COPEN: a CORBA-based intelligent push-engine

Abstract

Searching the right information on the Internet is a time-consuming task, mainly due to the large amount of information flow on the network and the nature of HTTP protocol. Push technology has been introduced in order to increase the search efficiency by providing only the pre-selected types of information. However, current push technique employs a simple method of defining users' preferences; letting the users select the right items in the menu or type the URLs interested. As a result, the information pushed by the push-engine does not satisfy much of the users' interest, lacking the search accuracy. Moreover, the HTML pages are transmitted to the user without any factoring, abstraction or re-packaging, yielding a large amount of information flow. In this paper, we present an intelligent push-engine, COPEN, that is based on Java & CORBA. The client part of COPEN monitors and intercepts the interaction between the user and the Web server in order to intelligently formulated the interests of the user. Based on the interests formulated, the server side of COPEN searches the Web, packages the information and transmits the packaged information to the user's local database. In order to overcome the storage access restriction of Java Security Manager, we developed an Java/ORB-based interceptor that can manage the cache database on the client's computer. As the result, the user's interest is dynamically and silently monitored, maintained in user profiles, information is intelligently searched based on the user's interests, and pre-transmitted to the cache database for faster retrieval.