Personalized Recommendation Model Of App Based On WLAN User Group

With the popularity of smart phones and tablet PCs,O2 O this new business model has improved,App has penetrated into every aspect of people's daily lives.The rapid development of the app market has brought great opportunities to the App market,but also makes the two-way choice between application developers and users become a prominent problem.Usually,the first batch of users who download an new App,on the one hand is the effect of some marketing tools from App operators,on the other hand may be the guys who are willing to try this kind of fresh app crowd.And the spread of app rely mainly on the recommendation of friends and marketing.In this paper,we propose a recommendation algorithm for app recommendation,which is based on the characteristics of each app's user group and the idea of the contact circle,which is based on WLAN,as a unit.Modeling and analysis of each app user group,based on the results of the model,we obtain the characteristic information of the user groups with different preferences,and judge whether the target users are interested in the app by matching the characteristics of the app user groups.This paper further puts forward to loyalty,activity to determine the preference of user to an app,rather than relying on the user rating.And the association rule mining algorithm Can-Mining based on the structure of Can-Tree was improved,and the improved algorithm is presented based on the incremental mining algorithm for growing app user group.In order to prove the hypothesis of this paper,a questionnaire about user app download habits has been designed and published.The results effectively support the hypothesis of this paper.Finally,Data acquisition and analysis are carried out through users' app download of a real contact circle,in order to find how deep the impact is,the contact circle to users.The app user group feature extraction algorithm is implemented and simulated.The experimental results show that the performance of the improved algorithm for app user group extraction is better than that of Can-Mining algorithm under most of the support.