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.
McKenzie, Grant; Janowicz, Krzysztof; Gao, Song; Gong, Li
How where is when? On the regional variability and resolution of geosocial temporal signatures for points of interest Journal Article
In: Computers, Environment and Urban Systems, vol. 54, pp. 336–346, 2015.
@article{7f,
title = {How where is when? On the regional variability and resolution of geosocial temporal signatures for points of interest},
author = {Grant McKenzie and Krzysztof Janowicz and Song Gao and Li Gong},
url = {http://grantmckenzie.com/academics/McKenzie_CEUS2015_2.pdf},
doi = {10.1016/j.compenvurbsys.2015.10.002},
year = {2015},
date = {2015-01-01},
journal = {Computers, Environment and Urban Systems},
volume = {54},
pages = {336–346},
abstract = {The temporal characteristics of human behavior with respect to Points of Interest (POI) differ significantly with place type. Intuitively, we are more likely to visit a restaurant during typical lunch and dinner times than at midnight. Aggregating geosocial check-ins of millions of users to the place type level leads to powerful temporal bands and signatures. In previous work these signatures have been used to estimate the place being visited based purely on the check-in time, to label uncategorized places based on their individual signaturetextquoterights similarity to a type signature, and to mine POI categories and their hierarchical structure from the bottom-up. However, not all hours of the day and days of the week are equally indicative of the place type, i.e., the information gain between temporal bands that jointly form a place type signature differs. To give a concrete example, places can be more easily categorized into weekend and weekday places than into Monday and Tuesday places. Nonetheless, research on the regional variability of temporal signatures is lacking. Intuitively, one would assume that certain types of places are more prone to regional differences with respect to the temporal check-in behavior than others. This variability will impact the predictive power of the signatures and reduce the number of POI types that can be distinguished. In this work, we address the regional variability hypothesis by trying to prove that all place types are created equal with respect to their temporal signatures, i.e., temporal check-in behavior does not change across space. We reject this hypothesis by comparing the inter-signature similarity of 321 place types in three major cities in the USA (Los Angeles, New York, and Chicago). Next, we identify a common core of least varying place types and compare it against signatures extracted from the city of Shanghai, China for cross-culture comparison. Finally, we discuss the impact of our findings on POI categorization and the reliability of temporal signatures for check-in behavior in general.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
The temporal characteristics of human behavior with respect to Points of Interest (POI) differ significantly with place type. Intuitively, we are more likely to visit a restaurant during typical lunch and dinner times than at midnight. Aggregating geosocial check-ins of millions of users to the place type level leads to powerful temporal bands and signatures. In previous work these signatures have been used to estimate the place being visited based purely on the check-in time, to label uncategorized places based on their individual signaturetextquoterights similarity to a type signature, and to mine POI categories and their hierarchical structure from the bottom-up. However, not all hours of the day and days of the week are equally indicative of the place type, i.e., the information gain between temporal bands that jointly form a place type signature differs. To give a concrete example, places can be more easily categorized into weekend and weekday places than into Monday and Tuesday places. Nonetheless, research on the regional variability of temporal signatures is lacking. Intuitively, one would assume that certain types of places are more prone to regional differences with respect to the temporal check-in behavior than others. This variability will impact the predictive power of the signatures and reduce the number of POI types that can be distinguished. In this work, we address the regional variability hypothesis by trying to prove that all place types are created equal with respect to their temporal signatures, i.e., temporal check-in behavior does not change across space. We reject this hypothesis by comparing the inter-signature similarity of 321 place types in three major cities in the USA (Los Angeles, New York, and Chicago). Next, we identify a common core of least varying place types and compare it against signatures extracted from the city of Shanghai, China for cross-culture comparison. Finally, we discuss the impact of our findings on POI categorization and the reliability of temporal signatures for check-in behavior in general.
McKenzie, Grant; Janowicz, Krzysztof; Gao, Song; Yang, Jiue-An; Hu, Yingjie
POI pulse: A multi-granular, semantic signatures-based information observatory for the interactive visualization of big geosocial data Journal Article
In: Cartographica, vol. 50, pp. 71–85, 2015.
