant McKenzie and Song Gao and Blake Regalia and Gengchen Mai and Rui Zhu and Benjamin Adams and Kerry Taylor},
url = {http://grantmckenzie.com/academics/giscience16_linkeddata_quality.pdf},
year = {2016},
date = {2016-09-01},
booktitle = {Geographic Information Science},
address = {Montreal, Canada},
abstract = {While the adoption of Linked Data technologies has grown dramatically over the past few years, it has not come without its own set of growing challenges. The triplification of domain data into Linked Data has not only given rise to a leading role of places and positioning information for the dense interlinkage of data about actors, objects, and events, but also led to massive errors in the generation, transformation, and semantic annotation of data. In a global and densely interlinked graph of data, even seemingly minor error can have far reaching consequences as different datasets make statements about the same resources. In this work we present the first comprehensive study of systematic errors and their potential causes. We also discuss lessons learned and means to avoid some of the introduced pitfalls in the future.},
keywords = {},
pubstate = {published},
tppubtype = {conference}
}
While the adoption of Linked Data technologies has grown dramatically over the past few years, it has not come without its own set of growing challenges. The triplification of domain data into Linked Data has not only given rise to a leading role of places and positioning information for the dense interlinkage of data about actors, objects, and events, but also led to massive errors in the generation, transformation, and semantic annotation of data. In a global and densely interlinked graph of data, even seemingly minor error can have far reaching consequences as different datasets make statements about the same resources. In this work we present the first comprehensive study of systematic errors and their potential causes. We also discuss lessons learned and means to avoid some of the introduced pitfalls in the future.
McKenzie, Grant; Baez, Carlos
Uber vs. Taxis: Event detection and differentiation in New York City Conference
Geographic Information Science, Montreal, Canada, 2016.
@conference{27,
title = {Uber vs. Taxis: Event detection and differentiation in New York City},
author = {Grant McKenzie and Carlos Baez},
url = {http://grantmckenzie.com/academics/McKenzie_GIScience2016.pdf},
year = {2016},
date = {2016-09-01},
booktitle = {Geographic Information Science},
address = {Montreal, Canada},
abstract = {The recent rise of ride-sourcing services such as Uber have significantly changed the transportation landscape. This work takes a first step in differentiating Uber and taxi transportation methods through events attended by their passengers. Using a sample of Uber and taxi pick-up times and locations in New York City, we show that events can be detected within each platform. Through identification of a select few of these events, this work takes a preliminary step in showing that there is a difference in the types of events that are attended by Uber users and taxi passengers.},
keywords = {},
pubstate = {published},
tppubtype = {conference}
}
The recent rise of ride-sourcing services such as Uber have significantly changed the transportation landscape. This work takes a first step in differentiating Uber and taxi transportation methods through events attended by their passengers. Using a sample of Uber and taxi pick-up times and locations in New York City, we show that events can be detected within each platform. Through identification of a select few of these events, this work takes a preliminary step in showing that there is a difference in the types of events that are attended by Uber users and taxi passengers.
McKenzie, Grant; Janowicz, Krzysztof; Seidl, Dara
Geo-privacy beyond coordinates Conference
The 19th AGILE Conference on Geographic Information Science: Geospatial Data in a Changing World, vol. Geospatial Data in a Changing World, Springer Springer, Helsinki, Finland, 2016, ISSN: 978-3-319-33782-1.
