Train Route Detection with using Sensor Data and GIS

• Item type: SIG Technical Reports(1)
• 公開日: 2016-08-17
• タイトル: Train Route Detection with using Sensor Data and GIS
• タイトル / 言語: en
• タイトル: Train Route Detection with using Sensor Data and GIS
• 言語: eng
• キーワード / 主題Scheme: Other
• キーワード / 主題: 位置情報と交通システム
• 資源タイプ / 資源タイプ識別子: http://purl.org/coar/resource_type/c_18gh
• 資源タイプ: technical report
• 著者所属: IBM Research - Tokyo
• 著者所属(英): en
• 著者所属(英): IBM Research - Tokyo
• 著者名: Masaki, Ono
• 著者名(英): Masaki, Ono
• 論文抄録 / 内容記述タイプ: Other
• 論文抄録 / 内容記述: Smart phones have many sensors (GPS/pedometer), so data can easily be obtained to estimate a user's context, which is an important factor for smart phone applications. In this paper, we present a method to detect past train routes (a sequence of stations) from GPS records using station location information. Although GPS sensors do not work well underground or in overcrowded areas, we solve this problem by thinking through all possible paths and weighting them using sensor records. To evaluate our method, we conducted an experiment with sensor data created by several people over more than two weeks and found that its accuracy exceeded the baseline accuracy.
• 論文抄録(英) / 内容記述タイプ: Other
• 論文抄録(英) / 内容記述: Smart phones have many sensors (GPS/pedometer), so data can easily be obtained to estimate a user's context, which is an important factor for smart phone applications. In this paper, we present a method to detect past train routes (a sequence of stations) from GPS records using station location information. Although GPS sensors do not work well underground or in overcrowded areas, we solve this problem by thinking through all possible paths and weighting them using sensor records. To evaluate our method, we conducted an experiment with sensor data created by several people over more than two weeks and found that its accuracy exceeded the baseline accuracy.
• 書誌レコードID / 収録物識別子タイプ: NCID
• 書誌レコードID / 収録物識別子: AA11851388
• 書誌情報: 研究報告モバイルコンピューティングとパーベイシブシステム（MBL）; 巻 2016-MBL-80, 号 16, p. 1-6, 発行日 2016-08-17
• ISSN / 収録物識別子タイプ: ISSN
• ISSN / 収録物識別子: 2188-8817
• Notice: SIG Technical Reports are nonrefereed and hence may later appear in any journals, conferences, symposia, etc.
• 出版者 / 言語: ja
• 出版者: 情報処理学会