Bayesian Question Classification In The Field Of Research, Based On Incremental Improvement

Question Answering System, as the new generation of intelligent search engine, allows users to ask questions by means of natural language, and can supply more accurate answers compared with traditional search engines. There are five parts in the Question Answering System:Question Classification,Question Expansion,Search Engine,Answer Extraction and Answer Selection. Question Classification is an important part in the Question Answering System, which is the basis of making the answer extraction strategy and determining the answer semantic type, the classification results can directly affect the accuracy rate of the answer extraction.How to use the incremental training corpus to improve the Question Classification accuracy rate in the process of Question Classification based on the statistic learning. A Question Classification method based on the incremental modified Bayes was presented in this paper. The method used the modified Bayes and combined the incremental learning to correct the parameter by the incremental training set stage by stage, and established the Question Classification model based on the incremental modified Bayes. In the processing of researching the Question Classification model, we achieved the main distinctive achievements which are as follows:1) A method is proposed which is used to deal with the questions of the special domain. The method realized key words extraction and domain question expansion by lexical and semantic analysis; and dealt with the question redundancies and identified the denominate entity; and obtained the question type, the question focus and the answer type by the mapping rules of question type and answer type.2) A method of Question Classification based on the incremental modified Bayes is proposed. The method is based on the idea of the modified Bayes, which can afresh define the parameter of the arithmetic of Question Classification, and learn the question classifier by using the different training set and the different incremental training set, and then modify the parameter of the question classifier continuously, so the parameter of the question classifier can reach to the optimization. The result of the experiment shows that the method can achieve good effects on the accuracy rate and the time efficiency.3) Using the incremental modified Bayes method, we design and realize the Question Classification system in the Yunnan tourism, and appraise the question classifier based on the experiment.