Avoiding Chinese Whispers: Controlling End-to-End Join Quality in Linked Open Data Stores

摘要: Today Linked Open Data is a central trend in information provisioning. collected distributed data stores, individually curated with high quality, and made available over the Web for wide variety of applications providing their own business logic utilization. Thus, key promise to provide holistic view range items or entities. But parallel problems database integration schema matching, linking several sources remains challenge currently severely hampering vision working Semantic Web. One possible solution are instance matching systems that automatically create owl:sameAs links between stores. According existing benchmarks, quality has even reached satisfying level. However, our extensive analysis shows not yet ready large-scale interlinking. This because query processors joining via single incorrectly created link implicitly use also all transitive may turn be mismatched again. The result similar game Chinese Whispers: watered-down sameAs semantics step-by-step lead terrible end-to-end joins. We develop innovative structural mechanisms on top significantly improve processing avoiding Whispers.