Detection of the App in Google Play by Categorization

A current challenge within the quickly evolving app market scheme is to take care of the integrity of app classes. At the time of registration, app developers need to choose, what they believe, is that the most acceptable class for his or her apps. Besides the inherent ambiguity of choosing the correct class, the approach leaves open the likelihood of misuse and potential recreation by the registrant. Sporadically the app store can refine the list of classes offered and doubtless designate the apps. However, it's been observed that the couple between the outline of the app and therefore the class it belongs to continues to persist. Though some common mechanisms (e.g. a complaint-driven or manual checking) exist, they limit the latent period to discover miscategorized apps and still open the challenge on categorization. We tend to introduce FRAC+: Framework for App Categorization. FRAC+ has the following salient features: (i) it's supported a data-driven topic model and mechanically suggests the classes acceptable for the app store, and (ii) it will discover miscategorizated apps. In depth experiments attest to the performance of FRAC+. Experiments on GOOGLE Play shows that FRAC+’s topics are a lot of aligned with GOOGLE’s new classes and zero.35%-1.10% game apps are detected to be miscategorized.