Abstract:

The World Wide Web is an omnipresent and ever-expanding source of data. The exponential increase of information on the web has affected the manner in which it is accessed, disseminated and delivered. The emphasis has shifted from mere surfing the web for information to efficient retrieval and monitoring of selective changes to its content. Hence, an effective monitoring system for change detection and notification based on user-profiles is needed. WebVigiL is a general-purpose, active capability-based information monitoring and notification system. It handles specification, management, and propagation of customized changes as requested by a user.

As change detection requires two versions of a page for comparison, and the same page is retrieved multiple times over a period of time, there is a need for storing, retrieving, and managing persistent versions of web pages of interest. This thesis presents an efficient mechanism to store, manage, and provide the appropriate pages as needed to perform change detection. It also deals with purging the versions when they are no longer needed. This thesis also deals with timely notifications of customized changes to users. It addresses various alternatives for notifying the users when change is detected. Finally, this thesis addresses presentation of changes as it is difficult to manually locate changes in a page. The changes are highlighted in order to present them in an easy-to-understand manner.