KPE: An Automatic Keyphrase Extraction Algorithm

摘要：

In this paper, an automatic Keyphrase extraction algorithm (KPE algorithm) is presented for single English document without using controlled vocabularies (corpus). It relies on n-gram filtration technique that filters complicated ngrams. The proposed algorithm proposesa ranking method to weight extracted keyphrases. The ranking method relies onthe position, frequency, as well as χ2-measure for testing the bias of co-occurrence distribution between frequent terms. Experimental results with publically available scientific text dataset show a comparable performance to both supervised algorithms (e.g. KEA) and unsupervised algorithms (e.g. Ngram filtration technique). Meanwhile, evaluations show that the proposed algorithm outperforms existing KEA and Ngram filtration technique methods producing keyphrases with higher precision and recall.

关键词： Automatickeyphrase extraction n-gram filtration technique KEA.

作者: Aliaa A.A. Youssif Atef Z.Ghalwash Islam A.Amer

作者单位: Computer Science DepartmentHelwan UniversityHelwan– Egypt Computer Science Department MSA University 6-October – Egypt

会议类型: 国际会议

会议名称: 2011 International Conference on Information System and Computational Intelligence(2011 IEEE信息系统与计算智能国际会议 ICISCI 2011)

会议地点: 哈尔滨

会议语种:英文

页码: 103-107

在线出版日期: 2011-01-18（万方平台首次上网日期，不代表论文的发表时间）