【作者】 陶玉香；

【导师】 高中文；

【作者基本信息】 哈尔滨理工大学 ， 模式识别与智能系统， 2008， 硕士

【摘要】 随着Internet技术的发展,网络资源也迅猛增长。如何使Internet用户快速有效地获得所需的资源,已成为网站设计者亟待解决的问题。基于中国国家企业信息网的决策支持系统,将数据挖掘技术应用到Web服务器日志的挖掘,即通过挖掘服务器中的日志文件,获得用户的访问模式,从而进一步分析和研究用户的访问规律,来改进网站的组织结构和服务。本文将理论与实际应用相结合,开发了基于中国国家企业信息网的决策支持系统,并在以下几个方面做了深入的研究:1.分析了Web日志挖掘的流程以及Web数据的收集,详细分析了Web日志预处理的四个步骤:数据净化、用户识别、会话识别和路径补充,给出了具体实现流程,并在传统数据预处理的基础上加入时态信息,提出了基于时态数据库的Web日志预处理的流程。2.具体分析了Apriori算法,并对Apriori算法做了改进,经实验证明了其正确性,并且节约了时间,提高了算法的执行效率。3.结合中国国家企业信息网的日志及其拓扑结构,开发设计了基于中国国家企业信息网的决策支持系统。本系统是采用MVC模式,用基于Java和SqlServer技术实现的。系统主要做了两方面的内容:一是对网站的访问信息的统计,如日访问量等;二是利用Apriori算法对其进行数据挖掘,来发现隐藏的用户访问的路径和规则。基于中国国家企业信息网的决策支持系统在改进网站结构和网站性能方面,为网站的设计者和决策者提供了依据。如果进一步研究就可以实现用户浏览行为预测,以及能自动控制的自适应性网站。更多还原

【Abstract】 Along with the development of the Internet technical,network resources also grow fast.So how to make the Internet customers acquire the resources which are needed effectively and quickly,become a very important problem for the website designers to resolve.The decision support system based on Chinese national enterprise information network,using the data mining technique in the web log mining, mining the log of the server to acquire the access mode of the customer, analyzing and studying the regulation of the log to improve the organization structure and the service of the website.The thesis combine the theories and physically applied together,developed the decision support system based on Chinese national enterprise information network,and do deep research follwing:1.Analyzed the mining processing of the web log and the web log collection,then analyzed the four steps in the preparing of the web log processing:data clean,user identification,session identification and path supple,also give the process flowing,a research of data preprocessing in web log based on temporal database.2.Analyzed the Apriori algorithm,and put forward a new algorithm of apriori.From the experiment we know it is right,also the algorithm can economized time and raised the efficiency.3.Using the topology structure and and web log of the Chinese national enterprise information network,develop the decision support system based on Chinese national enterprise information network.The system based on the MVC model,realize by using the Java and SqlServer technique.The system mainly do two things:statistic the access information,such as daliy accessing;to mining the website use the apriori algorithm,to discover the hidden path and rule which the user visited.The decision support system based on Chinese national enterprise information network can provide basis to the web decision maker and the web designer when improving the website structure and the website function.If further study , can carry out the self-adaptability of the website and the technique of the prediction of user browsing behavior.更多还原