본문 바로가기
  • Home

A Study on Information Resource Evaluation for Text Categorization

  • Journal of the Korean Society for Information Management
  • Abbr : JKOSIM
  • 2007, 24(4), pp.305~321
  • DOI : 10.3743/KOSIM.2007.24.4.305
  • Publisher : 한국정보관리학회
  • Research Area : Interdisciplinary Studies > Library and Information Science
  • Received : December 1, 2007
  • Accepted : December 1, 2007
  • Published : December 30, 2007

Eunkyung Chung 1

1이화여자대학교

Accredited

ABSTRACT

The purpose of this study is to examine whether the information resources referenced by human indexers during indexing process are effective on Text Categorization. More specifically, information resources from bibliographic information as well as full text information were explored in the context of a typical scientific journal article data set. The experiment results pointed out that information resources such as citation, source title, and title were not significantly different with full text. Whereas keyword was found to be significantly different with full text. The findings of this study identify that information resources referenced by human indexers can be considered good candidates for text categorization for automatic subject term assignment.

Journal Copyright Policy

BY-NC-ND 아이콘

This license enables reusers to copy and distribute the material in any medium or format in unadapted form only, for noncommercial purposes only, and only if attribution is given to the creator.

Citation status

* References for papers published after 2025 are currently being built.