構造化データから知識を発見するデータマイニングシステム

• 研究課題/領域番号: 13680459
• 研究種目: 基盤研究(C)
• 配分区分: 補助金
• 応募区分: 一般
• 研究分野: 知能情報学
• 研究機関: 広島市立大学
• 研究代表者: 宮原 哲浩 広島市立大学, 情報科学部, 助教授 (90209932)
• 研究分担者: 久保山 哲二 東京大学, 国際産学共同研究センター, 助手 (80302660) ; 正代 隆義 九州大学, 大学院・システム情報科学研究院, 助教授 (50226304) ; 内田 智之 広島市立大学, 情報科学部, 助教授 (70264934)
• 研究期間 (年度): 2001 – 2003
• 研究課題ステータス: 完了 (2003年度)
• 配分額: 3,600千円 (直接経費: 3,600千円); 2003年度: 900千円 (直接経費: 900千円); 2002年度: 900千円 (直接経費: 900千円); 2001年度: 1,800千円 (直接経費: 1,800千円)
• キーワード: データマイニング / 知識発見 / グラフ構造データ / 半構造データ / 木構造パターン / HTML / XMLファイル / 帰納推論

研究概要

本研究課題の目的は,グラフ構造や木構造を持つデータから知識を発見するデータマイニングシステムについて,その理論的基礎を確立し,知識発見システムを実現することである.インターネットの発展に伴い,Web文書も急速に増大している,本研究の目的の一つは,HTML/XMLファイルのような木構造を持つWeb文書から知識を発見することである.このようなWeb文書は,半構造データと呼ばれており,半構造データからのデータマイニングやテキストマイニングが注目を集めている.半構造Web文書から,意味がある知識を抽出するためには,まず,それらに頻出する木構造パターンを発見することが必要である.正事例とみなされる木構造データの構造的特徴を表すために,極大頻出タグ木パターンという木構造パターンを提案した.半構造データを非順序木とみなす場合と,順序木とみなす場合の両方について,極大頻出タグ木パターンをすべて生成するアルゴリズムを提案した.この発見アルゴリズムを実現し,その有効性を確認した.半構造データからのデータマイニングの理論的基礎を与えるため,項木という木構造パターンが表現する言語の学習可能性を調べた.項木言語の学習アルゴリズムを利用して,与えられた木構造データに共通な構造的パターンを抽出する手法を開発し,本研究課題の目標とするシステムの基本的データマイニング手法とすることができる.そのため,木構造データを,非順序木とみなす場合と,順序木とみなす場合の両方について,項木の表現能力,データ提示および質問に関する様々な設定のもとで,項木言語の学習アルゴリズムを与えた.

研究成果

• [文献書誌] Tetsuhiro Miyahara et al.: "Discovery of frequent tree structured patterns in semistructured web documents"Proc.PAKDD-2001, Lecture Notes in Artificial Intelligence, Springer-Verlag. 2035. 47-52 (2001)
  • 説明: 「研究成果報告書概要(和文)」より
• [文献書誌] Takayoshi Shoudai et al.: "Polynomial time algorithms for finding unordered tree patterns with internal variables"Proc.FCT-2001, Lecture Notes in Computer Science, Springer-Verlag. 2138. 335-346 (2001)
  • 説明: 「研究成果報告書概要(和文)」より
• [文献書誌] Osamu Maruyama et al.: "Learning conformation rules"Proc.DS-2001, Lecture Notes in Computer Science, Springer-Verlag. 2226. 243-257 (2001)
  • 説明: 「研究成果報告書概要(和文)」より
• [文献書誌] Tetsuhiro Miyahara et al.: "Discovery of frequent tag tree patterns in semistructured web documents"Proc.PAKDD-2002, Lecture Notes in Artificial Intelligence, Springer-Verlag. 2336. 341-355 (2002)
  • 説明: 「研究成果報告書概要(和文)」より
• [文献書誌] Kazuyoshi Furukawa et al.: "Extracting characteristic structures among words in semistructured documents"Proc.PAKDD-2002, Lecture Notes in Artificial Intelligence, Springer-Verlag. 2336. 356-367 (2002)
  • 説明: 「研究成果報告書概要(和文)」より
• [文献書誌] Yusuke Suzuki et al.: "Polynomial Time Inductive Inference of Ordered Tree Patterns with Internal Structured Variables from Positive Data"Proc.COLT02, Lecture Notes in Artificial Intelligence, Springer-Verlag. 2375. 169-184 (2002)
  • 説明: 「研究成果報告書概要(和文)」より
• [文献書誌] Yusuke Suzuki et al.: "A Polynomial Time Matching Algorithm of Structured Ordered Tree Patterns for Data Mining from Semistructured Data"Proc.ILP02, Lecture Notes in Artificial Intelligence, Springer-Verlag. 2583. 270-284 (2003)
  • 説明: 「研究成果報告書概要(和文)」より
