Semantic Keyword Query Mechanism Based on Entity Association Graph in Dataspaces

Currently keyword query is a hot research topic in dataspaces. Due to its unstructured feature it's difficult to understand the user's query intent and to explain the query properly. So we propose semantic keyword query mechanism based on entity association graph. First we propose a layered graph model to capture the associations between entities and entity classes. Then we give association finding and Goal Entity Class finding methods. Finally we propose three types of queries based on association graphs. We conduct experiments to compare the precision and recall of queries which show effectiveness and correctness of our approach.