An information retrieval model based on simple Bayesian networks

Abstract

In this article a new probabilistic information retrieval (IR) model, based on Bayesian networks (BNs), is proposed. We first consider a basic model, which represents only direct relationships between the documents in the collection and the terms or keywords used to index them. Next, we study two versions of an extended model, which also represents direct relationships between documents. In either case the BNs are used to compute efficiently, by means of a new and exact propagation algorithm, the posterior probabilities of relevance of the documents in the collection given a query. The performance of the proposed retrieval models is tested through a series of experiments with several standard document collections.

Structure-based inference of xml similarity for fuzzy duplicate detection

“…Later, a second model was proposed by Ribeiro-Neto and Muntz in [21], demonstrating that BNs can be effectively used to combine different types of information to further improve search results. More recently, Acid et al [1] presented a third model whose network topology is defined in such way that an exact propagation algorithm can be used to efficiently compute the relevance probabilities of documents.…”

The effectiveness of automatically structured queries in digital libraries

“…Bayesian network models were first used in IR problems by Turtle and Croft [36] and later by Ribeiro-Neto and Muntz [29] (upon whose work our model is based). More recently, Acid et al [1] further refined such models so that exact propagation algorithms can be used to efficiently compute probabilities. Bayesian networks also have been applied to other IR problems besides ranking as, for example, relevance feedback [24], automatic construction of hypertexts [33], query expansion [17], information filtering [9], ranking fusion [37], and document clustering and classification [7,18].…”

Probabilistic passage models for semantic search of genomics literature

“…The Belief Network Model (Ribeiro-Neto & Muntz, 1996;Silva, Ribeiro-Neto, Calado, Moura, & Ziviani, 2000) provides an efficient inference mechanism and is able to simulate vector space, probabilistic, and inference network models. More recently, the Bayesian Network Retrieval Model was designed to have a flexible topology that can take into account term relationships obtained from learning algorithms (Acid, de Campos, Fernández, & Huete, 2003). Each of these models is based on a directed graphical model, and assumes some causal relation between random variables.…”