The Browser User Behavior Recognition Technology Research

With the rapid development of network technology,networks are increasingly close to hu man life.People can obtain the latest information they wanted anytime and anywhere by using phone or computer,including purchasing their favorite items through the network.Network has been integrated into human life.At the same time,in order to provide users with more intellige nt and attentive network service,Web application developers and ad networks began to focus on the behavior of Internet users.Under the era of big data,advertise networks through a person's browsing habits and access records on the Internet,thus infer that the next person needs.In this way users not only access to resources they wanted conveniently,but also makes the network advertising precisely thereby reducing the running costs.Consequently,how to identify and track the user,which will become the main problem to be solved.In this paper,the research on this issue,through research and comparison of currently available technology of track user behavior,a new browser user identification technology-browser fingerprint recognition technology was founded.With its characteristics of unique means of identification?difficult to detect and remove,it has being widely admired.The technology uses the users' information of device configuration,operating system type,browser configuration as the characteristic parameter of the users' identification,with these characteristic parameters can be obtained from other user's browser fingerprints.The fingerprint will replace the browser cookies and other technologies to identify the user,in order to achieve the purpose track the user's lasting.The main work is as follows:Firstly,research status of the current browser user behavior tracking technology are introduced in detail.Secondly,the macro on the previously tracking technology are introduced and compared in detail;introduces the related content of the browser fingerprint recognition technology,including the characteristic parameters of the collection,the application of information entropy to describe the fingerprint recognition algorithm,HTML5 Canvas fingerprint technology and on the basis of the introduction.Again,for lack of a browser fingerprint technology,it proposes two improvements,improved training and selection of parameters characteristic fingerprint algorithm with the new threshold parameter,and through experiments on the effect of the improved algorithm is tested.Finally,the browser fingerprint recognition technology to practical application,it will be used on the hospital website registration system,to identify the identity of scalpers,ensure rational allocation of hospital resources registered.