An Evidential Model for Estimating the Salience of Attributes in Classification Schemes

TL;DRAbstract

In inexact information retrieval, it is often necessary to know the importance of the various descriptive features of the objects to be retrieved for appraising the extent to which they fit a query. Elegant solutions to this problem have been developed for documents with textual, flat formats but not for objects with complex, structural representations, highly interrelated and organised according to classification schemes that change frequently. Information structured according to this latter object-oriented form is normally found in engineering databases, such as software repositories. This paper presents a model for estimating the importance of the various descriptive features of such objects on the basis of their modelling in classification schemes. The model does not require any sort of importance related information from the user, is computationaly effective and has been applied successfully, as a constituent of a computational model of similarity, to the retrieval of software art

Chat with Paper

AI Agents for this Paper