Abstract: Systems and techniques are provided for the mining signals from developer clusters in an application ecosystem and the creation of rules to proactively ban applications when they are submitted to the application ecosystem. Applications may be received. Each application may be associated with a developer account that submitted the application to an application ecosystem. Signals may be extracted for one of the applications from one of the applications and the developer accounts. A banned prevalence percentage may be determined for the signal. A signal rule may be generated from the signal based on the banned prevalence percentage for the signal. The signal rule may be stored.

Abstract: The subject matter of this specification can be embodied in, among other things, a method that includes receiving, at a first portable program module in a first web-accessible container document at a client device, context information from a second portable program module in a second web-accessible container document at the client device. The method further includes identifying advertisement information for presentation to a user at the client device, wherein the advertisement information is based on the context information. The method further includes presenting the advertisement information by the first portable program module to the user.

Abstract: Applications that have been designed for a smaller format device such as a smartphone and simply ported to a larger format device such as a tablet can be discerned from applications designed specifically for the larger format device. An application can be evaluated based on tablet compatibility criteria and can be assigned a tablet compatibility score. The application can be evaluated based on quality criteria and can be assigned a quality score. The compatibility score and the quality score can be used to help find and rate any number of applications.

Abstract: Systems and techniques are provided for the classification of risks associated with developers in an application ecosystem. Signals associated with a developer account for an application ecosystem may be received. Each of the signals may include one of an account signal, an application signal, and a financial signal, and each of the signals may be associated with a weight and a score. The signals may be combined using the weights and the scores associated with the signals to obtain a risk probability for the developer account. The risk probability may be scored.

Abstract: Systems and techniques are provided for the mining signals from developer clusters in an application ecosystem and the creation of rules to proactively ban applications when they are submitted to the application ecosystem. Applications may be received. Each application may be associated with a developer account that submitted the application to an application ecosystem. Signals may be extracted for one of the applications from one of the applications and the developer accounts. A banned prevalence percentage may be determined for the signal. A signal rule may be generated from the signal based on the banned prevalence percentage for the signal. The signal rule may be stored.

Abstract: Applications that have been designed for a smaller format device such as a smartphone and simply ported to a larger format device such as a tablet can be discerned from applications designed specifically for the larger format device. An application can be evaluated based on tablet compatibility criteria and can be assigned a tablet compatibility score. The application can be evaluated based on quality criteria and can be assigned a quality score. The compatibility score and the quality score can be used to help find and rate any number of applications.