Targeted generation of messages for digital therapeutics using generative transformer models
Aspects of the present disclosure are directed to systems, methods, and computer readable media for generating messages targeted to address conditions in users. A computing system may identify (i) a condition of a user to be addressed and (ii) one or more parameters defining messages to be presented via an application towards achieving an endpoint associated with the condition for the user. The computing system may generate a prompt using the condition and the one or more parameters in accordance with a template. The computing system may apply the prompt to a generative model to output a message identifying at least one activity toward achieving the endpoint. The generative model may be trained using a corpus. The computing system may provide, the message for presentation to prompt the user to perform the activity via the application towards achieving the endpoint to address the condition.
1 . A method, comprising:
identifying, by one or more processors, (i) a condition of a user to be addressed and (ii) one or more parameters defining content to be presented via an application;
generating, by the one or more processors, an input based on the condition and the one or more parameters, in accordance with a template;
providing, by the one or more processors, the input to a generative model, wherein the generative model is trained by:
providing a set of inputs to the generative model to generate a set of outputs; and
updating, using back propagation or an optimization function, the generative model based at least on the set of inputs and the set of outputs;
generating, by the one or more processors, based on providing the input to the generative model, content associated with addressing the condition of the user, wherein the content is generated based at least on a probability of engagement with the content by the user and comprises (i) a lesson comprising one or more explanations, instructions, or education materials for the user to view or read, the lesson selected for addressing the condition of the user, or (ii) one or more interactive elements configured to receive an input from the user indicating performance of an activity for addressing the condition of the user, the input provided via a user interface of the application; and
providing, by the one or more processors, for presentation via the application, the content associated with addressing the condition of the user.
2 . The method of claim 1 , further comprising:
receiving, by the one or more processors, a response identifying interaction by a user via the application in response to the presentation of the content;
determining, by the one or more processors, based on the response, one or more second parameters to define content to be presented for the user;
generating, by the one or more processors, a second input based on the one or more second parameters, in accordance with the template;
providing, by the one or more processors, the second input to the generative model to generate second content associated with addressing the condition; and
providing, by the one or more processors, for presentation via the application, the second content associated with addressing the condition.
3 . The method of claim 1 , further comprising:
receiving, by the one or more processors, a response identifying interaction by a user via the application in response to the presentation of the content; and
updating, by the one or more processors, generation of content by the generative model using the response.
4 . The method of claim 1 , further comprising selecting, by the one or more processors, the template from a plurality of templates based on the condition to be addressed, the template including at least one placeholder to include at least one of the one or more parameters.
5 . The method of claim 1 , further comprising determining, by the one or more processors, a time at which to present the content to a user via the application, based on information associated with the user of the application.
6 . The method of claim 1 , wherein identifying the one or more parameters further comprises identifying, by the one or more processors, information on a user, to which content are to be presented, to use as at least one of the one or more parameters.
7 . The method of claim 1 , wherein identifying the one or more parameters further comprises selecting, by the one or more processors, a rule from a plurality of rules based on information on a user associated with the condition, to use as at least one of the one or more parameters.
8 . The method of claim 1 , wherein identifying the one or more parameters further comprises selecting, by the one or more processors, a plurality of candidate content based on an endpoint to be achieved towards addressing the condition, to use as at least one of the one or more parameters.
9 . The method of claim 1 , wherein generating the set of outputs further comprises selecting, by the one or more processors, at least one token to include in the set of outputs based on a probability of occurrence generated by the generative model for the at least one token.
10 . The method of claim 1 , wherein the generative model is established using a set of rules defining selection from a plurality of candidate content based on at least one of: (i) user behavior, (ii) user preference, or (iii) user profile information.
11 . The method of claim 1 , wherein the generative model is established using a plurality of corpora, at least a portion of the plurality of corpora comprising information associated with the condition to be addressed.
12 . The method of claim 1 , wherein a user is on a medication to address the condition, in partial concurrence with a session in which the content is provided, and wherein the content includes at least one of a short messaging service (SMS) message, a multimedia messaging service (MMS) message, or an in-app message.
13 . A system, comprising:
one or more processors coupled with memory, configured to:
identify (i) a condition in a user to be addressed and (ii) one or more parameters defining content to be presented via an application;
generate an input based on the condition and the one or more parameters, in accordance with a template;
provide the input to a generative model, wherein the generative model is trained by:
providing a set of inputs to the generative model to generate a set of outputs; and
updating, using back propagation or an optimization function, the generative model based at least on the set of inputs and the set of outputs;
generate, based on providing the input to the generative model, content associated with addressing the condition of a user, wherein the content is generated based at least on a probability of engagement with the content by the user and comprises (i) a lesson comprising one or more explanations, instructions, or education materials for the user to view or read, the lesson selected for addressing the condition of the user, or (ii) one or more interactive elements configured to receive an input from the user indicating performance of an activity for addressing the condition of the user, the input provided via a user interface of the application; and
provide, for presentation via the application, the content associated with addressing the condition of the user.
14 . The system of claim 13 , wherein the one or more processors are further configured to:
receive a response identifying interaction by a user with the application in response to the presentation of the content;
determine based on the response, one or more second parameters to define content to be presented for the user;
generate a second input based on the one or more second parameters, in accordance with the template;
provide the second input to the generative model to generate second content associated with addressing the condition; and
provide, for presentation via the application, the second content associated with addressing the condition.
15 . The system of claim 13 , wherein the one or more processors are further configured to:
receive a response identifying interaction by a user with the application in response to the presentation of the content; and
update generation of content by the generative model using the response.
16 . The system of claim 13 , wherein the one or more processors are further configured to select the template from a plurality of templates based on the condition to be addressed, the template including at least one placeholder to include at least one of the one or more parameters.
17 . The system of claim 13 , wherein the one or more processors are further configured to determine a time at which to present the content to a user via the application, based on information associated with the user of the application.
18 . The system of claim 13 , wherein the one or more processors are further configured to identify information on a user, to which content are to be presented, to use as at least one of the one or more parameters.
19 . The system of claim 13 , wherein the one or more processors are further configured to select a rule from a plurality of rules based on information on a user associated with the condition, to use as at least one of the one or more parameters.
20 . The system of claim 13 , wherein the one or more processors are further configured to select a plurality of candidate content based on an endpoint to be achieved towards addressing the condition, to use as at least one of the one or more parameters.
21 . The system of claim 13 , wherein the one or more processors are further configured to select at least one token to include in the set of outputs based on a probability of occurrence generated by the generative model for the at least one token.
22 . The system of claim 13 , wherein the generative model is established using a set of rules defining selection from a plurality of candidate content based on at least one of: (i) user behavior, (ii) user preference, or (iii) user profile information.
23 . The system of claim 13 , wherein the generative model is established using a plurality of corpora, at least a portion of the plurality of corpora comprising information associated with the condition to be addressed.
24 . The system of claim 13 , wherein a user is on a medication to address the condition, in partial concurrence with a session in which the content is provided, and wherein the content includes at least one of a short messaging service (SMS) message, a multimedia messaging service (MMS) message, or an in-app message.