The Design And Implementation Of The Real-time Analytic Platform Of Storm-based Customer's Behavior

Nowadays,with the development of IT product manufacturing enterprise informatization,Enterprise website are widely applied in the world,simultaneously the website are facing various problem such as the functionality getting complex,system becoming slower and slower.People start complaining it is hard to operate the system.As the enterprise website owner,in one hand,it is difficult for them to find the root cause so that they can hardly provide the solution for the relative problems,on the other hand,enterprise website owner can not realize the need of their owner,they cannot raise one optimization of the functionality of the enterprise website,these are commonly problems which enterprise website owner faced.This paper started with the current situation about enterprise website application,it analyzed the mainly problem about the enterprise website management and design the platform of the real-time Storm-based user behavior analysis.The platform not only monitoring those enterprise website,catching user behavior data and know about the user activity base on the data,but by analyzing these activity data real-time to display the various chart for explaining what the problem or user behavior regular is,enterprise website will be managed better after this platform.This paper mainly consists of the following sections:The architecture and design of the storm-based user behavior real-time analysis platform.Firstly,the paragraph analysis the mainly problem of enterprise website in daily operation,it sum up the needs of functionality and non-functionality for the commonly enterprise website and its owner.And it raise the solution which providing an analysis platform to analyze and resolve the problem and providing the optimization solution based on the user data.The analysis and design of the configurable rules of user behavior.Base on the needs of functionality and non-functionality,it design the models of the rule of user behavior,these models is able to be configured dynamically in the different sense.Also,it design the store model of user behavior and the conversion between models and data.The analysis and design of the store model of user behaviorAccording to the character of user behavior data from the enterprise website,paper design the store model of user behavior,it provide the reference about the analysis result for the enterprise website owner.These data models can adapt the different sense according to the different data of user behavior,so only if providing the rule of user behavior definition and enterprise website owner can retrieve different user behavior chart at different view corner,it imply the compatibility.This paper takes the user product configuration guide function in the management system of enterprise configuration as an example to verify the functionality of the Storm-based analysis platform of user behavior,it consist of the infrastructure of the analysis platform,grab the data of user behavior in configure system,making the rule for the user behavior and the verification of the analysis result of user behavior.As a result,it proved that this storm-based analysis platform of user behavior can help enterprise website owner find the problem easily and can optimize it.Meanwhile,through comparing the difference for the usage of memory on the platform between before and after,it indicated that there is little consumption for the system performance.This paper takes advantage of the real-time character to demonstrate the real feedback for enterprise website,and it provide the real chart display the obvious difference and bottleneck on the target enterprise website.It is helpful for the enterprise website owner to improve the satisfaction on their customers.