Browse Prior Art Database

A system and Method for Recommending Mobile App Installs

IP.com Disclosure Number: IPCOM000245815D
Publication Date: 2016-Apr-12
Document File: 2 page(s) / 40K

Publishing Venue

The IP.com Prior Art Database

Related People

Mitesh Patel: INVENTOR [+2]

Abstract

Disclosed is a method and system for identifying users who are likely to install a given mobile app. The method and system estimates a user's propensity to install a mobile app as a product of the user's propensity to install and the user's interest in the specific mobile app. The method models the user's propensity to install as Poisson process and estimates an install rate from mobile analytics platform. Further, the method models the user's interest in the mobile apps as a multinomial process over all mobile app categories. Thereafter, the method smooths the estimates using a Dirichlet prior. Finally, an overall propensity of a user to install a specific mobile app is defined as a product of the propensity to install and the interest in the mobile app's category.

This text was extracted from a Microsoft Word document.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 54% of the total text.

A system and Method for Recommending Mobile App Installs

Abstract

Disclosed is a method and system for identifying users who are likely to install a given mobile app.  The method and system estimates a user’s propensity to install a mobile app as a product of the user’s propensity to install and the user’s interest in the specific mobile app.  The method models the user’s propensity to install as Poisson process and estimates an install rate from mobile analytics platform.  Further, the method models the user’s interest in the mobile apps as a multinomial process over all mobile app categories.  Thereafter, the method smooths the estimates using a Dirichlet prior.  Finally, an overall propensity of a user to install a specific mobile app is defined as a product of the propensity to install and the interest in the mobile app’s category.

Description

A method and system is disclosed for identifying users who are likely to install a given mobile app. 

In accordance with the method and system, mobile app installation is modeled as a two phase process.  A user decides to install a mobile app according to a Poisson process with rate .  Once the user is installing, the user chooses which mobile app to install according to a multinomial distribution .  In theory, this multinomial can be over all mobile apps, however, in practice, due to high dimension, the method models this as multinomial over all mobile app categories.  The method further smooths the estimates...