Automated generation of predictive insights classifying user activity
Non-limiting examples of the present disclosure relate to application of artificial intelligence (AI) processing to generate classifications of user activity for a group of users. For example, a classification prediction is generated indicating whether students in an educational class are predicted, over a predetermined time period, to have a high or low activity level based on contextual analysis of multiple types of user-driven events. As user activity data is typically quite robust, the present disclosure applies dimensionality reduction processing to efficiently manage user activity data and further improve accuracy in generating downstream binary classifications. A dimensionality reduction transformation of user activity data results in a low-dimensional representation of input feature data that is contextually relevant for generating a binary classification. Derived classifications are then utilized to generate data insights pertaining to user activity levels of one or more users. Data insights are can then be rendered for presentation.
1 . A computer-implemented method to help guide learning by a plurality of student users comprising:
by one or more computer devices connected to a network:
storing user activity data of an educational software platform in one or more data storage devices connected to the network, wherein the user activity data comprises event count data indicating user activity, by each student user of the plurality of student users, with each of a plurality of applications of the educational software platform;
accessing, via the network, the user activity data on the one or more data storage devices;
executing one or more trained artificial intelligence (AI) models within a computing environment connected to the network to generate a classification of a level of user activity for each student user of the plurality of student users based on analysis of a dimensionality reduction transformation of the user activity data, wherein the one or more trained AI models, in generating the classification of user activity for each student user of the plurality of student users, executes processing operations that comprise:
identifying input feature data for the one or more trained AI models, wherein the input feature data comprises features indicating the event count data for each of the plurality of student users;
normalizing the input feature data based on an educational frame of reference designation to create normalized input feature data;
executing dimensionality reduction processing on the normalized input feature data to generate the dimensionality reduction transformation of the user activity data, wherein executing the dimensionality reduction processing comprises executing principal component analysis (PCA) on the normalized input feature data to generate one or more principal components and projecting the normalized input feature data in a PCA space based on the one or more principal components;
applying a null activity vector to the dimensionality reduction transformation of the user activity data in the PCA space to determine a frame of reference for evaluating a projection of the normalized input feature data in the PCA space; and
generating the classification of the level of user activity for each student user of the plurality of student users based on analysis of the dimensionality reduction transformation of the user activity data;
generating at least one data insight for visual presentation in a graphical user interface (GUI) of an application of the educational software platform, wherein the data insight indicates a comparative level of user activity in the educational software platform by at least one student user of the plurality of student users during a current temporal frame of reference as compared to a previous temporal frame of reference; and
causing the at least one data insight to be displayed in the GUI of the application of the educational software platform.
2 . The computer-implemented method of claim 1 , wherein the plurality of student users are students in an educational class, and wherein the educational frame of reference designation is the educational class.
3 . The computer-implemented method of claim 2 , wherein the one or more trained AI models generate an activity score for each student user of the plurality of student users based on an application of a classification algorithm to the dimensionality reduction transformation of the user activity data, and wherein the classification of the level of user activity for each student user of the plurality of student users is generated based on an analysis of a comparative evaluation of activity scores for the plurality of student users.
4 . The computer-implemented method of claim 3 , wherein the comparative evaluation comprises:
executing a threshold evaluation of the activity score, for each of the plurality of student users;
classifying one or more of the plurality of student users as low activity based on a result of the threshold evaluation; and
classifying one or more of the plurality of student users as high activity based on a result of the threshold evaluation.
5 . The computer-implemented method of claim 3 , wherein the classification algorithm generates the classification using a classifier trained as a logistic regression model.
6 . The computer-implemented method of claim 3 , wherein the one or more trained AI processing models are further adapted to generate labels for the classification of the level of user activity for each student user of the plurality of student users based on analysis of the projection in the PCA space using a trained machine learning model.
7 . The computer-implemented method of claim 1 , further comprising, prior to generating the at least one data insight, determining a change in the level of user activity in the educational software platform based on the classification of the level of user activity for each student user of the plurality of student users.
8 . The computer-implemented method of claim 1 , wherein generating the classification of the level of user activity for each student user of the plurality of student users comprises applying a classification algorithm to generate the classification from analysis of the projection of the normalized input feature data according to the dimensionality reduction transformation of the user activity data.
9 . The computer-implemented method of claim 1 , wherein the user activity data further comprises categorical representations of user feedback received from student surveys that are provided through the educational software platform, wherein the normalizing of the input feature data further comprises converting the categorical representations of user feedback to numerical representations.
