Open    Access



Collective Background Extraction for Station Market Area by Using Location Based Social Network



Author(s)

Kousuke Kikuchi1, Tatsuto Kihara2, Atsushi Enta3, Hideaki Takayanagi2, Takeshi Kimura4, Kazuto Hayashida1 and Hitoshi Watanabe1

Full-Text PDF XML     223 Views

DOI:10.17265/1934-7359/2013.03.004

1. Department of Creative Science and Engineering, Waseda University, Tokyo 169-8555, Japan
2. School of Environmental Science, University of Shiga Prefecture, Shiga 522-8533, Japan
3. Faculty of Science and Technology, Tokyo University of Science, Tokyo 121-0062, Japan
4. A&A Co. Ltd., Tokyo 101-0062, Japan

Half centuries of follow-up survey has enabled the architects and urban planners to design rationally by the aid of planning. Nonetheless, limitation has occurred at planning because city has been changing its utility in accordance with its users’ demand. In this paper, the authors proposed a method to analyze trait of users in market areas near stations by analyzing location based social network. After the datum collection from geotagged tweets, these GPS (global positioning system) datum were plotted to map attained from yahoo open location platform. Then the morphological analysis and terminology extraction system extracted the keywords and their scores. After calculating the distance from stations and users’ GPS coordination, the authors extracted the array of keywords and corresponding scores in some station market area. Lastly, ratios of all users’ scores and city’s scores were calculated to examine the locality. Full combination of data collection, natural language processing and visualization enabled the authors to envisage distribution of collective background in city.

Location based social networks, natural language processing, market analysis, visualization, big data.