Automatic Relevance Detection & Feedback in a Web Search Engine
Mark Grenham
December 2007
Abstract
This document outlines the functional speciļ¬cation for an Implicit Relevance Feedback (RF) implementation in an existing Information Retrieval (IR) System focusing on web search IR systems. The proposed relevance feedback implementation passively monitors users activities with search results and automatically updates the original query results with new ones based upon the output from the relevance feedback algorithm. This document also outlines the architecture of such a system and also the best algorithms and techniques chosen and the reasoning behind these choices.
Download
Download the complete document in PDF format.