Design And Implementation Of Mobile Applications Statistics And Analysis Platform

With the rapid development of mobile Internet, for different user needs more andmore mobile applications are born with transport.Subscribers have a mobile applicationsis growing, every day also produced a large number of user behavior and correspondingproduct logs, these logs user activity and can be used to analyze the same number ofuser behavior analysis,through these data we can What determines the user’s needs are.In this demand-based Internet thinking, we will get the user data and statistical analysisto improve our products, make more users to accept and use.Described in this paper is through our statistics and analysis of the mobileapplication products online log, some fragmented user behavior log, to quantify thedigital statistics, and then the digital concatenated into user behavior trajectory, throughstatistical analysis of a great deal of user behavior trajectory, will generally countedseveral clustering user behavior very similar, according to the main characteristics ofthese clustering analysis of the degree of use of each function of user product, then thefinal build these statistics and analysis of the data from one platform be demonstratedand provide the data support for the project manager in the project requirements.In this paper, according to the online real user log, use the python, JS, PHP andother language, using Linux as the development environment and running environment.The whole project and Realization of anti cheat into the design of URL, according to thelog by using cluster analysis of user behavior, user behavior data show the three parts ofthe platform. The main purpose of the design of URL anti cheat is the process In thispaper, according to the online real user log, use the python, JS, PHP language, usingLinux as the development environment and running environment. The whole project isdivided into the log processing and data display platform of two parts. The logprocessing includes the design and implementation of URL anti cheating and accordingto the log by the clustering analysis of user behavior. The main purpose of the design ofURL anti cheat is the process of filtering chain generated in the project, outside thechain of URL produced by manual annotation keywords and livejournal for comparison,will filter out the cheating URL think, equivalent to a layer of filter on the livejournal,interference analysis of the content that will weed out to analyze the online log. A largenumber of users log on, using a needle for K-means algorithm to improve the clusteringoperation, unsupervised clustering algorithm for small environment (UnsupervisedNiche Clustering), on the basis of this algorithm based on the hierarchical structure ofL-UNC(Level Unsupervised Niche Clustering) algorithm is proposed. By this algorithmanalysis line user data, the user behavior of different classified, mining user preferences, habits, to the product viscosity index, the results finally draws the user behavior. Datashow the platform part is the analysis of data and conclusions were shown in the processby showing Statistical Platform in the end and the analysis of design andimplementation of the end of the behavior of the user, user behavior data show platformis processed, the online journal, classification, storage, data according to demand for theshow.Through the design and research of this paper and the realize of the platform forthe show and the product iteration to user needs feedback is the most direct, to achievethe purpose of the study and provide data support for new products.