• [文献書誌] Yusuke Suzuki et al.: "Ordered Term Tree Languages Which Are Polynomial Time Inductively Inferable from Positive Data"Proc.ALT02, Lecture Notes in Artificial Intelligence, Springer-Verlag. 2533. 188-202 (2002)
  • 説明: 「研究成果報告書概要(和文)」より
• [文献書誌] Osamu Maruyama et al.: "Toward drawing an atlas of hypothesis classes"Proc.DS-2002, Lecture Notes in Computer Science, Springer-Verlag. 2534. 220-232 (2002)
  • 説明: 「研究成果報告書概要(和文)」より
• [文献書誌] Satoshi Matsumoto et al.: "Learning of Finite Unions of Tree Patterns with Internal Structured Variables from Queries"Proc.AI02, Lecture Notes in Artificial Intelligence, Springer-Verlag. 2557. 523-534 (2002)
  • 説明: 「研究成果報告書概要(和文)」より
• [文献書誌] Tetsuhiro Miyahara et al.: "Extraction of Tag Tree Patterns with Contractible Variables from Irregular Semistructured data"Proc.PAKDD03, Lecture Notes in Artificial Intelligence, Springer-Verlag. 2637. 430-436 (2003)
  • 説明: 「研究成果報告書概要(和文)」より
• [文献書誌] Yuko Itokawa et al.: "Finding Frequent Subgraphs from Graph Structured Data with Geometric Information and Its Application to Lossless Compression"Proc.PAKDD03, Lecture Notes in Artificial Intelligence, Springer-Verlag. 2637. 582-594 (2003)
  • 説明: 「研究成果報告書概要(和文)」より
• [文献書誌] Yusuke Suzuki et al.: "Efficient Learning of Unlabeled Term Trees with Contractible Variables from Positive Data"Proc.ILP03, Lecture Notes in Artificial Intelligence, Springer-Verlag. 2835. 347-364 (2003)
  • 説明: 「研究成果報告書概要(和文)」より
• [文献書誌] Kazunori Yamagata et al.: "An Effective Grammar-Based Compression Algorithm for Tree Structured Data"Proc.ILP03, Lecture Notes in Artificial Intelligence, Springer-Verlag. 2835. 383-400 (2003)
  • 説明: 「研究成果報告書概要(和文)」より
• [文献書誌] Yusuke Suzuki et al.: "Efficient Learning of Ordered and Unordered Tree Patterns with Contractible Variables."Proc.ALT03, Lecture Notes in Artificial Intelligence, Springer-Verlag. 2842. 114-128 (2003)
  • 説明: 「研究成果報告書概要(和文)」より
• [文献書誌] Satoshi Matsumoto et al.: "Learning of Finite Unions of Tree Patterns with Repeated Internal Structured Variables from Queries"Proc.ALT03, Lecture Notes in Artificial Intelligence, Springer-Verlag. 2842. 144-158 (2003)
  • 説明: 「研究成果報告書概要(和文)」より
• [文献書誌] Sachio Hirokawa et al.: "Semi-Automatic Construction of Metadata from a Series of Web Documents."Proc.AI03, Lecture Notes in Artificial Intelligence, Springer-Verlag. 2903. 942-953 (2003)
  • 説明: 「研究成果報告書概要(和文)」より
• [文献書誌] T.Miyahara, T.Shoudai, T.Uchida, K.Takahashi, H.Ueda: "Discovery of frequent tree structured patterns in semistructured web documents"Proceedings of the 5tg Pacific-Asia Conference on Knowledge Discovery and Data Mining (PAKDD-2001) (Springer-Verlag) Lecture Notes in Artificial Intelligence. Vol.2035. 47-52 (2001)
  • 説明: 「研究成果報告書概要(欧文)」より
• [文献書誌] T.Shodai, T.Uchida, T.Miyahara: "Polynomial time algorithms for finding unordered tree patterns with internal variables"Proceedings of the 13th International Symposium on Fundamentals of Computation Theory (FCT 2001) (Springer-Verlag) Lecture Notes in Computer Science. Vol.2138. 335-346 (2001)
