Research On The Software Architecture And Data Analysis Of Tor-based Dark Web

With the continuous development of network communication technology,the scale of the Internet has become increasingly large.Nowadays,the Internet can be divided into two parts:the surface web and the deep web.The deep web also contains a subset called dark web,which can only be accessed through an anonymous communication system and can provide anonymity that the deep web does not have.The dark web based on Tor,an anonymous communication system,is the largest dark web entity at present,and its essence is carried in the form of Tor hidden service.Tor hidden service allows providers to hide their physical location when providing network services.This strong anonymity makes the Tor dark web flooded with unsupervised sensitive data.Therefore,it is urgent to realize the reasonable supervision of Tor dark web.Firstly,this paper analyzes and introduces the communication principles and corresponding mechanisms of the anonymous system Tor,and summarizes the current academic research work on the data analysis of Tor dark web from three aspects:Tor dark web service discovery,data collection,and data classification.Secondly,this paper studies the software architecture of the Tor dark web,and on this basis,proposes a Tor-based dark web service discovery method(TWSD).This article analyzes the design of Tor dark web services,points out the current design flaws in dark web services,and uses the flaws to design the architecture of the TWSD method.The TWSD method implements service discovery by deploying controlled hidden service directories,optimizes the problem of high consumption of service discovery resources,and implements an avoidance strategy for the dark web service review mechanism.Then,experiments verified that TWSD method only needs to deploy 16 controlled hidden service directories to achieve efficient acquisition of non-public onion domains,and can effectively avoid the review mechanism of dark web service.Finally,this paper designs the Tor-based dark web sensitive data analysis scheme(TDSA)for the difficulty of collecting and analyzing the sensitive data of Tor dark web,and elaborates the overall architecture,key module design and code implementation of the scheme in detail.In the dark web element data collection,in order to solve the problem of low collection efficiency caused by the high load of the singleton onion proxy,a load balancing strategy based on the dark web request header is proposed.In the dark web sensitive data feature extraction,in view of the limitations of the TF-ICF algorithm,an improved TF-ICF algorithm based on the location of the dark web page is proposed.Then,it was verified through experiments that the TDSA scheme can realize the automatic classification of dark web sensitive data,analyzed the relationship between different categories of sensitive data,and designed a comparative experiment to prove that the improved TF-ICF algorithm can improve the effect of dark web sensitive data classification.