Research Of Key Techniques Of Intelligent Agricultural Information Service System

At present, agricultural intelligent information technology is a researching hotspot in the agricultural informationism field. As the Internet have reached into every corner of social life, become an important source of information for all kinds of people. Farmers also need a convenient, fast way of information retrieval to help themselves to find the necessary information. On the one hand, hope to raise information request query use the natural language way directly; On the other hand, do not want to fall into the complex information ocean, query information can be accurately and directly answer their requests. It aims to provide accurate solution timely for users and to solve the problems of excessive feedback information from search engine and of knowledge bottleneck in information system, To carry out a detailed and in-depth research for its key technologies involved in agricultural intelligent information service system.The questions classification was as an important component of information service system, provided central informations and solutions to let the system to know "What users are looking for". The accurate question classification results have a good guide for the follow-up processes, improved the performance of the system greatly. And that Ontology could put up clear, formal description for agricultural domain knowledge from a high level of semantic and expand question's semantic information expediently. Therefore, this thesis based the ideology and conception model of ontology, preliminary constructed an agricultural domain ontology knowledge base, used the ontology_based questions classification method, combined with ontological reasoning mechanism, did an in-depth analysis and semantic expansion of the question, realized the classification the question in agricultural information domain.The control of question state is essential for the whole question answering process control. If the question's status is confused, is likely to delay the answer for the question or lose the control of the question. This thesis studied the state control technology in workflow system, cleared the state-controlled methods and design ways. Through analyzing the process of the question answering, determined the system's response incidents and the states in the system, designed the question status switch chart, given the formalization expression for the question status control model, and then used the object-oriented approach implemented question state control module, realized the control of the question answering process, enhanced the system's reusability and robustness.This thesis given the framework of a web-based agricultural intelligent information service system, preliminary established a non-real-time exchange platform between users and experts, for the technical services intercommunion between users and experts.