The Similarity Calculation Based Automatic Question Answering System In The Body

With the Internet's development, on the one hand people can access a great deal of information through the network of resources, on the other hand although there are many internet search engines can help people search for information they want, the vast web of information, the speed of growth and renewal is faster and faster. To find the desired content is a challenging task from the mass of information. The current search technology has significant limitations. It only provides keyword-based search, while ignoring the semantic content contained keyword itself. It can not meet the individual needs of the user and can not meet people easily, quickly and accurately to get information. In this case, the traditional keyword-based search engine can not meet user's requirements. People look forward to new generation of search engines that can be intelligent processing and the results are more concise representation. Driven by the demand in this article combine Automatic Q & Processing to Information Retrieval in order to enhance the existing search engine to deal of information of Intelligent processing. Then Provide more Humane Human-computer interaction and use the concise, accurate answer, to auto-answer user's questions in natural language to help people quickly and effectively find the information.Statements similarity calculation is a key technology of automatic question answering system. This article will introduce ontology into the system, use Ontology Technology to build domain ontology. domain ontology provides Complete description about concept and the relationship's description between concepts and concept in a particular field. Therefore, we look on the domain ontology as the basic resources,let more comprehensive domain knowledge and rich semantic relations apply to question answering systems, in order to help us resolve the current lack of understanding of the problem of semantic in automatic question answering system.This paper builds on existing algorithms and proposes ontology-based semantic similarity calculation method. Similarity algorithm between the concept is in the main body through the various properties between different concepts which are compared to judge their individual properties on the same concept or not to get the semantic similarity. When we get similarity between the concept,we start to design vector model of the questions. We use vector space method to get the similarity between questions. Thus answering system allows users to be more intelligent and efficient information you want.Our main research works in this article is the following:1. We study the key technology automatic Q & A system and introduce the function module of automatic Q & A system.2. We study technology about ontology and describe concepts and ontology modeling language and modeling methods.3. We analysis the existing words and sentence similarity's calculation method, and point that its shortcoming in the field of automatic question answering system. Ontology is introduced into automatic Q & A system and propose a semantic similarity calculation method based on ontology automatic question answering system. We combinate ontology-based similarity with semantic similarity based on HowNet to service statements similarity calculation. We build concept semantic similarity calculation method based on ontology. Then we test algorithm's validity and correctness by the results.At last we use it in practical application. we design a simple ontology model based on automatic Q & A system through building a medicine of "rhinitis" ontology model, After we compared it with traditional information retrieval,we provide a more simple, convenient and personalized way of human-computer exchange.