Information Customization using SOMSE: A Self-Organizing Map Based Approach

Abstract

Information overload on the World-Wide Web is a well recognized problem. Research to subdue this problem and extract maximum benefit from the Internet is still in its infancy. Managing information overload on the Web is a challenge and the need for more precise techniques for assisting the user in finding the most relevant and most useful information is obvious. Search engines are very effective at filtering pages that match explicit queries. Search engines, however, require massive memory resources (to store an index of the Web) and tremendous network bandwidth (to create and continually refresh the index). These systems receive millions of queries per day, and as a result, the CPU cycles devoted to satisfying each individual query are sharply curtailed. There is no time for intelligence which is mandatory for offering ways to combat information overload. What is needed are systems, often referred to as information customization systems, that act on the user’s behalf and that can rely on existing information services like search engines that do the resource-intensive part of the work. These systems will be sufficiently lightweight to run on an average PC and serve as personal assistants. Since such an assistant has relatively modest resource requirements it can reside on an individual user’s machine. If the assistant resides on the user’s machine, there is no need to turn down intelligence. The system can have substantial local intelligence. In an attempt to circumvent the problems of search engines and contribute to resolving the problem of information overload over the Web, the authors propose SOMSE, a system that improves the quality of Web search by combining meta-search and unsupervised learning.