AUTOMATED APPLICATION ANALYTICS
An application analytics generation system employs data stored in a social networking system's social graph to automatically predict outcomes based on various observed attributes, and transmit these predictions to a social networking system's user interface, e.g., in a newsfeed, for an advertiser or application developer. These predictions correlate various attributes and/or actions to outcomes, e.g., in relation to advertisements or applications.
1 . A method performed by a computing device having one or more processors and memories, comprising:
receiving a collection of data, wherein the received collection of data includes at least a portion of a social graph, the social graph comprising a data structure representing entities, relationships between the entities, attributes associated with the entities, and events corresponding to actions performed on or by the entities;
receiving, via an application program interface, a request for correlation data, the correlation data correlating the attributes and the actions with outcomes, wherein the outcomes are selected other attributes or actions;
dividing a set of attributes and events not selected as outcomes into a set of dimensions;
grouping the dimensions into two or more groups of dimensions;
correlating the groups of dimensions with the outcomes;
selecting at least one statistically significant correlation between one of the groups of dimensions and the outcomes; and
causing to be displayed in a user interface the selected at least one correlation.
2 . The method of claim 1 , wherein the selecting includes performing a t-test between the two or more groups of dimensions to select the at least one statistically significant correlation.
3 . The method of claim 1 , further comprising receiving from the user interface an indication of whether the displayed at least one correlation is useful.
4 . The method of claim 3 , further comprising using a machine learning algorithm to improve subsequent selections of selected correlations.
5 . The method of claim 1 , wherein the displaying is in a newsfeed of a social network.
6 . The method of claim 5 , further comprising:
selecting multiple correlations;
prioritizing the selected multiple correlations; and
displaying the selected multiple correlations in order of priority.
7 . The method of claim 1 , further comprising receiving via the user interface a selection of a portion of the user interface, the selected portion defining a set of entities having common attributes and, in response, displaying via the user interface a set of additional entities also having the common attributes.
8 . The method of claim 1 , further comprising sorting the groups of dimensions.
9 . The method of claim 1 , wherein the attribute can be age, gender, country, operating system type, operating system version, screen size, memory capacity, or type of data communications network employed.
10 . The method of claim 1 , wherein an attribute can be one or more actions performed on or by an entity.
11 . The method of claim 1 , wherein outcome can be selecting an advertisement, purchasing a product, installing an application, uninstalling an application, or engaging with a feature of an application.
12 . A computer-readable storage device storing instructions, the instructions comprising:
instructions for receiving a collection of data, wherein the received collection of data includes at least a portion of a social graph, the social graph comprising a data structure representing entities, relationships between the entities, attributes associated with the entities, and events corresponding to actions performed on or by the entities;
instructions for receiving, via an application program interface, a request for correlation data, the correlation data correlating the attributes and the actions with outcomes, wherein the outcomes are selected other attributes or actions;
instructions for dividing a set of attributes and events not selected as outcomes into a set of dimensions;
instructions for grouping the dimensions into two or more groups of dimensions;
instructions for correlating the groups of dimensions with the outcomes;
instructions for selecting at least one statistically significant correlation between one of the groups of dimensions and the outcomes; and
instructions for causing to be displayed in a user interface the selected at least one correlation.
13 . The computer-readable storage device of claim 12 , wherein the instructions for selecting include instructions for performing a t-test between the two or more groups of dimensions to select the at least one statistically significant correlation.
14 . The computer-readable storage device of claim 12 , further comprising instructions for receiving from the user interface an indication of whether the displayed at least one correlation is useful.
15 . The computer-readable storage device of claim 14 , further comprising instructions for using a machine learning algorithm to improve subsequent selections of selected correlations.
16 . The computer-readable storage device of claim 16 , further comprising:
instructions for selecting multiple correlations;
instructions for prioritizing the selected multiple correlations; and
instructions for displaying the selected multiple correlations in order of priority.
17 . The computer-readable storage device of claim 12 , further comprising instructions for receiving via the user interface a selection of a portion of the user interface, the selected portion defining a set of entities having common attributes and, in response, displaying via the user interface a set of additional entities also having the common attributes.
18 . The computer-readable storage device of claim 12 , wherein the attribute can be age, gender, country, operating system type, operating system version, screen size, memory capacity, or type of data communications network employed.
19 . The computer-readable storage device of claim 12 , wherein outcome can be selecting an advertisement, purchasing a product, installing an application, uninstalling an application, or engaging with a feature of an application.
20 . A system having one or more processors and memories, comprising:
a component configured to receive a collection of data, wherein the received collection of data includes at least a portion of a social graph, the social graph comprising a data structure representing entities, relationships between the entities, attributes associated with the entities, and events corresponding to actions performed on or by the entities;
a component configured to receive, via an application program interface, a request for correlation data, the correlation data correlating the attributes and the actions with outcomes, wherein the outcomes are selected other attributes or actions;
a component configured to divide a set of attributes and events not selected as outcomes into a set of dimensions;
a component configured to group the dimensions into two or more groups of dimensions;
a component configured to correlate the groups of dimensions with the outcomes;
a component configured to select at least one statistically significant correlation between one of the groups of dimensions and the outcomes; and
a component configured to cause to be displayed in a user interface the selected at least one correlation.