Web document classification: managing context change
-
-
[1] University of Tasmania
University of Tasmania
Australia
-
- Localización: Proceedings of the IADIS International Conference WWW/INTERNET 2004: Madrid, Spain, October 6-9, 2004 / coord. por Pedro Isaías, Nitya Karmakar, Vol. 1, 2004 (Full papers), ISBN 972-99353-0-0, págs. 143-151
- Idioma: inglés
- Texto completo no disponible (Saber más ...)
-
Resumen
This paper focuses on the information management systems of the dynamic World Wide Web. Many individual web pages, such as news portals, provide periodic information and public announcements from various organizations. Pushing methods include semantic web methods or RSS (Really Simple Syndication), which is one of the related approaches to monitoring this type of information. Information filtering and classification of collected information are integral to information management systems because people have to locate and archive for the future, as well as for the present. One of the major problems in applying the current information filtering or classification approaches to this dynamic information is that the classification or filtering of knowledge is not static, so the system must be able to manage the change of knowledge in real time. Previous studies have proven that while the current approaches can handle presampled information very well, the maintenance issue is not well addressed. I-Web has been developed to locate the new information from the target (registered) Web sites and Multiple Classification Ripple Down Rules, the knowledge acquisition method, is applied to maintain the information classification knowledge in this system. The MCRDR-based document classification enables domain users to elicit their domain knowledge incrementally and revise their knowledge base (KB). They may then reclassify documents according to context changes. Our experiment results show the MCRDR document classifier performs these tasks successfully in the real world.
