受限领域问答系统的研究与设计

发布时间：2018-02-23 02:22

本文关键词： 问答系统 FAQ 问题理解相似度计算　出处：《内蒙古大学》2012年硕士论文　论文类型：学位论文

【摘要】：随着互联网的发展和应用,网上的信息迅速增长。人们希望能从海量的网络内容获取自己所需要的信息。搜索引擎的出现从很大程度上解决了这个问题。人们只需输入一些关键字,搜索引擎就会返回相关的网页。但是面对繁多的网页信息,用户很难迅速找到自己所需的内容。因此,为了满足人们能够更快速、准确地获取信息的愿望,自动问答系统(automatic Question Answering System, QA)逐渐发展起来。自动问答系统允许用户使用自然语言进行提问,并针对问题返回一个简洁准确的答案。它综合运用多种自然语言处理技术,是计算机应用领域研究的热点之一。目前,英文问答系统的研究已比较成熟,由于中文自然语言的复杂性,因此中文问答系统的研究还处于初步阶段。本文研究的是受限领域内的中文自动问答系统。本文根据计算机领域知识的特点,研究设计了一个针对计算机网络课程基于常问问题库(FAQ)的中文问答系统。本系统主要研究了领域知识库的构建,问题理解,计算句子相似度算法等方面的内容。在构建领域知识库部分,研究设计了课程知识点表结构、FAQ存储方式、对FAQ进行预处理；问题理解部分,主要研究了中文分词、关键词提取和扩展、问题分类方法等；句子相似度计算部分,采用了基于语义的相似度计算方法。并建立了相应的问题测试集进行试验,文章最后介绍了整个自动问答系统的实验结果及其评价。
[Abstract]:With the development and application of the Internet, Information on the Internet is growing rapidly. People want to get the information they need from a huge amount of online content. The emergence of search engines solves this problem to a large extent. People just need to enter some keywords. The search engine will return the relevant web pages. However, in the face of so many web pages, it is difficult for users to find the content they need quickly. Therefore, in order to satisfy people's desire to obtain information more quickly and accurately, Automatic Question Answering system (QA) is developing gradually. The automatic question answering system allows users to use natural language to ask questions and return a concise and accurate answer to the question. It synthetically uses a variety of natural language processing techniques and is one of the hotspots in the field of computer application. Due to the complexity of Chinese natural language, the study of Chinese question and answer system is still in its preliminary stage. According to the characteristics of computer domain knowledge, this paper studies and designs a Chinese question answering system for computer network courses based on FAQ. This system mainly studies the construction of domain knowledge base and the understanding of problems. In the part of constructing domain knowledge base, we design the structure of course knowledge point table and pre-process FAQ. In the part of problem understanding, we mainly study the Chinese word segmentation, and the main contents of this paper are as follows: (1) in the part of constructing domain knowledge base, we design the structure of course knowledge point table and preprocess FAQ. In the part of sentence similarity calculation, the semantic similarity calculation method is used, and the corresponding question test set is established for the experiment. Finally, the paper introduces the experimental results and evaluation of the whole automatic question and answer system.
【学位授予单位】：内蒙古大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：TP391.1

【引证文献】