Sen Hu's paper on Knowledge Base Question Answering (Answering Natural Language Questions by Subgraph Matching over Knowledge Graphs) is accepted by TKDE 2017.

This is an extension paper over previous work on Question Answering that is published on SIGMOD 2014. This paper also follow the semantic parsing based approach, it first translate the natural language question to semantic query graph, then find the appropriate matches over the knowledge graph. The framework in previous work is relation first which have some obstacles including can not tackle implicit relations. The core idea of the extension is to propose a node first framework to build semantic query graph, which can tackle implicit relations by utilizing data mining on offline data graph. It is robust because it does not rely on any templates to build the query graph structure and allow some false edge. Those false edges would be removed on the evaluation process according the matches. Thus the node first framework can tackle more complex questions. The experiment results show that our solution not only improves the precision but also speeds up query perfermence greatly.