Coincidencia y equiparación en los modelos de recuperación de información

Abstract

This paper is aimed at analysing the procedures by which the most renowned information retrieval models allow to lit question and answer according to underlying paradigms. It becomes so far crucial to establish a priori whether is searched for the adequacy between the terms for question and retrieved documents, or conversely, a semantic approach between both parts. In looking for exact match, the notion of f itness appears rather complete. However, there remain a pleyade of aditional possibilities, such as those based in similarity, question-answer or document-document, probability of relevance of documents against question, adjacency or nearbiness within the documental space, in co-wording, cociting, co-sitation, among others.It is concluded that, while the Boolean model search for exact matching, the remaining approaches show a tendency to cohere question and answer through different formulae.