@conference{12,
title = {Geo-privacy beyond coordinates},
author = {Grant McKenzie and Krzysztof Janowicz and Dara Seidl},
url = {http://grantmckenzie.com/academics/McKenzie_AGILE2016.pdf},
doi = {10.1007/978-3-319-33783-8_10},
issn = {978-3-319-33782-1},
year = {2016},
date = {2016-06-01},
booktitle = {The 19th AGILE Conference on Geographic Information Science: Geospatial Data in a Changing World},
volume = {Geospatial Data in a Changing World},
pages = {157-175},
publisher = {Springer},
address = {Helsinki, Finland},
edition = {Lecture Notes in Geoinformation and Cartography},
organization = {Springer},
abstract = {The desire to share onetextquoterights location with friends and family or to use location information for navigation and recommendations services is often overshadowed by the need to preserve privacy. As recent progress in big data analytics, ambient intelligence, and conflation techniques is met with the economytextquoterights growing hunger for data, even formerly negligible digital footprints become revealing of our activities. The majority of established geoprivacy research tries to protect an individualtextquoterights location by different masking or perturbation techniques or by suppressing and generalizing an individualtextquoterights characteristics to a degree where she cannot be singled out from a crowd. In this work we demonstrate that location privacy may already be compromised before these techniques take effect. More concretely, we discuss how everyday digital footprints such as timestamps, geosocial check-ins, and short social media messages, e.g., tweets, are indicative of the usertextquoterights location. We focus particularly on places and highlight how protecting place-based information differs from a purely spatial perspective. The presented research is based on so-called semantic signatures that are mined from millions of geosocial checkins and enable a probabilistic framework on the level of geographic feature types, here Points Of Interest (POI). While our work is compatible with leading privacy techniques, we take a user-centric perspective and illustrate how privacy-enabled services could guide the users by increasing information entropy.},
keywords = {},
pubstate = {published},
tppubtype = {conference}
}
The desire to share onetextquoterights location with friends and family or to use location information for navigation and recommendations services is often overshadowed by the need to preserve privacy. As recent progress in big data analytics, ambient intelligence, and conflation techniques is met with the economytextquoterights growing hunger for data, even formerly negligible digital footprints become revealing of our activities. The majority of established geoprivacy research tries to protect an individualtextquoterights location by different masking or perturbation techniques or by suppressing and generalizing an individualtextquoterights characteristics to a degree where she cannot be singled out from a crowd. In this work we demonstrate that location privacy may already be compromised before these techniques take effect. More concretely, we discuss how everyday digital footprints such as timestamps, geosocial check-ins, and short social media messages, e.g., tweets, are indicative of the usertextquoterights location. We focus particularly on places and highlight how protecting place-based information differs from a purely spatial perspective. The presented research is based on so-called semantic signatures that are mined from millions of geosocial checkins and enable a probabilistic framework on the level of geographic feature types, here Points Of Interest (POI). While our work is compatible with leading privacy techniques, we take a user-centric perspective and illustrate how privacy-enabled services could guide the users by increasing information entropy.
Regalia, Blake; McKenzie, Grant; Gao, Song; Janowicz, Krzysztof
Crowdsensing Smart Ambient Environments and Services Journal Article
In: Transactions in GIS, vol. 20, pp. 382–398, 2016.
@article{14,
title = {Crowdsensing Smart Ambient Environments and Services},
author = {Blake Regalia and Grant McKenzie and Song Gao and Krzysztof Janowicz},
url = {http://grantmckenzie.com/academics/McKenzie_TGIS2016_2.pdf},
doi = {10.1111/tgis.12233},
year = {2016},
date = {2016-06-01},
journal = {Transactions in GIS},
volume = {20},
pages = {382–398},
abstract = {Whether it be Smart Cities, Ambient Intelligence, or the Internet of Things, current visions for future urban spaces share a common core, namely the increasing role of distributed sensor networks and the on-demand integration of their data to power real-time services and analytics. Some of the greatest hurdles to implementing these visions include security risks, user privacy, scalability, the integration of heterogeneous data, and financial cost. In this work, we propose a crowdsensing mobile-device platform that empowers citizens to collect and share information about their surrounding environment via embedded sensor technologies. This approach allows a variety of urban areas (e.g., university campuses, shopping malls, city centers, suburbs) to become equipped with a free ad-hoc sensor network without depending on proprietary instrumentation. We present a framework, namely the GeoTracer application, as a proof of concept to conduct multiple experiments simulating use-case scenarios on a university campus. First, we demonstrate that ambient sensors (e.g., temperature, pressure, humidity, magnetism, illuminance, and audio) can help determine a change in environment (e.g., moving from indoors to outdoors, or floor changes inside buildings) more accurately than typical positioning technologies (e.g., global navigation satellite system, Wi-Fi, etc.). Furthermore, each of these sensors contributes a different amount of data to detecting events. for example, illuminance has the highest information gain when trying to detect changes between indoors and outdoors. Second, we show that through this platform it is possible to detect and differentiate place types on a university campus based on inferences made through ambient sensors. Lastly, we train classifiers to determine the activities that a place can afford at different times (e.g., good for studying or not, basketball courts in use or empty) based on sensor-driven semantic