Shuo Yang's paper (Efficiently Answering Technical Questions — A Knowledge Graph Approach) is accepted by AAAI 2017.

In this paper, we propose a novel question understanding method. We construct a hierarchical knowledge graph automatically which contains abundant technical information, to understand a question. Next, an augmented knowledge graph is used for relating the questions to documents via the knowledge graph. Then we use a question-driven method to select candidate solutions. Finally, random walk approach is applied to candidates ranking and an index-based random walk is proposed to support the online search.