@article{9f,
title = {POI pulse: A multi-granular, semantic signatures-based information observatory for the interactive visualization of big geosocial data},
author = {Grant McKenzie and Krzysztof Janowicz and Song Gao and Jiue-An Yang and Yingjie Hu},
url = {http://grantmckenzie.com/academics/McKenzie_POIPulse.pdf},
doi = {10.3138/carto.50.2.2662},
year = {2015},
date = {2015-01-01},
journal = {Cartographica},
volume = {50},
pages = {71–85},
abstract = {The volume, velocity, and variety at which data are now becoming available allow us to study urban environments based on human behavior at a spatial, temporal, and thematic granularity that was not achievable until now. Such data-driven approaches opens up additional, complementary perspectives on how urban systems function, especially if they are based on User-Generated Content (UGC). While the data sources, e.g., social media, introduce specific biases, they also open up new possibilities for scientists and the broader public. For instance, they provide answers to questions that previously could only be addressed by complex simulations or extensive human participant surveys. Unfortunately, many of the required datasets are locked in data silos that are only accessible via restricted APIs. Even if these data could be fully accessed, their natextasciidieresisive processing and visualization would surpass the abilities of modern computer architectures. Finally, the established place schemata used to study urban spaces differ substantially from UGC-based Point of Interest (POI) schemata. In this work, we present a multi-granular, datadriven, and theory-informed approach that addressed the key issues outlined above by introducing the theoretical and technical framework to interactively explore the pulse of a city based on social media.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
The volume, velocity, and variety at which data are now becoming available allow us to study urban environments based on human behavior at a spatial, temporal, and thematic granularity that was not achievable until now. Such data-driven approaches opens up additional, complementary perspectives on how urban systems function, especially if they are based on User-Generated Content (UGC). While the data sources, e.g., social media, introduce specific biases, they also open up new possibilities for scientists and the broader public. For instance, they provide answers to questions that previously could only be addressed by complex simulations or extensive human participant surveys. Unfortunately, many of the required datasets are locked in data silos that are only accessible via restricted APIs. Even if these data could be fully accessed, their natextasciidieresisive processing and visualization would surpass the abilities of modern computer architectures. Finally, the established place schemata used to study urban spaces differ substantially from UGC-based Point of Interest (POI) schemata. In this work, we present a multi-granular, datadriven, and theory-informed approach that addressed the key issues outlined above by introducing the theoretical and technical framework to interactively explore the pulse of a city based on social media.
McKenzie, Grant; Janowicz, Krzysztof
Where is also about time: A location-distortion model to improve reverse geocoding using behavior-driven temporal semantic signatures Journal Article
In: Computers, Environment and Urban Systems, vol. 54, pp. 1–13, 2015.
@article{8c,
title = {Where is also about time: A location-distortion model to improve reverse geocoding using behavior-driven temporal semantic signatures},
author = {Grant McKenzie and Krzysztof Janowicz},
url = {http://grantmckenzie.com/academics/McKenzie_WhereIsAlsoAboutTime2015.pdf},
doi = {10.1016/j.compenvurbsys.2015.05.003},
year = {2015},
date = {2015-01-01},
journal = {Computers, Environment and Urban Systems},
volume = {54},
pages = {1–13},
abstract = {While geocoding returns coordinates for a full or partial address, the converse process of reverse geocoding maps coordinates to a set of candidate place identifiers such as addresses or toponyms. For example, numerous Web APIs map geographic point coordinates, e.g., from a usertextquoterights smartphone, to an ordered set of nearby Places Of Interest (POI). Typically, these services return the k nearest POI within a certain radius and measure distance to order the results. Reverse geocoding is a crucial task for many applications and research questions as it translates between spatial and platial views on geographic location. What makes this process difficult is the uncertainty of the queried location and of the point features used to represent places. Even if both could be determined with a high level of accuracy, it would still be unclear how to map a smartphonetextquoterights GPS fix to one of many possible places in a multi-story building or a shopping mall. In this work, we break up the dependency on space alone by introducing time as a second variable for reverse geocoding. We mine the geosocial behavior of users of online location-based social networks to extract temporal semantic signatures. In analogy to the notion of scale distortion in cartography, we present a model that uses these signatures to distort the location of POI relative to the query location and time, thereby reordering the set of potentially matching places. We demonstrate the strengths of our method by evaluating it against a purely spatial baseline by determining the Mean Reciprocal Rank and the normalized Discounted Cumulative Gain. Our method performs substantially better than said baseline.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
While geocoding returns coordinates for a full or partial address, the converse process of reverse geocoding maps coordinates to a set of candidate place identifiers such as addresses or toponyms. For example, numerous Web APIs map geographic point coordinates, e.g., from a usertextquoterights smartphone, to an ordered set of nearby Places Of Interest (POI). Typically, these services return the k nearest POI within a certain radius and measure distance to order the results. Reverse geocoding is a crucial task for many applications and research questions as it translates between spatial and platial views on geographic location. What makes this process difficult is the uncertainty of the queried location and of the point features used to represent places. Even if both could be determined with a high level of accuracy, it would still be unclear how to map a smartphonetextquoterights GPS fix to one of many possible places in a multi-story building or a shopping mall. In this work, we break up the dependency on space alone by introducing time as a second variable for reverse geocoding. We mine the geosocial behavior of users of online location-based social networks to extract temporal semantic signatures. In analogy to the notion of scale distortion in cartography, we present a model that uses these signatures to distort the location of POI relative to the query location and time, thereby reordering the set of potentially matching places. We demonstrate the strengths of our method by evaluating it against a purely spatial baseline by determining the Mean Reciprocal Rank and the normalized Discounted Cumulative Gain. Our method performs substantially better than said baseline.
