MISC

2017年7月19日

Enumeration of Maximally Frequent Ordered Tree Patterns with Wildcards for Edge Labels

情報処理学会論文誌数理モデル化と応用(TOM)
  • Tetsuhiro Miyahara
  • ,
  • Yusuke Suzuki
  • ,
  • Takayoshi Shoudai
  • ,
  • Tomoyuki Uchida
  • ,
  • Tetsuji Kuboyama

10
2
開始ページ
59
終了ページ
69
記述言語
英語
掲載種別

We consider representing tree structured features of structured data which are represented by rooted trees with ordered children. As representations of tree structured features, we use ordered tree patterns, called ordered wildcard tree patterns, which have structures of rooted ordered trees, structured variables and wildcards for edge labels. A structured variable can be replaced with an arbitrary rooted ordered tree. First we show that it is hard to compute two types of optimum frequent ordered wildcard tree patterns. Then we present an algorithm for enumerating all maximally frequent ordered wildcard tree patterns. Finally we consider extended ordered wildcard tree patterns, called ordered tag tree patterns, which have structured variables, wildcards, tags and keywords, and present an algorithm for enumerating all maximally frequent ordered tag tree patterns.We consider representing tree structured features of structured data which are represented by rooted trees with ordered children. As representations of tree structured features, we use ordered tree patterns, called ordered wildcard tree patterns, which have structures of rooted ordered trees, structured variables and wildcards for edge labels. A structured variable can be replaced with an arbitrary rooted ordered tree. First we show that it is hard to compute two types of optimum frequent ordered wildcard tree patterns. Then we present an algorithm for enumerating all maximally frequent ordered wildcard tree patterns. Finally we consider extended ordered wildcard tree patterns, called ordered tag tree patterns, which have structured variables, wildcards, tags and keywords, and present an algorithm for enumerating all maximally frequent ordered tag tree patterns.

リンク情報
CiNii Articles
http://ci.nii.ac.jp/naid/170000148790
CiNii Books
http://ci.nii.ac.jp/ncid/AA11464803
URL
http://id.nii.ac.jp/1001/00182648/
ID情報
  • ISSN : 1882-7780
  • CiNii Articles ID : 170000148790
  • CiNii Books ID : AA11464803

エクスポート
BibTeX RIS