《面向网络舆情任务的本体学习方法研究论文》由会员分享，可在线阅读，更多相关《面向网络舆情任务的本体学习方法研究论文（61页珍藏版）》请在金锄头文库上搜索。
1、 Nanjing University of Aeronautics and Astronautics The Graduate School College of Economics and Management Research on Internet Public Opinion-oriented Method of Ontology Learning A Thesis in Management Science and Engineering by Hu Dongrun Advised by Professor Ma Jing Submitted in Partial Fulfillm
2、ent of the Requirements for the Degree of Master of Management March, 2013 承诺书 本人郑重声明：所呈交的学位论文，是本人在导师指导下，独立进行研究工作所取得的成果。尽我所知，除文中已经注明引用的内容外，本学位论文的研究成果不包含任何他人享有著作权的内容。对本论文所涉及的研究工作做出贡献的其他个人和集体，均已在文中以明确方式标明。 本人授权南京航空航天大学可以有权保留送交论文的复印件，允许论文被查阅和借阅,可以将学位论文的全部或部分内容编入有关数据库进行检索，可以采用影印、缩印或其他复制手段保存论文。 （保密的学位论文在
3、解密后适用本承诺书） 作者签名： 日 期： 南京航空航天大学硕士学位论文 i 摘 要 随着互联网的高速发展，网络舆情的重要性越来越受到人们的重视，学术界、业界及政府部门对网络舆情进行跟踪及监测的需求日益增大。然而，互联网及信息技术的发展现状使得互联网对网络舆情缺乏语义层面的表达，导致通过互联网直接对信息进行搜检时，查全率和查准率严重不足。 为解决这个问题，本体及其相关技术被提了出来，学术界也在本体构建、本体进化、本体维护和本体评价等方面进行了深入的研究。但是，手工构建本体过程中难度很大、成本也很高这一突出问题，一直极大地限制着本体技术在对网络舆情进行监测和跟踪过程中的运用与推广。学术界提出了不
4、少本体学习的解决方案，但是对本体中概念间关系的获取能力依然不高，依旧没有很好地解决问题。 本文首先依据基于本体的网络舆情监测与跟踪系统的应用需求，结合本体的构成要素，分析了互联网新闻领域本体学习的目标；并围绕本体学习，对现有本体学习技术进行了分析与研究。之后，本文设计了面向网络舆情任务的本体学习方案，将实现的技术方案分解成流程框架与实现算法两部分分别进行讨论。在流程框架设计上，本文对网络舆情任务进行了需求分析，提出了基于自由文本的本体学习框架；在实现算法上，本文对主题词获取算法、概念获取算法以及概念间关系获取算法进行了讨论。 本文有如下特色及成果：首先，将网络舆情任务引入本体学习过程，使得所构
5、建的本体具有独特的领域性，本体学习过程也被附加上了与网络舆情相关的特点。其次，文本提出了基于本体的网络舆情跟踪与监测系统的框架，设计了本体学习过程中概念获取和概念间关系获取的算法，并在一定范围内验证了算法的有效性。 关键词：关键词：本体学习，网络舆情，概念获取 面向网络舆情任务的本体学习方法研究 ii ABSTRACT With the rapid development of the Internet, the importance of network public opinion is growing more and more attention, academia, industry
6、, and government departments for tracking and monitoring the network public opinion demand. However, the current development of the Internet and information technology makes the Internet network public opinion the lack of expression of the semantic level, resulting in, check the full rate and precis
7、ion rate is a serious shortage of information directly through the Internet search seized. To address this problem, the body and its related technologies, academia also ontology construction, ontology evolution, ontology maintenance and ontology evaluation carried out in-depth research. However, the
8、 hand-built body process difficult and costly this outstanding problems have been greatly restricted the use and promotion of the body in the process of monitoring and tracking network public opinion. The proposed solution for many ontology learning academia still not high, but the ability to obtain
9、 the relationship between the concepts in the ontology, still did not solve the problem. Firstly, based on application requirements based of ontology network public opinion monitoring and tracking system, combined with elements of the body, analysis of Internet news domain ontology learning goals; l
10、earning around the body, on existing ontology learning technology, analysis and research. The paper designs a the ontology learning program for network public opinion task, which will achieve the technical solution is decomposed into a process framework to achieve the algorithm two parts separately
11、discussion. Process framework design tasks of network public opinion needs analysis, free text-based ontology learning framework; algorithm key words acquisition algorithm, concept acquisition algorithms, the relationship between concepts acquisition algorithm discussed . This article is like featur
12、es and results: First, the introduction of network public opinion task ontology learning process, the body has built unique areas of the ontology learning process can also be attached on the network public opinion. Secondly, the text proposed design concepts in the ontology learning process to acces
13、s and obtain the relationship between the concept of the algorithm based on the framework of the bodys network of public opinion tracking and monitoring system, and verify the effectiveness of the algorithm within a certain range. Key words: Ontology Learning, Inter Public Opinion, Concept Accepting 南京航空航天大学硕士学位论文 iii 目 录 第一章 绪论 . 1 1.1选题的背景及意义. 1 1.1.1选题背景 . 1 1.1.2研究意义 . 2 1.2国内外相关研究现状 . 3 1.2.1关于本体学习的国内外研究现状. 3 1.2.2关于本体生成技术的国内外研究现状 . 5 1.2.3关于本体半自动构建方法的国内外研究现状 . 6 1.3研究内容和技术路线 .
