The Mathematical Model of Hybrid Schema Matching based on Constraints and Instances Similarity

Schema matching is a crucial issue in applications that involve multiple databases from heterogeneous sources. Schema matching evolves from a manual process to a semi-automated process to effectively guide users in finding commonalities between schema elements. New models are generally developed using a combination of methods to improve the effectiveness of schema matching results. Our previous research has developed a prototype of hybrid schema matching utilizing a combination of constraints-based method and an instance-based method. The innovation of this paper presents a mathematical formulation of a hybrid schema matching model so it can be run for different cases and becomes the basis of development to improve the effectiveness of output and or efficiency during schema matching process. The developed mathematical model serves to perform the main task in the schema matching process that matches the similarity between attributes, calculates the similarity value of the attribute pair, and specifies the matching attribute pair. Based on the test results, a hybrid schema matching model is more effective than the constraints-based method or instance-based method run individually. The more matching criteria used in the schema matching provide better mapping results. The model developed is limited to schema matching processes in the relational model database.