Searching Semantic Associations Based on Virtual Document

As the explosive growth of online linked data, enormous RDF triples are produced every minute in various fields such as health, transportation, chemical, etc. There is an urgent need for an approach to finding and searching semantic association from massive data. However, the complex graph structure of the semantic association brings a great barrier to the process of searching. Transforming the complex graph into text-based structure is a better idea. To characterize the semantics of each association, a virtual document of each association is built with the help of a neighboring operation. A searching model of virtual documents of associations and a ranking schema are also discussed in this paper. Experiments show that our approach is feasible and efficient.