Mobile application daily user engagement scores and user profiles
AssigneeNuance Communications, Inc.
InventorsDan Hendrick, Eric Jun Fu
Life of this patent
Estimating the term…
Abstract
A system logs application usage data on a mobile device, processes the data on an analysis system and outputs a current and predicted score to, e.g. third parties. The system logs application-related usage data, which is collected via, e.g. a keyboard application running in the background on the mobile device. The system then evaluates the logged usage data and the events corresponding to a particular application. The events can be analyzed to score the user engagement level with the application, e.g., more events recorded for a given application per day, the more engaged a user is with that application. The engagement level can further be predicated based on historical usage log data from which a score decay model can be generated.
Claims20
A computer-implemented method for assessing a user engagement level of a particular application used on a mobile device, the method comprising:
- detecting user input received on a mobile device via an input application,
- wherein the input application is at least partially stored on the mobile device, and
- wherein the input application is configured to receive user input data for input to multiple, different applications, including the particular application;
- when the user input is detected,
- at the mobile device, associating the user input with the particular application;
- at the mobile device, logging data corresponding to user interaction with the particular application,
- wherein the user interaction includes data currently input by the user to the particular application via the input application;
- and
- at the mobile device, creating a usage log of the logged data,
- wherein the usage log includes at least an identifier for the user or the mobile device, an application identifier, and a timestamp, and
- wherein the usage log is configured to determine a user engagement level for the particular application.
A computer-implemented method for determining a user engagement level for an application stored on a mobile device, the method comprising:
- receiving, at a computing system, usage logs for one or more events detected on the mobile device,
- wherein the one or more events indicate user interactions with a keypad application stored on the mobile device,
- wherein the usage logs include an application identifier for a user-selected application,
- wherein the usage logs are associated with a user identifier, and wherein the user identifier is associated with the user or the mobile device;
- and,
- determining a user engagement level for the selected application, wherein the determining includes analyzing—
- a first time interval associated with the usage logs for the selected application;
- a usage value during the first time interval, wherein the usage value indicates when input events were recorded for the selected application in relation to the first time interval;
- a usage events value associated with the selected application, wherein usage events occur at times within the first time interval, and wherein the usage events value indicates a number of separate events recorded in association with the selected application;
- and,
- a usage count value indicating a number of characters typed, a number of characters typed per minute, a number of characters typed per day for the selected application.
A method for predicting user engagement level of an application used by a computing device, the method comprising:
- selecting an application, the application including an application identifier and being associated with a user of a mobile device on which the application is stored;
- accessing an application database to determine a last engagement level score for a selected application,
- wherein the last engagement level score is a numeric value calculated from user usage logs,
- wherein the usage logs identify one or more events detected on the computing device, and
- wherein the one or more events indicate user interactions with a keypad application stored on the computing device;
- defining a first interval of no engagement activity, the first interval reflecting a time interval during which no events associated with the application are detected in the usage logs and a current engagement level score equals to the last engagement level score,
- wherein the first interval is within a predetermined range of an averaged time distribution associated with the application, and
- wherein the averaged time distribution is calculated across multiple users having at least one detected event corresponding to the application;
- defining a second interval of no engagement activity, the second interval reflecting a time interval during which no events associated with the application are detected in the usage logs and a current engagement level score declines from the last engagement level score,
- wherein the second interval occurs subsequent to the first interval, and
- wherein the second interval is outside the predetermined range of the averaged time distribution associated with the application;
- and
- modeling the application engagement level score with the first interval and the second interval to predict the application engagement level score when no events associated with the application are detected in the usage logs.
Dependencies are inferred from the claim text.