  • 説明: 「研究成果報告書概要(欧文)」より
• [文献書誌] O.Maruyama, T.Shoudai, E.Furuichi, S.KUhara: "Learning Conformation Rules"Proceedings of the 4nd International Conference of Discovery Science (DS-2001) (Springer-Verlag) Lecture Notes in Artificial Intelligence. Vol.2226. 243-257 (2001)
  • 説明: 「研究成果報告書概要(欧文)」より
• [文献書誌] T.Miyahara, Y.Suzuki, T.Shoudai, T.Uchida, K.Takahashi, H.Ueda: "Discovery of Frequent Tag Tree Patterns in Semistructured Web Document"Proceedings of the 6th Pacific-Asia Conference on Knowledge Discovery and Data Mining (PAKDD-2002) (Springer-Verlag) Lecture Notes in Artificial Intelligence. Vol.2336. 341-355 (2002)
  • 説明: 「研究成果報告書概要(欧文)」より
• [文献書誌] K.Furukawa, T.Uchida, K.Yamada, T.Miyahara, T.Shoudai, Y.Nakamura: "Extracting Characteristic Structures among Words in Semistructured Documents"Proceedings 6th Pacific-Asia Conference on Advances in Knowledge Discovery and Data Mining (PAKDD-2002) (Springer-Verlag) Lecture Notes in Artificial Intelligence. Vol.2336. 356-367 (2002)
  • 説明: 「研究成果報告書概要(欧文)」より
• [文献書誌] Y.Suzuki R.akanuma, T.Shoudai, T.Miyahara, T.Uchida: "Polynomial Time Inductive Inference of Ordered Tree Languages with Height-Constrained Variables from Positive Data"Proceedings of the 15th Annual Conference on Inductive Logic Programming (ILP-2002) (Springer-Verlag) Lecture Notes in Artificial Intelligence. Vol.2375. 169-184 (2002)
  • 説明: 「研究成果報告書概要(欧文)」より
• [文献書誌] Y.Suzuki, K.Inomae, T.Shoudai, T.Miyahara, T.Uhicda: "A polynomial time matching algorithm of structured ordered tree patterns for data mining from semistructured data"Proceedings of the 12th International Conference on Inductive Logic Programming (ILP-2002) (Springer-Verlag) Lecture Notes in Artificial Intelligence. Vol.2583. 270-284 (2003)
  • 説明: 「研究成果報告書概要(欧文)」より
• [文献書誌] Y.Suzuki, T.Shoudai, T.Uchida, T.Miyahara: "Ordered Term Tree Languages Which Are Polynomial Time Inductively Inferable from Positive Data"Proceedings 13th International Conference on Algorithmic Learning Theory (ALT-2002) (Springer-Verlag) Lecture Notes in Artificial Intelligence. Vol.2533. 188-203 (2002)
  • 説明: 「研究成果報告書概要(欧文)」より
• [文献書誌] O.Maruyama, T.shoudai, S.Miyano: "Toward Drawing an Atlas of Hypothesis Classes"Proceedings of the 5th International Conference on Discovery Science (DS-2002) (Springer-Verlag) Lecture Notes in Computer Science. Vol.2534. 220-232 (2002)
  • 説明: 「研究成果報告書概要(欧文)」より
• [文献書誌] S.Mtsumoto, T.shoudai, T.Miyahara, T.Uchida: "Learning of Finite Unions of Tree Patterns with Repeated Internal Structured Variables from Queries"Proceedings of the 15th Australian Joint Conference on Artificial Intelligence (AI-2002) (Springer-Verlag) Lecture Notes in Artificial Intelligence. Vol.2557. 523-534 (2002)
  • 説明: 「研究成果報告書概要(欧文)」より
