Title: Building of field association terms based on links

Authors: Mahmoud Rokaya, El-Sayed Atlam

Addresses: Department of Information Systems, Faculty of Computers and Information Systems, Taif University, 808, El-Hawia, Taif, Saudi Arabia. ' Department of Information Systems, Faculty of Computers and Information Systems, Taif University, 808, El-Hawia, Taif, Saudi Arabia

Abstract: Studying co-word relations between field association (FA) terms and candidate terms can provide valuable information which helps the machine to take the correct decision and append the candidate terms in their suitable place inside FA terms dictionary. In this paper, a pure automatic tool to build a dynamic FA terms dictionary is described. The concept of link power as well as modifications of the rules used to determine the FA terms level depending on the concept of links will be presented. The results showed that the precession and recall were improved. In our simulation, the average improvement was about 16%. Also the complete dynamic system was built and tested.

Keywords: field association terms; co-word analysis; power links; concentration rules; concentration ratio; term co-occurrence; recall; precision; content analysis.

DOI: 10.1504/IJCAT.2010.034530

International Journal of Computer Applications in Technology, 2010 Vol.38 No.4, pp.298 - 305

Published online: 07 Aug 2010 *

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