Techniques for expanding a target audience for messaging
This disclosure relates to systems and methods that include configuring a machine learning system to train on a plurality of messages transmitted to target groups of an online social networking service, determining a threshold differential and a weight value using responses to the plurality of messages, and send the input message to the target in response to a differential between the expected number of positive responses and the weight multiplied by the expected number of negative responses being greater than the threshold differential.
1. A system comprising:
at least one processor;
a machine-readable storage medium having instructions stored thereon, which, when executed by the at least one processor, cause the system to:
configure a machine learning system to train on a plurality of messages transmitted to a portion of members for a first target group of an online social networking service, the machine learning system, once trained, to output an expected number of positive responses and an expected number of negative responses based on an input message being selected for transmission to a group of the online social networking service;
identify a plurality of target groups from a set of target groups for respective input messages;
for each of the plurality of target groups, determine a threshold differential and a weight value using expected responses from each respective target group, the threshold differential and weight value for use in minimizing a number of messages to send while satisfying a constraint requirement, the constraint requirement comprising a maximum number of negative responses and a minimum number of positive responses; and
send the respective input message to each target group in the plurality of target groups in response to the threshold differential for the respective group being exceeded by a differential between the expected number of positive responses and a first quantity, the first quantity being the weight multiplied by the expected number of negative responses.
2. The system of claim 1 , wherein each target group is selected from target groups consisting of a language group, a regional group, a cultural group, and a membership group at a network system.
3. The system of claim 1 , wherein the positive responses include at least one of a response that involves a page view, a clicked link, a purchase, a like, and a comment, and the negative responses include at least one of a response that involves an unsubscribe action, a complaint, a dislike, and a spam report.
4. The system of claim 1 , wherein the machine learning system is configured to train on responses for a predetermined threshold period of time starting with a first response from a user.
5. The system of claim 1 , wherein determining the threshold differential and the weight value uses one of multi-objective optimization and a grid search.
6. A method comprising:
configure a machine learning system to train on a plurality of messages transmitted to a portion of members for a first target group of an online social networking service, the machine learning system, once trained, to output an expected number of positive responses and an expected number of negative responses based on an input message being selected for transmission to a of the online social networking service;
identify a plurality of target groups from a set of target groups for respective input messages;
for each of the plurality of target groups, determining a threshold differential and a weight value using expected responses from each respective target group, the threshold differential and weight value for use in minimizing a number of messages to send while satisfying a constraint requirement, the constraint requirement comprising a maximum number of negative responses and a minimum number of positive responses; and
sending the respective input message to each target group in the plurality of target groups in response to the threshold differential for the respective group being exceeded by a differential between the expected number of positive responses and a first quantity, the first quantity being the weight multiplied by the expected number of negative responses.
7. The method of claim 6 , wherein the plurality of messages on which the machine learning system is configured to train are messages selected from a recent period of time.
8. The method of claim 6 , wherein each target group is selected from target groups consisting of a language group, a regional group, a cultural group, and a membership group at a network system.
9. The method of claim 6 , wherein the positive responses include at least one of a response that involves a page view, a clicked link, a purchase, a like, and a comment, and the negative responses include at least one of a response that involves an unsubscribe action, a complaint, a dislike, and a spam report.
10. The method of claim 6 , wherein the machine learning system is configured to train on responses for a predetermined threshold period of time starting with a first response from a user.
11. The method of claim 6 , wherein determining the threshold differential and the weight value uses one of multi-objective optimization and a grid search.
12. A tangible machine-readable storage medium having instructions stored thereon, which, when executed by a processor, cause the processor to perform:
configuring a machine learning system to train on a plurality of messages transmitted to a portion of members for a first target group of an online social networking service, the machine learning system, once trained, to output an expected number of positive responses and an expected number of negative responses based on an input message being selected for transmission to a group of the online social networking service;
identify a plurality of target groups from a set of target groups for respective input messages;
for each of the plurality of target groups, determining a threshold differential and a weight value using expected responses from each respective target group, the threshold differential and the weight value for use in minimizing a number of messages to send while satisfying a constraint requirement, the constraint requirement comprising a maximum number of negative responses and a minimum number of positive responses; and
sending the respective input message to each target group in the plurality of target groups in response to the threshold differential for each respective group being exceeded by a differential between the expected number of positive responses and a first quantity, the first quantity being the weight multiplied by the expected number of negative responses.
13. The machine-readable storage medium of claim 12 , wherein determining the threshold differential and the weight value uses one of multi-objective optimization and a grid search.
14. The machine-readable storage medium of claim 12 , wherein the machine learning system is configured to train on responses for a predetermined threshold period of time starting with a first response from a user.