Client Fingerprint Collecting System Design And Implementation

In the internet plus period,thousands of new platforms come out every day.People has more and more chance to choice.Clients can use a site on different kind of devices,such as personal PC,Pad,mobile phone and so on.To improve the used experience of client users,each enterprise hopes that they can trace and recode the customers operations and personal habits.When user login in by different kind of devices,how to recognize these users,and set a relationship between the user and the devices,it's really a challenge.To recognize users' devices and set a relationship between users and the devices,this article proposes to collect the attributes of browser and system as device fingerprint.This article may introduce how to collect client user's Html Canvas,browser attributes,system fonts etc.through Javascript.This article proposes to get client user's devices fingerprint through analysis the attributes of devices.Such as entropy of each attributes,uniqueness properties of each attributes,the changes properties of each attributes.To get the client user's devices fingerprint and do data analysis,this article designs and developes a client device fingerprint collecting system.The Web site collects users' device fingerprint and send to server.The server part then sends the Client fingerprint to HDFS,where we can analysis the huge data throw big data algorithm.Finally,this article analysis the data,calculates each device attribute's entropy,uniqueness rate and change properties.After balance the three values,finally get the attributes that most suitable for using as device fingerprint collection and calculates the most numbers of devices this system can distinct out.