Utilizing Geospatial Information in Cellular Data Usage for Key Location Prediction

dc.contributor.author Yu, Yinan
dc.contributor.author Ma, Baojun
dc.contributor.author Chen, Hailiang
dc.contributor.author Yen, Benjamin
dc.date.accessioned 2017-12-28T00:43:14Z
dc.date.available 2017-12-28T00:43:14Z
dc.date.issued 2018-01-03
dc.description.abstract Previous research on the identification of key locations (e.g., home and workplace) for a user largely relies on call detail records (CDRs). Recently, cellular data usage (i.e., mobile internet) is growing rapidly and offers fine-grained insights into various human behavior patterns. In this study, we introduce a novel dataset containing both voice and mobile data usage records of mobile users. We then construct a new feature based on the geospatial distribution of cell towers connected by mobile users and employ bivariate kernel density estimation to help predict users’ key locations. The evaluation results suggest that augmented features based on both voice and mobile data usage improve the prediction precision and recall.
dc.format.extent 8 pages
dc.identifier.doi 10.24251/HICSS.2018.123
dc.identifier.isbn 978-0-9981331-1-9
dc.identifier.uri http://hdl.handle.net/10125/50010
dc.language.iso eng
dc.relation.ispartof Proceedings of the 51st Hawaii International Conference on System Sciences
dc.rights Attribution-NonCommercial-NoDerivatives 4.0 International
dc.rights.uri https://creativecommons.org/licenses/by-nc-nd/4.0/
dc.subject Decision Support for Smart Cities
dc.subject cellular data usage, geospatial information, home and workplace, precision marketing, kernel density estimation
dc.title Utilizing Geospatial Information in Cellular Data Usage for Key Location Prediction
dc.type Conference Paper
dc.type.dcmi Text
Original bundle
Now showing 1 - 1 of 1
No Thumbnail Available
973.61 KB
Adobe Portable Document Format