MatchBench: Benchmarking Schema Matching Algorithms for Schematic Correspondences

Schema matching algorithms aim to identify relationships between database schemas, which underpin a wide range of model management operations, such as merge, compose and difference, as required for tasks such as data integration, exchange or schema evolution. However, the results of matching algorithms are typically expressed as lowlevel, 1-to-1 associations between pairs of attributes or entity types, rather than as the higher-level characterisations of relationships required by many implementations of model management operations. This paper presents a benchmark for evaluating schema matching algorithms that is based on the well established classification of schematic heterogeneities of Kim et al., which explores the extent to which the matching algorithms are effective at diagnosing schematic heterogeneities. The paper contributes: (i) a wide range of scenarios that manifest different types of schematic heterogeneities; (ii) a collection of experiments over the scenarios that can be used to investigate the performance of different matching algorithms; and (iii) an application of the experiments for the evaluation of matchers from three well-known and publicly available schema matching platforms, namely COMA++, Rondo and OpenII.