UUM Repository | Universiti Utara Malaysian Institutional Repository
FAQs | Feedback | Search Tips | Sitemap

A knowledge based system for automatic classification of web pages


Fathy, Sherif Kassem (2006) A knowledge based system for automatic classification of web pages. In: Knowledge Management International Conference and Exhibition 2006 (KMICE 2006), 6-8 June 2006, The Legend Hotel Kuala Lumpur.

[img] PDF
Restricted to Registered users only

Download (52kB)

Abstract

The paper describes design and implementation of a new knowledge based system for Automatic Information Retrieval DataBase (AIRDB).AIRDB helps the end-user to cluster and classify web pages on the basis of information filtering combined with an Artificial Neural Network (ANN).The classification depends mainly on keyword indexes.A large sample set consists of 11043 web pages of several formats are collected automatically and randomly from various resources.The AIRDB feature selection algorithm is summarized.The feature selection depends upon stemming words of web page. Each stem word is generated with local profile. This local profile contains information that indicates the weight of each stem with the possible related classes of web pages.A statistical analysis process is illustrated to reduce the noise stems.The various components of the AIRDB are described.The knowledge based system is tested with various web pages that disseminate their content in English.The average discrimination performance of the AIRDB reaches 84%.

Item Type: Conference or Workshop Item (Paper)
Additional Information: ISBN 983-3282-90-3 Organized by: Faculty of Information Technology, UUM
Uncontrolled Keywords: Knowledge Based,Information Retrieval, Classification, Feature Selection, Artificial Neural Network (ANN),Stemming.
Subjects: T Technology > T Technology (General)
Divisions: College of Arts and Sciences
Depositing User: Mrs. Norazmilah Yaakub
Date Deposited: 02 Jul 2014 08:31
Last Modified: 02 Jul 2014 08:31
URI: http://repo.uum.edu.my/id/eprint/11537

Actions (login required)

View Item View Item