Question answering method combining sequence model and knowledge graph

The invention discloses a sequence model and knowledge graph combined question answering method. The sequence model and knowledge graph combined question answering method comprises a knowledge acquisition processing method, a knowledge graph construction method, a model construction method and a question judgment and answer generation method. Corresponding question and answer strategies are designed for different types of data sets, field professional knowledge semantic data are constructed from field professional knowledge data sets, and professional answers in a knowledge graph are obtained through keyword extraction and template matching. According to the method, field knowledge basic question and answer semantic data is constructed from a field knowledge basic question and answer data set, and an optimal answer of a field knowledge chat data set is obtained by using a sentence vector similarity algorithm, so that a Seq2seq model is trained to serve as a supplement of a knowledge graph. In combination with