Applying Semantic Social Graphs to Disambiguate Identity References

摘要: Person disambiguation monitors web appearances of a person by disambiguating information belonging to different people sharing the same name. In this paper we extend incorporate abstract notion identity. This extension utilises semantic technologies represent identity be found and resources disambiguated as graphs. Our approach extracts complete social graph from distributed Web 2.0 services. containing possible references are converted into graphs describing available features. We disambiguate these identify correct performing random walks through space, measuring distances between resource graphs, clustering similar resources. present new distance measure called "Optimum Transitions" evaluate accuracy our using retrieval f-measure.