signatures},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Whether it be Smart Cities, Ambient Intelligence, or the Internet of Things, current visions for future urban spaces share a common core, namely the increasing role of distributed sensor networks and the on-demand integration of their data to power real-time services and analytics. Some of the greatest hurdles to implementing these visions include security risks, user privacy, scalability, the integration of heterogeneous data, and financial cost. In this work, we propose a crowdsensing mobile-device platform that empowers citizens to collect and share information about their surrounding environment via embedded sensor technologies. This approach allows a variety of urban areas (e.g., university campuses, shopping malls, city centers, suburbs) to become equipped with a free ad-hoc sensor network without depending on proprietary instrumentation. We present a framework, namely the GeoTracer application, as a proof of concept to conduct multiple experiments simulating use-case scenarios on a university campus. First, we demonstrate that ambient sensors (e.g., temperature, pressure, humidity, magnetism, illuminance, and audio) can help determine a change in environment (e.g., moving from indoors to outdoors, or floor changes inside buildings) more accurately than typical positioning technologies (e.g., global navigation satellite system, Wi-Fi, etc.). Furthermore, each of these sensors contributes a different amount of data to detecting events. for example, illuminance has the highest information gain when trying to detect changes between indoors and outdoors. Second, we show that through this platform it is possible to detect and differentiate place types on a university campus based on inferences made through ambient sensors. Lastly, we train classifiers to determine the activities that a place can afford at different times (e.g., good for studying or not, basketball courts in use or empty) based on sensor-driven semantic signatures
Zhu, Rui; Hu, Yingjie; Janowicz, Krzysztof; McKenzie, Grant
Spatial signatures for geographic feature types: examining gazetteer ontologies using spatial statistics Journal Article
In: Transactions in GIS, vol. 20, pp. 333–355, 2016.
@article{15,
title = {Spatial signatures for geographic feature types: examining gazetteer ontologies using spatial statistics},
author = {Rui Zhu and Yingjie Hu and Krzysztof Janowicz and Grant McKenzie},
url = {http://grantmckenzie.com/academics/McKenzie_TGIS2016_1.pdf},
doi = {10.1111/tgis.12232},
year = {2016},
date = {2016-06-01},
journal = {Transactions in GIS},
volume = {20},
pages = {333–355},
abstract = {Digital gazetteers play a key role in modern information systems and infrastructures. They facilitate (spatial) search, deliver contextual information to recommended systems, enrich textual information with geographical references, and provide stable identifiers to interlink actors, events, and objects by the places they interact with. Hence, it is unsurprising that gazetteers, such as GeoNames, are among the most densely interlinked hubs on the Web of Linked Data. A wide variety of digital gazetteers have been developed over the years to serve different communities and needs. These gazetteers differ in their overall coverage, underlying data sources, provided functionality, and geographic feature type ontologies. Consequently, place types that share a common name may differ substantially between gazetteers, whereas types labeled differently may, in fact, specify the same or similar places. This makes data integration and federated queries challenging, if not impossible. To further complicate the situation, most popular and widely adopted geo-ontologies are lightweight and thus under-specific to a degree where their alignment and matching become nothing more than educated guesses. The most promising approach to addressing this problem, and thereby enabling the meaningful integration of gazetteer data across feature types, seems to be a combination of top-down knowledge representation with bottom-up data-driven techniques such as feature engineering and machine learning. In this work, we propose to derive indicative spatial signatures for geographic feature types by using spatial statistics. We discuss how to create such signatures by feature engineering and demonstrate how the signatures can be applied to better understand the differences and commonalities of three major gazetteers, namely DBpedia Places, GeoNames, and TGN.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Digital gazetteers play a key role in modern information systems and infrastructures. They facilitate (spatial) search, deliver contextual information to recommended systems, enrich textual information with geographical references, and provide stable identifiers to interlink actors, events, and objects by the places they interact with. Hence, it is unsurprising that gazetteers, such as GeoNames, are among the most densely interlinked hubs on the Web of Linked Data. A wide variety of digital gazetteers have been developed over the years to serve different communities and needs. These gazetteers differ in their overall coverage, underlying data sources, provided functionality, and geographic feature type ontologies. Consequently, place types that share a common name may differ substantially between gazetteers, whereas types labeled differently may, in fact, specify the same or similar places. This makes data integration and federated queries challenging, if not impossible. To further complicate the situation, most popular and widely adopted geo-ontologies are lightweight and thus under-specific to a degree where their alignment and matching become nothing more than educated guesses. The most promising approach to addressing this problem, and thereby enabling the meaningful integration of gazetteer data across feature types, seems to be a combination of top-down knowledge representation with bottom-up data-driven techniques such as feature engineering and machine learning. In this work, we propose to derive indicative spatial signatures for geographic feature types by using spatial statistics. We discuss how to create such signatures by feature engineering and demonstrate how the signatures can be applied to better understand the differences and commonalities of three major gazetteers, namely DBpedia Places, GeoNames, and TGN.
