由于传统网络管理体系较少考虑互联网的社会性,因而对互联网的管理问题缺乏充分支持,互联网秩序正在遭受不良上网行为和不良信息的严重挑战,网络管理的主要矛盾也正在由网络基础设施管理转变为网络用户行为管理。现有的过滤策略等被动防御措施无法从根本上解决网络用户行为管理的问题。本文研究基于信用评估的网络管理方法,把金融领域的信用评估思想引入到网络用户行为管理中,实现对不良用户行为的主动调控,达到降低网络管理成本、提高网络管理水平的目的。并结合网络技术、数据挖掘等相关领域的研究成果,设计并实现了适合互联网管理特点的网络信用管理系统,可以及时获得各网络用户的信用情况,基于此信用情况的网络管理将更有针对性,加强了可操作性。
As less consideration on the social nature of the Internet, traditional network management system lacks adequate support on the management of Internet. Currently, the order of Internet is facing a serious challenge from bad behavior and bad information. The core contradiction of network management has been changed from service management to behavior management. Traditional pas- sive defense cannot solve the complex problems of management completely. This paper proposed a credit evaluation based approach for network management to actively control the bad behavior. In this way the network management costs can be reduced and the level of network management can be improved. Using the SVM distance, a credit evaluation system was built up for network management, and evaluated credit scoring for network user. In the system, the credit situation of each network user can be accessed and the management of network will be more targeted and operability based on the credit.