• [文献書誌] T.Miyahara, Y.suzuki, T.Shoudai, T.Uchida, S.Hirokawa, K.Takahashi, H.Ueda: "Extraction of Tag Tree Patterns with Contractible Variables from Irregular semistructured data"Proceedings of the 7th Pacific-Asia Conference on Knowledge discovery and Data Mining (PAKDD-2003) (Springer-Verlag) Lecture Notes in Artificial Intelligence. Vol.2637. 430-436 (2003)
  • 説明: 「研究成果報告書概要(欧文)」より
• [文献書誌] Y.Itokawa, T.Uchida, T.Shoudai, T.Miyahara, Y.Nakamura: "Finding Frequent Subgraphs from Graph Structured Data with Geometric Information and Its Application to Lossless Compression"Proceedings of the 7th Pacific-Asia Conference on Knowledge Discovery and Data Mining (PAKDD-2003) (Springer-Verlag) Lecture Notes in Artificial Intelligence. Vol.2637. 582-594 (2003)
  • 説明: 「研究成果報告書概要(欧文)」より
• [文献書誌] Y.Suzuki, T.Shoudai, S.Matsumoto, T.Uchida: "Efficient Learning of Unlabeled Term Trees with Contractible Variables from Positive Data"Proceedings of the 13th International Conference on Inductive Logic Programming (ILP-2003) (Springer-Verlag) Lecture Notes in Artificial Intelligence. Vol.2835. 347-364 (2003)
  • 説明: 「研究成果報告書概要(欧文)」より
• [文献書誌] K.Yamagata, T.Uchida, T.Shoudai, Y.Nakamura: "An Effective Grammar-Based Compression Algorithm for Tree Structured Data"Proceedings 13th International Conference on Inductive Logic Programming (ILP-2003) (Springer-Verlag) Lecture Notes in Artificial Intelligence. Vol.2835. 383-400 (2003)
  • 説明: 「研究成果報告書概要(欧文)」より
• [文献書誌] Y.Suzuki, T.Shoudai, S.Matsumoto, T.Uchida, T.Miyahara: "Efficient Learning of Ordered and Unordered Tree Patterns with Contractible Variables"Proceedings of the 14th Work-shop on Algorithmic Learning Theory (ALT-2003) (Springer-Verlag) Lecture Notes in Artificial Intelligence. Vol.2842. 114-128 (2003)
  • 説明: 「研究成果報告書概要(欧文)」より
• [文献書誌] S.Matsumoto, Y.Suzuki, T.Shoudai, T.Miyahara, T.Uchida: "Learning of Finite Unions of Tree Patterns with Repeated Internal Structured Variables from Queries"Proceedings of the 14th Workshop on algorithmic Learning Theory (ALT-2003) (Springer-Verlag) Lecture Notes in Artificial Intelligence. Vol.2842. 144-158 (2003)
  • 説明: 「研究成果報告書概要(欧文)」より
• [文献書誌] S.Hirokawa, E.Itoh, T.Miyahara: "Semi-Automatic Construction of Metadata from a Series of Web Documents"Proceedings of the 16th Australian Joint Conference on Artificial (AI-2003) (Springer-Verlag) Lecture Notes in Artificial Intelligence. Vol.2903. 942-953 (2003)
  • 説明: 「研究成果報告書概要(欧文)」より
• [文献書誌] Yusuke Suzuki: "Efficient Learning of Unlabeled Term Trees with Contractible Variables from Positive Data"Proc.ILP03, Lecture Notes in Artificial Intelligence, Springer-Verlag. 2835. 347-364 (2003)
• [文献書誌] Kazunori Yamagata: "An Effective Grammar-Based Compression Algorithm for Tree Structured Data"Proc.ILP03, Lecture Notes in Artificial Intelligence, Springer-Verlag. 2835. 383-400 (2003)
• [文献書誌] Yusuke Suzuki: "Efficient Learning of Ordered and Unordered Tree Patterns with Contractible Variables."Proc.ALT03, Lecture Notes in Artificial Intelligence, Springer-Verlag. 2842. 114-128 (2003)
