Dark Web Domain Name Collection And Content Analvsis

Dark web is a network space on the Internet that is difficult to access and retrieve through public channels.While protecting the privacy of users,the dark web has also become a bed of criminal activities such as guns,drugs and credit card transactions.Therefore,how to comprehensively collect the dark web resources,analyze its network organization structure,and classify the content based on the degree of harmfulness has practical urgency and important application value for protecting cyberspace security.The domain name of the dark web is not publicly released.It has a short survival time or is often changed.It is highly dynamic,and there is almost no link between the dark web and the open web.As a result,the domain name of the dark web is difficult to find.Many features of dark web is different from what of the open web which limit the applicability of standard technologies and increase the difficulty of studying the structure and content distribution of dark space.Based on the above problems,this paper analyzes and studies the domain name collection,web structure and content harmfulness of three anonymous networks of Tor,I2P and ZeroNet.The main contributions include:(1)For the problem that the dark domain name is difficult to find,based on the way that domain name addresses of Tor anonymous networks are collected by searching keywords in Open Web,a method for discovering more search keywords based on Tor2web software project is proposed.Based on this method,sixteen new search keywords have been found on the basis of the existing ones..Starting from the existing domain name collection methods of Tor and I2P anonymous networks,according to the working principle and operation mechanism of ZeroNet,four methods of domain names collection about ZeroNet anonymous network were proposed.19,561 unique ZeroNet domain names were collected in total.(2)Aiming at how to effectively analyze the dark web structure,a method of constructing complex network graph based on hyperlinks between websites is proposed.By analyzing the complex network structure of dark web,It is found that the dark web has the characteristics of loose network structure and excessive number of isolated nodes and its complex network based on hyperlink structure has scale-free characteristics and small world characteristics,but does not have hierarchical module characteristics;By using network attacks method according to its scale-free characteristics,the importance of the nodes is evaluated,and the point-centered indicators are selected as the basis for ranking the importance of the website.(3)Aiming at the problem of how to define the illegality of websites content,a method of websites classification based on the degree of harm of the website is proposed.The main idea is to mark the degree of harm of illegal websites according to the relevant legal provisions in the sub-commentary of criminal law.According to the correlation analysis of the harm degree of illegal websites,the importance,influence and popularity of websites,classified websites are divided into three levels:serious harmful,harmful and influential,harmful.Then,according to whether there are links to illegal websites,other websites are divided into potential harmful and harmless levels;In this stage,PageRank algorithm is improved by using the number of domain names collected according to the way of publishing and collecting secret domain names and the behavior habits of secret network users,which improves the popularity of links on homepages.