2014
McKenzie, Grant; Klippel, Alexander
The Interaction of Landmarks and Map Alignment in You-Are-Here Maps Journal Article
In: The Cartographic Journal, 2014.
@article{11,
title = {The Interaction of Landmarks and Map Alignment in You-Are-Here Maps},
author = {Grant McKenzie and Alexander Klippel},
url = {http://grantmckenzie.com/academics/McKenzie_2015_YAH_preprint.pdf},
doi = {10.1080/00087041.2015.1108675},
year = {2014},
date = {2014-01-01},
journal = {The Cartographic Journal},
abstract = {Knowing where one is located within an environment is one of the most fundamental tasks humans have to master in their daily routines. Maps, as external representations of the environment offer intuitive ways to extend the capacities of the human cognitive systems. Operations such as planning a route can be performed on maps instead of in the environment. Question of how to design maps that support cognitive processes such as wayfinding in novel environments have been discussed in several disciplines. The research reported here addresses the question of how map alignment and the presence of landmarks in maps interact during wayfinding. For the purpose of systematically analyzing the relationship between map alignment and landmark presence, 9 virtual environments were designed. Routes learned from maps with different alignments and different numbers of landmarks present at decision points were used. While generally landmarks are assumed to foster wayfinding performance, our results indicate that misaligned maps can cancel out positive effects obtained through landmarks.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Knowing where one is located within an environment is one of the most fundamental tasks humans have to master in their daily routines. Maps, as external representations of the environment offer intuitive ways to extend the capacities of the human cognitive systems. Operations such as planning a route can be performed on maps instead of in the environment. Question of how to design maps that support cognitive processes such as wayfinding in novel environments have been discussed in several disciplines. The research reported here addresses the question of how map alignment and the presence of landmarks in maps interact during wayfinding. For the purpose of systematically analyzing the relationship between map alignment and landmark presence, 9 virtual environments were designed. Routes learned from maps with different alignments and different numbers of landmarks present at decision points were used. While generally landmarks are assumed to foster wayfinding performance, our results indicate that misaligned maps can cancel out positive effects obtained through landmarks.
McKenzie, Grant; Janowicz, Krzysztof; Adams, Benjamin
A weighted multi-attribute method for matching user-generated Points of Interest Journal Article
In: Cartography and Geographic Information Science, vol. 41, pp. 125–137, 2014.
@article{10,
title = {A weighted multi-attribute method for matching user-generated Points of Interest},
author = {Grant McKenzie and Krzysztof Janowicz and Benjamin Adams},
url = {http://grantmckenzie.com/academics/McKenzie_CaGIS2014.pdf},
doi = {10.1080/15230406.2014.880327},
year = {2014},
date = {2014-01-01},
journal = {Cartography and Geographic Information Science},
volume = {41},
pages = {125–137},
abstract = {To a large degree, the attraction of Big Data lies in the variety of its heterogeneous multi-thematic and multidimensional data sources and not merely its volume. To fully exploit this variety, however, requires conflation. This is a two step process. First, one has to establish identity relations between information entities across the different data sources; and second, attribute values have to be merged according to certain procedures which avoid logical contradictions. The first step, also called matching, can be thought of as a weighted combination of common attributes according to some similarity measures. In this work, we propose such a matching based on multiple attributes of Points of Interests (POI) from the Location-based Social Network Foursquare and the Yelp local directory service. While both contain overlapping attributes that can be use for matching, they have specific strengths and weaknesses which makes their conflation desirable. For instance, Foursquare offers information about user check-ins to places, while Yelp specializes in user-contributed reviews. We present a weighted multi-attribute matching strategy, evaluate its performance, and discuss application areas that benefit from a successful matching. Finally, we also outline how the established POI matches can be stored as Linked Data on the Semantic Web. Our strategy can automatically match 97% of randomly selected Yelp POI to their corresponding Foursquare entities},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
To a large degree, the attraction of Big Data lies in the variety of its heterogeneous multi-thematic and multidimensional data sources and not merely its volume. To fully exploit this variety, however, requires conflation. This is a two step process. First, one has to establish identity relations between information entities across the different data sources; and second, attribute values have to be merged according to certain procedures which avoid logical contradictions. The first step, also called matching, can be thought of as a weighted combination of common attributes according to some similarity measures. In this work, we propose such a matching based on multiple attributes of Points of Interests (POI) from the Location-based Social Network Foursquare and the Yelp local directory service. While both contain overlapping attributes that can be use for matching, they have specific strengths and weaknesses which makes their conflation desirable. For instance, Foursquare offers information about user check-ins to places, while Yelp specializes in user-contributed reviews. We present a weighted multi-attribute matching strategy, evaluate its performance, and discuss application areas that benefit from a successful matching. Finally, we also outline how the established POI matches can be stored as Linked Data on the Semantic Web. Our strategy can automatically match 97% of randomly selected Yelp POI to their corresponding Foursquare entities
2013
McKenzie, Grant; Janowicz, Krzysztof; Adams, Benjamin
Weighted multi-attribute matching of user-generated points of interest Conference
Proceedings of the 21st ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems, ACM ACM, Orlando, Florida, USA, 2013.
