An Intelligent Question Answering System of the Liao Dynasty Based on Knowledge Graph

Abstract The Liao Dynasty was a minority regime established by the Khitan on grasslands of northern China. To promote and spread cultural knowledge Dynasty, an intelligent question-and-answer system is constructed based graph in historical field Dynasty. In traditional question answering system, quality answers not high due to incomplete data distinctive vocabulary. solve this problem, combination method database KB proposed realize answering, joint model Siamese LSTM fusion MatchPyramid for semantic matching between questions database. With model, it easy perform fusing sentence-level word-level interactive features through MatchPyramid. Furthermore, sentence with same semantics as user input retrieved database, answer corresponding returned result. experimental results show that our has achieved relatively good performance domain open-domain graph, improved accuracy answer.