基于本体与问句模板的教学内容的主动搜索

发布时间：2018-05-11 13:23

本文选题：本体 + 本体知识库　；参考：《广西师范大学》2013年硕士论文

【摘要】：目前计算机信息技术的快速发展,给人们生活和生产都带来了很大的便利,其中由于网络的快速发展和普及,学生获取知识的途径也不再局限于传统的书本和课堂,虽然通过网络的搜索可以增加学生自主学习的能力和兴趣,但传统的教学资源搜索引擎主要是基于关键字、词、短语的简单匹配,这种方法忽略了资源中的数据和数据之间的各种语义关系,致使这类搜索系统的查准率和查全率都很低,它的结果会返回给学生很多无关的网页,学生若要想获取所需的信息资源只能做进一步的筛选处理,并且还有诸如交互性差、智能性差等一系列的不足。在教学领域中,人们正期盼推出能理解学生输入语义、高效检索信息资源的搜索系统。因此,本论文提出的教学资源搜索引擎以本体为核心,构建出了一个对教学领域中的概念层次、概念关系、概念的属性进行形式化描述的领域本体库,本搜索系统的各个模块充分利用了本体的语义性,在论文的问句处理模块中,采用了句型模板的问句处理方法来处理学生输入的问句。另一方面,本体知识库的构建使得系统在查找学习资源时的查全率和查准率得到了提高。本文从以下几个方面进行研究： (1)教学资源领域本体的构建。通过给出教学领域本体中重要的概念词、概念之间各种的关系、概念的对应属性,可以形式化的描述教学领域,并给出了教学资源本体的部分实现和数据结构领域本体的详细实现,最后对构建的本体给出了相应的检测。 (2)学生输入问句的查询处理。本论文允许学生以自然语言的方式进行输入查询,在本系统的前台获得学生的输入查询请求后,通过问句处理方法中的分词、词性标注、去除停用语和客套语后,通过计算了输入问句与句型模版的相似度,将预处理后的相关部分与所设定的句型模板进行匹配,若与所有的句型模板匹配失败,则转向本论文的词库匹配,从而在最大程度上提高了学生输入问句的匹配率,使系统能更好的理解学生的输入查询请求。 (3)教学资源本体知识库的构建。本论文中采用动态的方式构建本体知识库,其构建过程主要由Web网页抓取模块和Web信息语义处理这2个模块组成,通过定时器利用HTML Parser每隔一定的时间对所设定的相关URL网页进行抓取,并结合传统的向量模型将所抓取的网页内容转换成本体知识库中的实例。 (4)教学内容的主动搜索系统的设计与实现。系统主要分为三个部分：问句处理、资源检索与相关答案的返回。其中问句处理用来处理学生的输入查询,资源检索是对本体知识库中的资源进行检索,答案的返回是将与查询相关的URL网址、网页标题等相关信息按权重的大小以降序的形式返回给用户。最后给出了系统的性能分析,用实验结果证明了本论文方法的可行性与优越性。在论文的最后部分,对已经完成的工作做出了总结,并且还对需要进行改进的工作做了阐述。
[Abstract]:At present , the rapid development of computer information technology brings great convenience to people ' s life and production .

Therefore , the teaching resource search engine proposed in this paper takes the ontology as the core , constructs a domain ontology library for formal description of the concept hierarchy , the concept relation and the concept of the teaching field , and the various modules of the search system fully utilize the semantic of the ontology . In the question sentence processing module of the thesis , the question sentence processing method of the sentence pattern template is adopted to deal with the question sentences input by the students .

( 1 ) The construction of the ontology in the field of teaching resources . By giving the important concept words in the field of teaching , the relations between the concepts and the corresponding attributes of the concept can be formalized description of the teaching field , and the partial realization of the teaching resource body and the detailed implementation of the ontology in the data structure field are given , and finally the corresponding detection is given to the constructed ontology .

( 2 ) The inquiry process of the student ' s input question sentence . This thesis allows students to input query in natural language . After obtaining the students ' input query request in front desk of the system , the similarity of input question sentence and sentence pattern is calculated . By calculating the similarity between the input question sentence and the sentence template , the relevant part after preprocessing is matched with the set period template , so that the matching rate of the student ' s input question sentence can be improved to a maximum extent , so that the system can better understand the student ' s input query request .

( 3 ) The construction of the knowledge base of the teaching resource ontology . In this paper , the ontology knowledge base is constructed in a dynamic way . The construction process is mainly composed of two modules of Web page grabbing module and Web information semantic processing . The timer is used to grab the set related URL web page at regular intervals by using the timer , and the retrieved web page content is converted into an instance in the cost volume knowledge base in combination with the traditional vector model .

( 4 ) The design and implementation of the active search system of the teaching content . The system is divided into three parts : question sentence processing , resource retrieval and return of the relevant answers . The question sentence processing is used to process the input query of the student . The retrieval of resources is the retrieval of resources in the ontology knowledge base . The return of the answer is to return the relevant information related to the query to the user in descending order according to the size of the weight . Finally , the performance analysis of the system is given , and the feasibility and superiority of the method are proved by the experimental results .

In the last part of the paper , the work done has been summarized , and the work to be improved is also described .

【学位授予单位】：广西师范大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：TP391.1

【参考文献】