Study on the Question Processing of Farmer-oriented Question Answering System
面向农民的问答系统问句处理研究*

In order to improve the convenience of farmers’ access to information, this paper is focused on developing the farmer-oriented question answering system and proposes that the system is composed of four modules: knowledge base construction, question processing, information retrieval, answer extraction, in which question processing is the research priority. On the basis of concluding the characteristics of farmers’ questions, it proposes the question classification method based on interrogatives and phrases. During the question processing, methods are adopted to classify the questions and extract the keywords effectively, such as removing polite words, establishing special rule table for informal interrogatives and no interrogative and so on. While, the authors take advantage of the synonym extended table to expand keywords, and set different weight benchmarks. The research can lay the foundation for the processing of the information retrieval module.