Schema Matching Based on Weighted Fuzzy Concept Lattice

This paper introduces a new schema matching approach based on weighted fuzzy concept lattice. The procedure contains three steps. Firstly, we increase the evidence about each element being matched by applying naive Bayes classifier to classify the names and descriptions of the elements. Secondly, we use weighted fuzzy concept lattice to integrate the classified results as well as type messages and constrains. At last, a structural similarity measure is introduced to calculate the final matching. We present experimental results that demonstrate WFCL-based matching outperforms direct matching (without the benefit of WFCL).