• [文献書誌] Satoshi Matsumoto: "Learning of Finite Unions of Tree Patterns with Repeated Internal Structured Variables from Queries"Proc.ALT03, Lecture Notes in Artificial Intelligence, Springer-Verlag. 2842. 144-158 (2003)
• [文献書誌] Sachio Hirokawa: "Semi-Automatic Construction of Metadata from a Series of Web Documents."Proc.ALT03, Lecture Notes in Artificial Intelligence, Springer-Verlag. 2903. 942-953 (2003)
• [文献書誌] Tetsuhiro Miyahara: "Discovery of Maximally Frequent Tag Tree Patterns with Contractible Variables from Semistructured Documents"Proc.PAKDD04, Lecture Notes in Artificial Intelligence, Springer-Verlag. (発表予定). (2004)
• [文献書誌] Tomoyuki Uchida: "Finding Frequent Structural Features among Words in Tree-Structured Documents"Proc.PAKDD04, Lecture Notes in Artificial Intelligence, Springer-Verlag. (発表予定). (2004)
• [文献書誌] Yusuke Suzuki: "Polynomial Time Inductive Inference of Ordered Tree Patterns with Internal Structured Variables from Positive Data"Proc. COLT02, Lecture Notes in Artificial Intelligence, Springer-Verlag. 2375. 169-184 (2002)
• [文献書誌] Yusuke Suzuki: "Ordered Term Tree Languages Which Are Polynomial Time Inductively Inferable from Positive Data"Proc. ALT02, Lecture Notes in Artificial Intelligence, Springer-Verlag. 2533. 188-202 (2002)
• [文献書誌] Satoshi Matsumoto: "Learning of Finite Unions of Tree Patterns with Internal Structured Variables from Queries"Proc.A102, Lecture Notes in Artificial Intelligence, Springer-Verlag. 2557. 523-534 (2002)
• [文献書誌] Yusuke Suzuki: "A Polynomial Time Matching Algorithm of Structured Ordered Tree Patterns for Data Mining from Semistructured Data"Proc. ILP02, Lecture Notes in Artificial Intefligence, Springer-Verlag. 2583. 270-284 (2002)
• [文献書誌] Tetsuhiro Miyahara: "Extraction of Tag Tree Patterns with Contractible Variables from Irregular Semistructured data"Proc. PAKDD03, Lecture Notes in Artificial Intelligence, Springer-Verlag. (発表予定). (2003)
• [文献書誌] Yuko Itokawa: "Finding Frequent Subgraphs from Graph Structured Data with Geometric Information and Its Application to Lossless Compression"Proc. PAKDD03, Lecture Notes in Artificial Intelligence, Springer-Verlag. (発表予定). (2003)
• [文献書誌] Tetsuhiro Miyahara: "Discovery of Frequent Tree Structured Patterns in Semistructured Web Documents"Lecture Notes in Artificial Intelligence, Springer-verlag. 2035. 47-52 (2001)
• [文献書誌] Takayoshi Shoudai: "Polynomial Time Algorithms for Finding Unordered Tree Patterns with Internal Variables"Lecture Notes in Computer Science, Springer-verlag. 2138. 335-346 (2001)
• [文献書誌] Tetsuhiro Miyahara: "Discovery of Frequent Tag Tree Patterns in Semistructured Web Documents"Lecture Notes in Artificial Intelligence, Springer-verlag. 2336. (2002)
• [文献書誌] Kazuyoshi Furukawa: "Extracting Characteristic Structures among Words in Semistructured Documents"Lecture Notes in Artificial Intelligence, Springer-verlag. 2336. (2002)