Mai, Gengchen; Janowicz, Krzysztof; Hu, Yingjie; McKenzie, Grant
A Linked Data Driven Visual Interface for the Multi-Perspective Exploration of Data Across Repositories Conference
VOILA 2016, Kobe, Japan, 2016.
@conference{26,
title = {A Linked Data Driven Visual Interface for the Multi-Perspective Exploration of Data Across Repositories},
author = {Gengchen Mai and Krzysztof Janowicz and Yingjie Hu and Grant McKenzie},
year = {2016},
date = {2016-01-01},
booktitle = {VOILA 2016},
address = {Kobe, Japan},
abstract = {As more data from heterogeneous sources become available, interfaces that support the federated exploration of these data are gaining importance to uncover relations between entities across multiple sources. Instead of explicit queries, visual interfaces enable a follow-your-nose style of exploration by which a user can seamlessly navigate between entities from different data sources. This requires an alignment of the ontologies used by said sources as well as the coreference resolution of entities across them. Together with Semantic Web technologies, the Linked Data paradigm provides the technological foundations to address these challenges. Nonetheless, the majority of work studies these components in isolation, focusing either on the alignment, coreference resolution, or visualization. Some interesting aspects, however, only arise when all puzzle pieces are in place. Two of these aspects are the seamless transitions be- tween visualization and interaction paradigms as well as the combination of entity and type queries. In this work, we present a multi-perspective visual interface that enables the seamless exploration of major scientific geo-data sources that contain millions of RDF triples.},
keywords = {},
pubstate = {published},
tppubtype = {conference}
}
As more data from heterogeneous sources become available, interfaces that support the federated exploration of these data are gaining importance to uncover relations between entities across multiple sources. Instead of explicit queries, visual interfaces enable a follow-your-nose style of exploration by which a user can seamlessly navigate between entities from different data sources. This requires an alignment of the ontologies used by said sources as well as the coreference resolution of entities across them. Together with Semantic Web technologies, the Linked Data paradigm provides the technological foundations to address these challenges. Nonetheless, the majority of work studies these components in isolation, focusing either on the alignment, coreference resolution, or visualization. Some interesting aspects, however, only arise when all puzzle pieces are in place. Two of these aspects are the seamless transitions be- tween visualization and interaction paradigms as well as the combination of entity and type queries. In this work, we present a multi-perspective visual interface that enables the seamless exploration of major scientific geo-data sources that contain millions of RDF triples.
McKenzie, Grant; Hegarty, Mary; Barrett, Trevor; Goodchild, Michael
Assessing the effectiveness of different visualizations for judgments of positional uncertainty Journal Article
In: International Journal of Geographic Information Science, vol. 30, pp. 221-239, 2016.
@article{5h,
title = {Assessing the effectiveness of different visualizations for judgments of positional uncertainty},
author = {Grant McKenzie and Mary Hegarty and Trevor Barrett and Michael Goodchild},
url = {http://grantmckenzie.com/academics/McKenzie_PositionalUncertainty2015_preprint.pdf},
doi = {10.1080/13658816.2015.1082566},
year = {2016},
date = {2016-01-01},
journal = {International Journal of Geographic Information Science},
volume = {30},
pages = {221-239},
abstract = {Many techniques have been proposed for visualizing uncertainty in geospatial data. Previous empirical research on the effectiveness of visualizations of geospatial uncertainty has focused primarily on user intuitions rather than objective measures of performance when reasoning under uncertainty. Framed in the context of Googletextquoterights blue dot, we examined the effectiveness of four alternative visualizations for representing positional uncertainty when reasoning about self-location data. Our task presents a mobile mapping scenario in which GPS satellite location readings produce location estimates with varying levels of uncertainty. Given a known location and two smartphone estimates of that known location, participants were asked to judge which smartphone produces the better location reading, taking uncertainty into account. We produced visualizations that vary by glyph type (uniform blue circle with border vs. Gaussian fade) and visibility of a centroid dot (visible vs. not visible) to produce the four visualization formats. Participants viewing the uniform blue circle are most likely to respond in accordance with the actual probability density of points sampled from bivariate normal distributions and additionally respond most rapidly. Participants reported a number of simple heuristics on which they based their judgments, and consistency with these heuristics was highly predictive of their judgments.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Many techniques have been proposed for visualizing uncertainty in geospatial data. Previous empirical research on the effectiveness of visualizations of geospatial uncertainty has focused primarily on user intuitions rather than objective measures of performance when reasoning under uncertainty. Framed in the context of Googletextquoterights blue dot, we examined the effectiveness of four alternative visualizations for representing positional uncertainty when reasoning about self-location data. Our task presents a mobile mapping scenario in which GPS satellite location readings produce location estimates with varying levels of uncertainty. Given a known location and two smartphone estimates of that known location, participants were asked to judge which smartphone produces the better location reading, taking uncertainty into account. We produced visualizations that vary by glyph type (uniform blue circle with border vs. Gaussian fade) and visibility of a centroid dot (visible vs. not visible) to produce the four visualization formats. Participants viewing the uniform blue circle are most likely to respond in accordance with the actual probability density of points sampled from bivariate normal distributions and additionally respond most rapidly. Participants reported a number of simple heuristics on which they based their judgments, and consistency with these heuristics was highly predictive of their judgments.
