An opportunity cost approach for measuring performance of an intelligent knowledge based search assistant

Abstract

The Web has myriad of useful information, but its dynamic, unstructured nature makes them difficult to locate the desired information. A general-purpose search engine, such as Google or AltaVista usually generates thousands of hits, many of them irrelevant to the user query. A knowledge based search assistant is developed which reduces the time and cost of information accumulating of common interest groups. When the users from a common network search on similar topics then an intelligent agent minimize the searching effort of a user by utilizing the previous experience of the users they have gathered from their surfing behaviours. Additionally the agent incrementally updates its database by analyzing its perception, which gradually increases its recall rate. A search assistant that accumulates knowledge from user activity and gathers information would provide a convenient searching environment with minimum effort within shortest possible time.