Problem Definition, goal and Importance:

Abstract

Document length can affect calculated relevance between a document and a query. Yet, information retrieval systems must deal with documents regardless of their length. Thus, document length normalization can be used to abstract relevance form length. A good retrieval effectiveness results when a normalization strategy retrieves documents with chances similar to their probability of relevance. Pivot length normalization tries to achieve this goal. It can be used to modify any normalization function, thereby reducing the gap between the relevance and the retrieval probabilities. Performance of a pivot length normalization system highly depends on the values of the two parameters: pivot and slope. These values are collection dependent. In order to adjust these values for best performance many works have been done on English language collections; however there is no similar work in Persian language. In this study, we will tune these values for Hamshahri document collection.