2015
Adams, Benjamin; McKenzie, Grant; Gahegan, Mark
Frankenplace: interactive thematic mapping for ad hoc exploratory search Conference
The 24th International Conference on World Wide Web, ACM ACM, Florence, Italy, 2015.
@conference{17,
title = {Frankenplace: interactive thematic mapping for ad hoc exploratory search},
author = {Benjamin Adams and Grant McKenzie and Mark Gahegan},
url = {http://www.www2015.it/documents/proceedings/proceedings/p12.pdf},
doi = {10.1145/2736277.2741137},
year = {2015},
date = {2015-05-01},
booktitle = {The 24th International Conference on World Wide Web},
publisher = {ACM},
address = {Florence, Italy},
organization = {ACM},
abstract = {Ad hoc keyword search engines built using modern information retrieval methods do a good job of handling fine-grained queries. However, they perform poorly at facilitating spatial and spatially-embedded thematic exploration of the results, despite the fact that many queries, e.g. civil war, refer to different documents and topics in different places. This is not for lack of data: geographic information, such as place names, events, and coordinates are common in unstructured document collections on the web. The associations between geographic and thematic contents in these documents can provide a rich groundwork to organize information for exploratory research. In this paper we describe the architecture of an interactive thematic map search engine, Frankenplace, designed to facilitate document exploration at the intersection of theme and place. The map interface enables a user to zoom the geographic context of their query in and out, and quickly explore through thousands of search results in a meaningful way. And by combining topic models with geographically contextualized search results, users can discover related topics based on geographic context. Frankenplace utilizes a novel indexing method called geoboost for boosting terms associated with cells on a discrete global grid. The resulting index factors in the geographic scale of the place or feature mentioned in related text, the relative textual scope of the place reference, and the overall importance of the containing document in the document network. The system is currently indexed with over 5 million documents from the web, including the English Wikipedia and online travel blog entries. We demonstrate that Frankenplace can support four distinct types of exploratory search tasks while being adaptive to scale and location of interest.},
keywords = {},
pubstate = {published},
tppubtype = {conference}
}
Ad hoc keyword search engines built using modern information retrieval methods do a good job of handling fine-grained queries. However, they perform poorly at facilitating spatial and spatially-embedded thematic exploration of the results, despite the fact that many queries, e.g. civil war, refer to different documents and topics in different places. This is not for lack of data: geographic information, such as place names, events, and coordinates are common in unstructured document collections on the web. The associations between geographic and thematic contents in these documents can provide a rich groundwork to organize information for exploratory research. In this paper we describe the architecture of an interactive thematic map search engine, Frankenplace, designed to facilitate document exploration at the intersection of theme and place. The map interface enables a user to zoom the geographic context of their query in and out, and quickly explore through thousands of search results in a meaningful way. And by combining topic models with geographically contextualized search results, users can discover related topics based on geographic context. Frankenplace utilizes a novel indexing method called geoboost for boosting terms associated with cells on a discrete global grid. The resulting index factors in the geographic scale of the place or feature mentioned in related text, the relative textual scope of the place reference, and the overall importance of the containing document in the document network. The system is currently indexed with over 5 million documents from the web, including the English Wikipedia and online travel blog entries. We demonstrate that Frankenplace can support four distinct types of exploratory search tasks while being adaptive to scale and location of interest.