@conference{13,
title = {Weighted multi-attribute matching of user-generated points of interest},
author = {Grant McKenzie and Krzysztof Janowicz and Benjamin Adams},
url = {http://grantmckenzie.com/academics/McKenzie_Short_SIGSPATIAL2013.pdf},
year = {2013},
date = {2013-09-01},
booktitle = {Proceedings of the 21st ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems},
publisher = {ACM},
address = {Orlando, Florida, USA},
organization = {ACM},
abstract = {To a large degree, the attraction of Big Data lies in the variety of its heterogeneous multi-thematic and multidimensional data sources and not merely its volume. To fully exploit this variety, however, requires conflation. This is a two step process. First, one has to establish identity relations between information entities across the different data sources; and second, attribute values have to be merged according to certain procedures which avoid logical contradictions. The first step, also called matching, can be thought of as a weighted combination of common attributes according to some similarity measures. In this work, we propose such a matching based on multiple attributes of Points of Interests (POI) from the Location-based Social Network Foursquare and the Yelp local directory service. While both contain overlapping attributes that can be use for matching, they have specific strengths and weaknesses which makes their conflation desirable. We present a weighted multi-attribute matching strategy and evaluate its performance. Our strategy can automatically match 97% of randomly selected Yelp POI to their corresponding Foursquare entities.},
keywords = {},
pubstate = {published},
tppubtype = {conference}
}
To a large degree, the attraction of Big Data lies in the variety of its heterogeneous multi-thematic and multidimensional data sources and not merely its volume. To fully exploit this variety, however, requires conflation. This is a two step process. First, one has to establish identity relations between information entities across the different data sources; and second, attribute values have to be merged according to certain procedures which avoid logical contradictions. The first step, also called matching, can be thought of as a weighted combination of common attributes according to some similarity measures. In this work, we propose such a matching based on multiple attributes of Points of Interests (POI) from the Location-based Social Network Foursquare and the Yelp local directory service. While both contain overlapping attributes that can be use for matching, they have specific strengths and weaknesses which makes their conflation desirable. We present a weighted multi-attribute matching strategy and evaluate its performance. Our strategy can automatically match 97% of randomly selected Yelp POI to their corresponding Foursquare entities.
McKenzie, Grant; Adams, Benjamin; Janowicz, Krzysztof
A thematic approach to user similarity built on geosocial check-ins Conference
The 16th AGILE Conference on Geographic Information Science, Springer Springer, Leuven, Belgium, 2013.
@conference{14b,
title = {A thematic approach to user similarity built on geosocial check-ins},
author = {Grant McKenzie and Benjamin Adams and Krzysztof Janowicz},
url = {http://grantmckenzie.com/academics/McKenzie_AGILE2013.pdf},
year = {2013},
date = {2013-01-01},
booktitle = {The 16th AGILE Conference on Geographic Information Science},
pages = {39–53},
publisher = {Springer},
address = {Leuven, Belgium},
organization = {Springer},
abstract = {Computing user similarity is key for personalized locationbased recommender systems and geographic information retrieval. So far, most existing work has focused on structured or semi-structured data to establish such measures. In this work, we propose topic modeling to exploit sparse, unstructured data, e.g., tips and reviews, as an additional feature to compute user similarity. Our model employs diagnosticity weighting based on the entropy of topics in order to assess the role of commonalities and variabilities between similar users. Finally, we offer a validation technique and results using data from the locationbased social network Foursquare.},
keywords = {},
pubstate = {published},
tppubtype = {conference}
}
Computing user similarity is key for personalized locationbased recommender systems and geographic information retrieval. So far, most existing work has focused on structured or semi-structured data to establish such measures. In this work, we propose topic modeling to exploit sparse, unstructured data, e.g., tips and reviews, as an additional feature to compute user similarity. Our model employs diagnosticity weighting based on the entropy of topics in order to assess the role of commonalities and variabilities between similar users. Finally, we offer a validation technique and results using data from the locationbased social network Foursquare.