DYMS (Dynamic Matcher Selector) - Scenario-based Schema Matcher Selector

Schema matching is one of the main challenges in different information system integration contexts. Over the past 20 years, different schema matching methods have been proposed and shown to be successful in various situations. Although numerous advanced matching algorithms have emerged, schema matching research remains a critical issue. Different algorithms are implemented to resolve different types of schema heterogeneities, including differences in design methodologies, naming conventions, and the level of specificity of schemas, amongst others. The algorithms are usually too generic regardless of the schema matching scenario. This situation indicates that a single matcher cannot be optimized for all matching scenarios. In this research, I proposed a dynamic matcher selector (DYMS) as a probable solution to the aforementioned problem. The proposed DYMS analyzes the schema matching scenario and selects the most appropriate matchers for a given scenario. Selecting matchers are weighted based on the parameter optimization process, which adopts the heuristic learning approach. The DYMS returns the alignment result of input schemas.