First Experiments on the BMIR - J2 Collection using the NEAT System

• Item type: SIG Technical Reports(1)
• 公開日: 1998-05-14
• タイトル: First Experiments on the BMIR - J2 Collection using the NEAT System
• タイトル / 言語: en
• タイトル: First Experiments on the BMIR - J2 Collection using the NEAT System
• 言語: jpn
• 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: Research and Development Center Toshiba Corporation
• 著者所属: Research and Development Center Toshiba Corporation
• 著者所属: Research and Development Center Toshiba Corporation
• 著者所属: Research and Development Center Toshiba Corporation
• 著者所属(英): en
• 著者所属(英): Research and Development Center, Toshiba Corporation
• 著者所属(英): en
• 著者所属(英): Research and Development Center, Toshiba Corporation
• 著者所属(英): en
• 著者所属(英): Research and Development Center, Toshiba Corporation
• 著者所属(英): en
• 著者所属(英): Research and Development Center, Toshiba Corporation
• 著者名: Jones, Gareth; Sakai, Tetsuya; Kajiura, Masahiro; Sumita, Kazuo
• 著者名(英): Gareth, Jones; Tetsuya, Sakai; Masahiro, Kajiura; Kazuo, Sumita
• 論文抄録 / 内容記述タイプ: Other
• 論文抄録 / 内容記述: The BMIR-J2 Collection is the largest generally available information retrieval test set for Japanese text retrieval. This paper describes first experiments on BMIR-J2 using the Toshiba NEAT information retrieval system. Experimental retrieval results indicate that the probabilistic information retrieval model originally developed for English text transfers well to the Japanese language using both word-based and character-based indexing. Additional experiments report retrieval results comparing various approaches to query processing and retrieval performance for the different query categories defined within BMIR-J2.
• 論文抄録(英) / 内容記述タイプ: Other
• 論文抄録(英) / 内容記述: The BMIR-J2 Collection is the largest generally available information retrieval test set for Japanese text retrieval. This paper describes first experiments on BMIR-J2 using the Toshiba NEAT information retrieval system. Experimental retrieval results indicate that the probabilistic information retrieval model originally developed for English text transfers well to the Japanese language using both word-based and character-based indexing. Additional experiments report retrieval results comparing various approaches to query processing, and retrieval performance for the different query categories defined within BMIR-J2.
• 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: AN10114171
• 書誌情報: 情報処理学会研究報告情報学基礎（FI）; 巻 1998, 号 34(1998-FI-049), p. 57-64, 発行日 1998-05-14
• 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