LSI-based semantic characterisation for automated text categorisation

Tan, Ping Ping (2009) LSI-based semantic characterisation for automated text categorisation. Masters thesis, Universiti Malaysia Sarawak.

PDF (Please get the password by email to repository@unimas.my , or call ext: 3914 / 3942 / 3933) LSI-based semantic characterization for automated text categorization (fulltext).pdf Restricted to Registered users only Download (17MB)

Abstract

As knowledge acquisition remains a bottleneck, incorporating human judgement within intelligent systems is still a challenge. Supervised learning methods have shown to be able to assist humans in automated text categorization (ATC). However, the performance of such systems is largely dependent on the characteristics of the datasets. Without the understanding of why a classifier works well for certain datasets, it is difficult to generalise its application across domains. Furthermore, most training sets used in supervised ATC have category labels provided by human experts. Expert knowledge used in the task of categorization is often not captured via the mere process of manipulating category labels. This has resulted in lose of intended meanings while performing supervised ATC. Besides that, large text datasets often contain a greater deal of noise.

Actions (For repository members only: login required)

View Item