Title: Relevant estimation among fields using field association words

Authors: Akihiro Tanaka, El-Sayed Atlam, Kazukiro Morita, Yohei Tsukuda, Masao Fuketa, Jun-ichi Aoe

Addresses: Department of Computer Science and Information Management, Tokushima University, Japan. ' Department of Computer Science and Information Management, Tokushima University, Japan. ' Department of Computer Science and Information Management, Tokushima University, Japan. ' Department of Computer Science and Information Management, Tokushima University, Japan. ' Department of Computer Science and Information Management, Tokushima University, Japan. ' Department of Computer Science and Information Management, Tokushima University, Japan

Abstract: In recent years, there has been a tremendous growth of online text information related to the explosive growth of the web which provides a very useful information resource to all types of users who access the Internet for various purposes. The major demand of these users is to get required information within the stipulated time. Humans can recognise subjects of document fields by reading only some relevant specific words called field association words in the field. This paper presents a method of relevant estimation among fields by using field association words. Two methods are proposed in this paper: first is a method of extraction of co-occurrence among fields and the second is a method of judgment of similarity among fields as the methods of relevant estimation among fields. From experimental results, precision of the first method is high when relevance among fields is very high and considering direction of fields, preferable results are obtained in the second method.

Keywords: co-occurrence; field association words; similarity; relevance estimation; automatic extraction; subject recognition; information retrieval.

DOI: 10.1504/IJCAT.2009.026605

International Journal of Computer Applications in Technology, 2009 Vol.35 No.2/3/4, pp.296 - 306

Available online: 20 Jun 2009 *

Full-text access for editors Access for subscribers Purchase this article Comment on this article