Improving Document Representation for Story Link Detection by Modeling Term Topicality

• Item type: Trans(1)
• 公開日: 2008-12-26
• タイトル: Improving Document Representation for Story Link Detection by Modeling Term Topicality
• タイトル / 言語: en
• タイトル: Improving Document Representation for Story Link Detection by Modeling Term Topicality
• 言語: eng
• キーワード / 主題Scheme: Other
• キーワード / 主題: 研究論文
• 資源タイプ / 資源タイプ識別子: http://purl.org/coar/resource_type/c_6501
• 資源タイプ: journal article
• 著者所属: National Institute of Informatics Tokyo Presently with the University of North Carolina at Chapel Hill USA
• 著者所属: National Institute of Informatics Tokyo Kobe University Kobe Japan
• 著者所属(英): en
• 著者所属(英): National Institute of Informatics, Tokyo,Presently with the University of North Carolina at Chapel Hill, USA
• 著者所属(英): en
• 著者所属(英): National Institute of Informatics, Tokyo,Kobe University, Kobe, Japan
• 著者名: Chirag, Shah; Koji, Eguchi
• 著者名(英): Chirag, Shah; Koji, Eguchi
• 論文抄録 / 内容記述タイプ: Other
• 論文抄録 / 内容記述: Several information organization access and filtering systems can benefit from different kind of document representations than those used in traditional Information Retrieval (IR). Topic Detection and Tracking (TDT) is an example of such a domain. In this paper we demonstrate that traditional methods for term weighing do not capture topical information and this leads to inadequate representation of documents for TDT applications. We present various hypotheses regarding the factors that can help in improving the document representation for Story Link Detection (SLD)－a core task of TDT. These hypotheses are tested using various TDT collections. From our experiments and analysis we found that in order to obtain a faithful representation of documents in TDT domain we not only need to capture a term's importance in traditional IR sense but also evaluate its topical behavior. Along with defining this behavior we propose a novel measure that captures a term's importance at the collection level as well as its discriminating power for topics. This new measure leads to a much better document representation as reflected by the significant improvements in the results.
• 論文抄録(英) / 内容記述タイプ: Other
• 論文抄録(英) / 内容記述: Several information organization, access, and filtering systems can benefit from different kind of document representations than those used in traditional Information Retrieval (IR). Topic Detection and Tracking (TDT) is an example of such a domain. In this paper we demonstrate that traditional methods for term weighing do not capture topical information and this leads to inadequate representation of documents for TDT applications. We present various hypotheses regarding the factors that can help in improving the document representation for Story Link Detection (SLD)－a core task of TDT. These hypotheses are tested using various TDT collections. From our experiments and analysis we found that in order to obtain a faithful representation of documents in TDT domain, we not only need to capture a term's importance in traditional IR sense, but also evaluate its topical behavior. Along with defining this behavior, we propose a novel measure that captures a term's importance at the collection level as well as its discriminating power for topics. This new measure leads to a much better document representation as reflected by the significant improvements in the results.
• 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: AA11464847
• 書誌情報: 情報処理学会論文誌データベース（TOD）; 巻 1, 号 3, p. 11-19, 発行日 2008-12-26
• ISSN / 収録物識別子タイプ: ISSN
• ISSN / 収録物識別子: 1882-7799
• 出版者 / 言語: ja
• 出版者: 情報処理学会