Improving Information Retrieval by Concept-Based Ranking

Abstract

With the Internet getting available to more and more people in the last decade and with the rapidly growing number of webpages the internet is a vast resource of information. Millions of people are using the internet to search for information every day. However, the search result is not satisfying in the case of an ambiguous search query. In this paper an algorithm for re-ranking pages according to concepts - the content of webpages — is proposed. This algorithm uses Association rules, a Data Mining technique, to derive concepts from webpages. The preliminary experiment shows a promising result compared with hyperlink-based ranking algorithms.