An object-oriented, logic based approach to document retrieval

Abstract

There have been many advances in the field of information retrieval but some issues remain to be resolved. There is yet a best way of representing the contents of text documents and the user's information need so that they can be compared and relevant documents are retrieved. A comprehensive survey of conventional information retrieval systems has led to this proposal of an object-oriented, logic-based information retrieval approach. This approach recommends a way to model a document's semantic representation and to apply logic inferential rules in the document retrieval. The proposed system will accept an user's query in natural language form and returns a ranked set of documents. In such a framework, a document is relevant to a query if the latter can be deduced from a set of axioms associated with the document using inference rules. To model the document semantically, an object-oriented data model is used.