10 . A system comprising:
at least one processor; and
a memory, operatively connected with the at least one processor, storing computer-executable instructions to help guide learning by a plurality of student users, wherein the executable instructions, when executed by the at least one processor, causes the at least one processor to execute a method that comprises:
storing user activity data of an educational software platform in one or more data storage devices connected to a network, wherein the user activity data comprises event count data indicating user activity, by each student user of the plurality of student users, with each of a plurality of applications of the educational software platform;
accessing, via the network, the user activity data on the one or more data storage devices;
executing one or more trained artificial intelligence (AI) models within a computing environment connected to the network to generate a classification of a level of user activity for each student user of the plurality of student users based on analysis of a dimensionality reduction transformation of the user activity data, wherein the one or more trained AI models, in generating the classification of user activity for each student user of the plurality of student users, executes processing operations that comprise:
identifying input feature data for the one or more trained AI models, wherein the input feature data comprises features indicating the event count data for each of the plurality of student users;
normalizing the input feature data based on an educational frame of reference designation to create normalized input feature data;
executing dimensionality reduction processing on the normalized input feature data to generate the dimensionality reduction transformation of the user activity data, wherein executing the dimensionality reduction processing comprises executing principal component analysis (PCA) on the normalized input feature data to generate one or more principal components and projecting the normalized input feature data in a PCA space based on the one or more principal components;
applying a null activity vector to the dimensionality reduction transformation of the user activity in the PCA space to determine a frame of reference for evaluating a projection of the normalized input feature data in the PCA space; and
generating the classification of the level of user activity for each student user of the plurality of student users based on analysis of the dimensionality reduction transformation of the user activity data;
generating at least one data insight for visual presentation in a graphical user interface (GUI) of an application of the educational software platform, wherein the data insight indicates, a comparative level of user activity in the educational software platform by at least one student user of the plurality of student users during a current temporal frame of reference as compared to a previous temporal frame of reference; and
causing the at least one data insight to be displayed in the GUI of the application of the educational software platform.
11 . The system of claim 10 , wherein the plurality of student users are students in an educational class, and wherein the educational frame of reference designation is the educational class.
12 . The system of claim 11 , wherein the one or more trained AI models generate an activity score for each student user of the plurality of student users based on an application of a classification algorithm to the dimensionality reduction transformation of the user activity data, and wherein the classification of the level of user activity for each student user of the plurality of student users is generated based on an analysis of a threshold evaluation of activity scores for the plurality of student users.
13 . The system of claim 12 , wherein the threshold evaluation comprises: classifying, from analysis of the activity scores for the plurality of student users relative to one or more set thresholds, one or more student users of the plurality of student users as low activity and one or more student users of the plurality of student users as high activity.
14 . The system of claim 13 , wherein the method, executed by the at least one processor, further comprises, prior to generating the at least one data insight, determining a change in the level of user activity in the educational software platform based on the classification of the level of user activity for each student user of the plurality of student users.
15 . The system of claim 10 , wherein the user activity data further comprises categorical representations of user feedback received from student surveys that are provided through the educational software platform, and wherein the normalizing of the input feature data further comprises converting the categorical representations of user feedback to numerical representations.
16 . One or more non-transitory computer-readable storage media having program instructions stored thereon to help guide learning by a plurality of student users, wherein the program instructions, when executed by a computing system, direct the computing system to perform operations, the operations comprising:
by one or more devices connected to a network:
storing user activity data of an educational software platform in one or more data storage devices connected to the network, wherein the user activity data comprises event count data indicating user activity, by each student user of the plurality of student users, with each of a plurality of applications of the educational software platform;
accessing, via the network, the user activity data on the one or more data storage devices;
executing one or more trained artificial intelligence (AI) models within a computing environment connected to the network to generate a classification of a level of user activity for each student user of the plurality of student users based on analysis of a dimensionality reduction transformation of the user activity data, wherein the one or more trained AI models, in generating the classification of user activity for each student user of the plurality of student users, executes processing operations that comprise:
identifying input feature data for the one or more trained AI models, wherein the input feature data comprises features indicating the event count data for each of the plurality of student users;
normalizing the input feature data based on an educational frame of reference designation to create normalized input feature data;
executing dimensionality reduction processing on the normalized input feature data to generate the dimensionality reduction transformation of the user activity data, wherein executing the dimensionality reduction processing comprises executing principal component analysis (PCA) on the normalized input feature data to generate one or more principal components and projecting the normalized input feature data in a PCA space based on the one or more principal components;
applying a null activity vector to the dimensionality reduction transformation of the user activity data in the PCA space to determine a frame of reference for evaluating a projection of the normalized input feature data in the PCA space;
generating the classification of the level of user activity for each student user of the plurality of student users based on analysis of the dimensionality reduction transformation of the user activity data; and
outputting the classification of the level of user activity for one or more student users of the plurality of student users; and
generating at least one data insight for visual presentation in a graphical user interface (GUI) of an application of the educational software platform, wherein the data insight indicates a comparative level of user activity in the educational software platform by at least one student user of the plurality of student users during a current temporal frame of reference as compared to a previous temporal frame of reference; and
causing the at least one data insight to be displayed in the GUI of the application of the education software platform.
17 . The one or more non-transitory computer-readable storage media of claim 16 , wherein the one or more trained AI models generate an activity score for each student user of the plurality of student users based on an application of a classification algorithm to the dimensionality reduction transformation of the user activity data, and wherein the classification of the level of user activity for each student user of the plurality of student users is generated based on an analysis of a threshold evaluation of activity scores for the plurality of student users.
18 . The one or more non-transitory computer-readable storage media of claim 16 , the operations further comprising, prior to generating the at least one comparative data insight, determining the number of student users of the plurality of student users that are less active during the current temporal frame of reference than during the previous temporal frame of reference.