Research On App Network Construction And Visualization Based On Tag Recommendation

In recent years,with the popularity of mobile smart terminals,mobile application(App)have also prospered,and the number has reached millions in Google Play and Apple App Store.There are many different types of Apps that have similar functions or similar functions.Studying the relationship between Apps can make App Store better sort the App,which can help users find the App that best suits their needs faster and more accurately in the vast Apps.This paper proposes to build an App network based on tag recommendation and its visualization system,which can better explore the relationship between App and display it intuitively.First,define the App network,with the App as the node,and use the tags that can describe the App and use the relationship between the tags as the edge.After that,the App data is obtained from the application store and the App tag is extracted,and the App description information and the App comment information are merged to filter out the tags that best represent the App.Then,use Word Net to calculate the relationship between the tags,then map the relationship to the App,and then build the App network according to the definition.Finally,the visualization system of the App network is not only presented in 2D effect,but also the visualization effect is improved-it is presented in 3D effect,which facilitates the deep meaning hidden behind the data from different angles.Based on the real App data of Google Play and Apple App Store,the experimental verification of the label recommendation algorithm and the implementation of the App network and visualization system are carried out.The results show that the label recommendation algorithm proposed in this paper meets the expected effect,and the visualization effect is good.3D visualization can avoid the traditional 2D visualization brings edge and edge overlap issues,also presents App-related statistical results,which can help users search for App quickly and accurately and help developers understand similar Apps to improve their products.