基于个性化知识搜索的E-learning系统的研究

发布时间：2018-06-01 16:14

本文选题：E-learning系统 + Nutch　；参考：《湖北工业大学》2012年硕士论文

【摘要】：随着网络化技术进入高校的趋势愈演愈烈，社会对网络化的教学越来越重视，这种教学要求俨然成为各高校信息化建设的核心价值体现，E-learning系统便在此背景下应运而生。其不仅为学习者提供丰富的学习资源，也提供了不分时间和空间的灵活的学习方式。然而，目前大多数E-learning系统只提供人机交互的平台，缺乏有效的互助行为。而且不提供知识检索的功能，当学习者希望通过搜索获得具体信息时，便需要转至全文搜索引擎,而这类搜索引擎为学习者提供无差别的、“千人一面”的搜索结果，从而使学习者失去学习兴趣。据于此，个性化知识搜索成为在线学习系统开发者追求的目标。个性化知识搜索在传统搜索方式的基础上建立基于本体的用户兴趣模型，该模型能够将用户的教育背景、兴趣偏好和搜索目的等信息考虑其中，并重视信息间的语义联系，对索引进行优化，对提交结果进行重新排序，实现个性化搜索服务。本文旨在实现一个基于个性化知识搜索的E-learning系统，所做的工作如下：（1）搭建实验环境。这个工作包括两个部分：一是建立E-learning平台。参考开源E-learning系统框架结构根据需求定制属于自己的E-learning系统，结果显示建立的系统在功能上符合在线学习平台的要求，在效果上令人满意。二是建立Nutch搜索引擎。配置环境变量，，在Eclipse里调试代码，给定初始网址测试实验数据，并将爬虫抓取好的文件部署到Tomcat服务器上，通过浏览器即可以得到Nutch搜索引擎查询入口。（2）从两个方面对Nutch搜索引擎进行优化。一是Nutch中文分词优化，针对Nutch搜索引擎只是单字切分，对中文不敏感的问题，提出使用Imdict智能词典的中文分词模块imdict-chinese-analyzer，同时从三个方面对新的中文分词模块进行功能上的扩展，实验表明扩展后的分词模块在分词工作上更精确。（3）具体讨论了领域本体的构建方法和用户模型的建模流程，在其基础上建立了基于计算机课程本体的用户兴趣模型。由于整个教学资源范围过大，构建本体比较复杂，本文只对其中计算机课程部分进行本体构建，最后基于该本体和提取用户兴趣特征词及其权值计算来建立基于本体的用户兴趣模型。本体可以用来描述信息间的语义关系，基于本体的用户兴趣模型能更好地反映用户的查询偏好。（4）实现了基于个性化搜索的E-learning系统。将通过如中文分词优化、排序优化和基于计算机课程本体的用户兴趣模型优化后的Nutch源代码用Ant工具重新打包后，放入Eclipse调试，并将Nutch搜索引擎嵌入到原先搭建好的E-learning系统上，完成对基于个性化搜索的E-learning系统的搭建。
[Abstract]:With the increasing trend of network technology into colleges and universities, the society pays more and more attention to the network teaching. This teaching requirement has become the core value of the information construction of colleges and universities, and E-learning system emerges as the times require under this background. It not only provides learners with abundant learning resources, but also provides flexible learning methods regardless of time and space. However, at present, most E-learning systems only provide a platform for human-computer interaction, and lack of effective mutual aid behavior. And without the function of knowledge retrieval, when learners want to get specific information through search, they need to go to full-text search engines, which provide learners with undifferentiated, "one-to-one" search results. So that learners lose interest in learning. According to this, personalized knowledge search has become the goal of online learning system developers. On the basis of traditional search methods, personalized knowledge search establishes user interest model based on ontology. This model can take user's educational background, interest preference and search purpose into account, and attach importance to the semantic relation between information. The index is optimized, the submitted results are reordered, and the personalized search service is realized. The purpose of this paper is to implement a E-learning system based on personalized knowledge search. The work of this paper is as follows: 1) setting up the experimental environment. This work includes two parts: one is the establishment of E-learning platform. According to the framework of open source E-learning system, the E-learning system is customized according to the requirements. The results show that the system meets the requirements of online learning platform in function and the effect is satisfactory. Second, the establishment of Nutch search engine. Configure the environment variable, debug the code in Eclipse, give the experimental data of initial URL test, and deploy the crawler grab files to the Tomcat server, and get the Nutch search engine query entry through the browser. 2) optimize Nutch search engine from two aspects. The first is Nutch Chinese word segmentation optimization. Aiming at the problem that Nutch search engine is only single word segmentation and is insensitive to Chinese, the Chinese word segmentation module imdict-chinese-analyzer, which uses Imdict intelligent dictionary, is put forward, and the new Chinese word segmentation module is extended from three aspects. Experiments show that the extended segmentation module is more accurate in word segmentation. 3) the construction method of domain ontology and the modeling flow of user model are discussed in detail. Based on it, user interest model based on computer course ontology is established. Because the whole teaching resources range is too large and the construction of ontology is more complicated, this paper only carries on ontology construction to the part of computer course. Finally, the user interest model based on the ontology is established based on the ontology and the extraction of user interest feature words and their weights. Ontology can be used to describe the semantic relationship between information, and the user interest model based on ontology can better reflect the query preference of users. The E-learning system based on personalized search is implemented. The Nutch source code, which is optimized by Chinese word segmentation, sorting optimization and user interest model based on computer course ontology, is repackaged with Ant tools and put into Eclipse debugging, and the Nutch search engine is embedded into the previously built E-learning system. The E-learning system based on personalized search is built.
【学位授予单位】：湖北工业大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：TP391.6

【参考文献】