基于语义web涉鄂网络舆情分析系统.doc

本科毕业论文 题目： 基于语义web的涉鄂网络舆情分析系统 学 院: 计算机科学与技术 专 业: 日 期: 二〇一一年六月  摘 要 随着网络的普及和网民数量的快速增长，网络舆情已经成为映射社会舆情的实时晴雨表，最直接、最快速地反映了经济、政治、文化的状况和发展态势。网络舆情分析平台成为了解社情民意的有力工具之一。然而，监管机构往往更关注某个领域、层次的舆情情况，从基于关键词的搜索结果中的海量信息中找到与特定领域的相关的局部信息是非常困难的，另一方面，由于互联网的开放性、共享性、扁平性使得网络舆情的出现和监管不再具有地域性，不再具有边界，使关键词在正确表达搜索期望时总显得力不从心，容易产生语义漂移现象。 本文提出了一个基于语义web的网络舆情分析系统，系统主要包括网络舆情信息采集，舆情智能分析和处理，信息的分词索引，舆情简报和舆情检索，以及本体知识构建等功能，通过建立领域本体，采集网络舆情信息，并利用本体对舆情信息进行分析和聚类，实现舆情的垂直化、专业化分析，解决了舆情分析中的语义漂移问题。 该系统以湖北省相关领域为样本，重点采集了湖北省地域比较活跃的论坛上舆情信息，对湖北省舆情信息进行监测分析，发现湖北地域比较热门的事件，并将基于关键词的搜索结果与基于语义的搜索结果进行对比，发现基于语义的搜索结果更加准确可靠，验证了系统模型的有效性。 关键词: 语义Web；本体；网络舆情；Web信息挖掘  Abstract With the pervasion of Internet and the sharp increase of the number of netizens , Network public opinion has been serving as an effective method in the monitoring and reflection of social public opinion . However, Regulatory Authorities are inclined to concern more about public opinion in a special field while it is too difficult to acquisite the expected information that relates to a special area from the massive searching results based on key words, on the other hand , the characteristics of openness ,sharability and flatness of Internet deprive the network public opinion of its localization and boundaries to a great extent and lead to the fact that it is powerless for the key words to express the searching desire since its easiness of generating semantic drift phenomenon. This article introduced a semantic Web - based Network public opinion analysis system which includes many functions,such as Network public opinion gathering, Information intelligence analysis and processing , Information participle and index, Public opinion briefing, Public opinion retrieval and Ontology knowledge construction etc.Also this system realizes the vertical search of public opinion ,and the analysis of specialization ,besides ,it completely solve the semantic drift phenomenon problem inpublic opinion analysis by establishing domain ontology . T