Research Of Question Answering Technology And Application Based On Natural Language Understanding

Question Answering Technology is a hot research field of Natural Language Understanding. Question Answering Technology is relative to many forefront and difficult research techniques. At present, no mature products have appeared. The key technologies in this paper include the building of data base, question understanding technology and answers matching technology.This paper is mainly research on Question Answering Technology based on Natural Language Understanding. The main work of this paper is as the following:1. This paper designs a multi-level knowledge base and a knowledge representation method which can expand arbitrarily. It is easy facilitate to increase and update the knowledge in the knowledge base.2. The system is designed to be multilingual, so it is platform independence.3. This paper designs a word index tree. The paper improves the two-way word segmentation algorithm and other word segmentation technologies. It also improves the new word formation method based the statistics of word frequency.4. This paper designs an answer extraction technical which is based on a profuse knowledge base and the context of the talk scene. It designs a three-step question matching model.As can be seen from the test, The Question Answering Technology gets further improved for using the advanced NLU technology, knowledge representation method and answer extraction technology.The Question Answering Technology is relative to many techniques. This paper discussed three key technologies. There are some other technologies such as syntactic analysis, pronouns and name recognition need further study.