Integration of information retrieval and database management systems

摘要

The problem of integrating database management systems and information retrieval systems has received increasing attention in recent years. In a database management environment, the records are formatted. The attributes with which the record characteristics and the user needs are described are precise. In contrast, an information retrieval system provides facilities for identifying references to documents, usually in textual form, from which the user information needs can be satisfied. The descriptors used to represent the content of the documents for this purpose are normally imprecise. The motivation for integrating these two types of systems is presented. Several features of such an integrated system in terms of what kinds of retrieval options such an integrated system should provide are identified. In particular, it is argued that the integrated system should allow the selection of the degree of structuredness of the search, depending on how precisely the user can specify his or her needs. A formulation of the underlying retrieval problem is presented. It is observed that the kind of integrated system considered has a broader scope of application than either one of its components.