An information retrieval model based on modal logic

摘要

Most currently used information retrieval models are unsuitable to describe and integrate many recent techniques, namely those for the semantic-based retrievals. The aim of this paper is to construct a more general model that is able to describe the general functionalities of information retrieval systems. The implicit basis in all information retrieval systems is considered as logical implications. The measurement of the correspondence between a query and a document thus becomes the estimation of the implication strength. This estimation can be suitably described in terms of modal logic. The model is shown to be more general than the existing ones, such as the Boolean model, the probabilistic model and so on. A basic necessary constraint is given for the consistency of the correspondence evaluation.