Font Size: a A A

Research And Implementation Of Key Technologies Of User Behavior Recognition Based On Depth Packet Detection

Posted on:2018-04-28Degree:MasterType:Thesis
Country:ChinaCandidate:Y WangFull Text:PDF
GTID:2348330536479862Subject:Electronic and communication engineering
Abstract/Summary:PDF Full Text Request
With the rapid development of the Internet,there are more than 700 million people using the network in China.All kinds of applications emerge in an endless stream,mainly including information inquiry,online communication,e-commerce,online games,e-mail,audio and video playback.The vast amounts of data and information generated by these applications are a very important strategic resource that hides great social,economic,scientific,cultural and political values.It has important application value that how to legally and efficiently use these data to help users quickly find their own information from the complex data through the user identity and user network behavior recognition research.This dissertation analyzes the main methods of user behavior identification.For people using the network by mobile phones,pad and other mobile devices in the public,these methods can not meet the commercial or public occasions wifi scenarios for user behavior analysis and web push requirements.First,this dissertation carries out the overall design of the behavior analysis system on the intelligent router.The data acquisition method and the application feature extraction method used in the system are analyzed in detail,the user behavior table and push the information table and other database design are completed.Second,the user behavior recognition system is designed in detail,and the network protocol is designed and implemented.The user behavior recognition system is designed in detail,and the network protocol is designed and implemented.The DPI technology is used to identify and record the user's Internet behavior,and the depth analysis work of the application layer data is completed with the AC multi-mode matching algorithm.Using the crawler and TF-IDF model to complete web page keyword extraction,the flow chart,code language and data structure of the main modules of the system are designed.Finally,a complete system test environment is built,combined with hardware equipment on the system functions for a large number of field tests,the successful identification of user behavior and web push.Combined with hardware equipment on the system function for a large number of tests,the system successfully identifies user behavior and makes web push.This dissertation studies and realizes the user's Internet behavior recognition system.After the function test,it shows that the system is running normally,and the resource consumption of the intelligent router is small,and it does not affect the operation of other systems.The real-time and the success rate are higher,indicating that the system meets the customer's requirements for the user behavior identification system for intelligent routers.
Keywords/Search Tags:deep packet inspection, user behavior recognition, web push, term frequency–inverse document frequency, multi-mode matching algorithm
PDF Full Text Request
Related items