Skip to Main Content
The relation among data objects is not strictly symmetric and transitive in data grid environment, so it is rather difficult to compare their relativity. Combining membership with rough set and introducing them into information retrieval, this paper proposes a fuzzy rough set retrieval model. First the relationship between two classes is considered sufficiently to construct similarity relationship for classifying index space, then the membership of index word in regard to data object is designed with fuzzy set as theoretical principle, and lower and upper approximation of data object with respect to index space is defined with rough set as theoretical principle, finally the retrieval result is ranked and outputted. Experiment emulation shows the feasibility and validity of this algorithm.