Text Mining using PrefixSpan constrained by Item Interval and Item Attribute

Issei Sato, Yu Hirate, Hayato Yamana

研究成果: Conference contribution

抄録

Applying conventional sequential pattern mining methods to text data extracts many uninteresting patterns, which increases the time to interpret the extracted patterns. To solve this problem, we propose a new sequential pattern mining algorithm by adopting the following two constraints. One is to select sequences with regard to item intervals-The number of items between any two adjacent items in a sequence-And the other is to select sequences with regard to item attributes. Using Amazon customer reviews in the book category, we have confirmed that our method is able to extract patterns faster than the conventional method, and is better able to exclude uninteresting patterns while retaining the patterns of interest.

本文言語English
ホスト出版物のタイトルICDEW 2006 - Proceedings of the 22nd International Conference on Data Engineering Workshops
編集者Roger S. Barga, Xiaofang Zhou
出版社Institute of Electrical and Electronics Engineers Inc.
ページ35-38
ページ数4
ISBN(電子版)0769525717, 9780769525716
DOI
出版ステータスPublished - 2006
外部発表はい
イベント22nd International Conference on Data Engineering Workshops, ICDEW 2006 - Atlanta, United States
継続期間: 2006 4 32006 4 7

出版物シリーズ

名前ICDEW 2006 - Proceedings of the 22nd International Conference on Data Engineering Workshops

Other

Other22nd International Conference on Data Engineering Workshops, ICDEW 2006
国/地域United States
CityAtlanta
Period06/4/306/4/7

ASJC Scopus subject areas

  • 情報システム
  • コンピュータ ネットワークおよび通信
  • 情報システムおよび情報管理

フィンガープリント

「Text Mining using PrefixSpan constrained by Item Interval and Item Attribute」の研究トピックを掘り下げます。これらがまとまってユニークなフィンガープリントを構成します。

引用スタイル