【作者】 潘晔；

【导师】 任广伟；

【作者基本信息】 贵州大学 ， 计算机应用技术， 2007， 硕士

【摘要】 随着网络和分布式应用的发展和普及，大量的异构数据源应运而生。特别是当前Web技术的飞速发展，人们获取信息的途径大量来自于网络。而Web是一个巨大的信息集散地，又同时也伴随着一个巨大的问题——信息泛滥、混杂。它给Web上信息的有效获取及利用带来一定的困难。如何使用户快速有效的获取和利用Web上的大量信息，这是目前国际上许多机构和组织正在研究的焦点问题，也是Web信息管理的关键技术之一。为此，本文引入本体概念来实现异构Web数据集成，本体的使用有利于解决数据集成中的语义异构问题，这是其它数据集成技术考虑较少的方面。本文在介绍了基本的语义网的原理及相关标准后，重点研究了数据的获取和集成。数据获取的过程分为三个阶段：定义抽取规则，获取网页及清洗页面，格式转换。首先在用户参与下，根据样本页面学习获得抽取规则，之后下载页面并利用Tidy工具清洗，最后由系统完成数据的抽取和转换。文中介绍了基本的基于DOM树的方法，以及改进的基于语义块的方式。数据集成的过程包括：构建领域本体，本体与抽取数据库的映射，查询处理。本体是语义集成的基础，文中阐述了领域本体的构建方式，建模工具，数据库实例的语义标注及查询处理过程。最后对论文工作进行了总结和展望，提出了此方面尚需研究和改进的问题。更多还原

【Abstract】 Along with the development of network and distribute application and universality, a great deal of difference Heterogeneous data emerge with the tide of the times.Especially with current Web technique rapid development, people obtain the a great deal of information come from the Web. The Web is not only a huge information commercial center, but also in the meantime a huge problem-information deluge, congestion. It brings a lot of difficults in obtaintion of Web top information and use.How to obtain a great deal of information from web and use it effectively, this is a hotspot which many top organizations and associations are researching, and it is also one of the key techniques in Web information management.So this paper introduce Ontology concept to realize difference Heterogeneous Web data integration, Usage of Ontology has many advantages to solve a data integration medium of semantic Heterogeneous problem, this is other data integration technique be very difficult realization. This paper puts the main foucs on obtaining data and data integration after introducting of basic principle of the Semantic Web and related standard.The process that the data obtain is divided into three stage: The definition sample rule, obtain a web page and clean a page, format conversion. Firstly, with the users’ participation, the process studies the rules according to the sample page. Secondly, the process downloads pages and makes use of Tidy tool to clean web page. Finally, system completes sample of data and converses the data. This paper introduces the basic method bases on DOM tree, and improvement of the method bases on semantic block.The process of data integration include: Setting up a realm Ontology, mapping Ontology and database, search processing.The Ontology is foundation of semantic integration, The dissertation demonstrates a set of realm Ontology ways, model tools, the database example semantic mark and search processing process.The end of dissertation refers to the summary of this paper, outlook, and puts forward the problem in this aspect that need to be researched and improved constantly.更多还原