Authors:

Abstract

Providing relevant information through the determination of priorities for a user request is important as it saves both labour and time of the inquirer as well as the information provider. Therefore, the development of information retrieval models to compute these priorities as numerical representations of their relevancies is becoming a major task of the modern information retrieval arena. This paper introduces a new model for information retrieval. This model appears as a vector multiplication of the distances among the terms in the query with the distances among the same terms in the document. Therefore, it determines the relevancies among the query and the documents by matching the distances between each two terms in the query with the distances between the same terms in each document of the corpus. Also this paper uses an example to show that the new model is able to distinctly identify two or more documents according to the given query which cannot be done by using the vector space model when the documents include same term frequencies and same number of total terms.

Journal of the University Librarians Association, Sri Lanka, Vol. 17, Issue 2, January 2013, Page 119-135

DOI: http://dx.doi.org/10.4038/jula.v17i2.6649

Keywords: Information Retrieval, Vector Space Model, tf-idf, Distance Model