UTILIZING MACHINE LEARNING MODELS TO GENERATE PREDICTED REFEREE INTERACTION METRICS FOR GENERATING AND TRANSMITTING DIGITAL NOTIFICATIONS ACROSS COMPUTER NETWORKS TO REFERRER CLIENT DEVICES
The present disclosure relates to systems, non-transitory computer-readable media, and methods for utilizing machine learning to generate predicted referee interaction metrics for building a digital notification distribution policy for tiers of referrer client devices and transmitting digital notifications to referrer client devices across computer networks. In particular, in one or more embodiments, the disclosed systems utilize a referee interaction prediction machine learning model that generate predicted referee interaction metrics indicating likelihoods of downstream interactions of referee client devices based on features of referrer client devices. The disclosed systems generate referrer client device tiers for referrer client devices based on the predicted referee interaction metrics and then utilizes an optimization model to generate a digital notification distribution policy for the tiers of the referrer client devices. Further, the disclosed systems transmit digital notifications to referrer client devices in accordance with the digital notification policy and the referrer client device tiers.
1 . A computer-implemented method comprising:
generating, utilizing a trained referee interaction prediction machine learning model, a plurality of predicted referee interaction metrics from a plurality of referrer client device features;
generating a plurality of referrer client device tiers based on the plurality of predicted referee interaction metrics;
generating, utilizing an optimization model, a digital notification distribution policy for the plurality of referrer client device tiers from a set of measured referee client device interactions for the plurality of referrer client device tiers and a target interaction metric; and
transmitting digital notifications to referrer client devices in accordance with the digital notification distribution policy and the plurality of referrer client device tiers.
2 . The computer-implemented method of claim 1 , wherein generating the plurality of predicted referee interaction metrics comprises:
generating, utilizing the trained referee interaction prediction machine learning model, a first plurality of predicted referee interaction metrics; and
generating, utilizing an additional trained referee interaction prediction machine learning model, a second plurality of predicted referee interaction metrics.
3 . The computer-implemented method of claim 2 , further comprising generating the plurality of predicted referee interaction metrics by combining the first plurality of predicted referee interaction metrics from the trained referee interaction prediction machine learning model and the second plurality of predicted referee interaction metrics from the additional trained referee interaction prediction machine learning model.
4 . The computer-implemented method of claim 1 , wherein generating the plurality of referrer client device tiers based on the plurality of predicted referee interaction metrics comprises:
generating a ranked order of the plurality of predicted referee interaction metrics; and
identifying tier thresholds based on the ranked order of the plurality of predicted referee interaction metrics.
5 . The computer-implemented method of claim 1 , further comprising:
generating a referrer client device tier database by aggregating measured referee client device interactions from the set of measured referee client device interactions according to the plurality of referrer client device tiers and referrer invitation values transmitted to historical referrer client devices corresponding to the measured referee client device interactions; and
generating, utilizing the optimization model, referrer invitation values corresponding to the plurality of referrer client device tiers based on the referrer client device tier database.
6 . The computer-implemented method of claim 1 , wherein transmitting the digital notifications comprises:
extracting client device features for a candidate referrer client device; and
generating, utilizing the trained referee interaction prediction machine learning model, a predicted referee interaction metric for the candidate referrer client device from the client device features.
7 . The computer-implemented method of claim 6 , wherein transmitting the digital notifications comprises:
selecting a referrer client device tier for the candidate referrer client device from the plurality of referrer client device tiers based on the predicted referee interaction metric for the candidate referrer client device; and
transmitting to the candidate referrer client device a digital notification comprising a referrer invitation value according to the digital notification distribution policy and the referrer client device tier.
8 . The computer-implemented method of claim 1 , further comprising generating the set of measured referee client device interactions for the plurality of referrer client device tiers by:
providing a set of digital notifications comprising referrer invitation values to a plurality of test referrer client devices; and
measuring interactions of referee client devices associated with test referrer client devices of the plurality of test referrer client devices.
9 . The computer-implemented method of claim 1 , further comprising training the trained referee interaction prediction machine learning model utilizing historical referrer client device features and a set of training referee client device interactions.
10 . A non-transitory computer-readable medium storing instructions that, when executed by at least one processor, cause a computer system to:
generate, utilizing a trained referee interaction prediction machine learning model, a plurality of predicted referee interaction metrics from a plurality of referrer client device features;
generate a plurality of referrer client device tiers based on the plurality of predicted referee interaction metrics;
generate, utilizing an optimization model, a digital notification distribution policy for the plurality of referrer client device tiers from a set of measured referee client device interactions for the plurality of referrer client device tiers and a target interaction metric; and
transmit digital notifications to referrer client devices in accordance with the digital notification distribution policy and the plurality of referrer client device tiers.
11 . The non-transitory computer-readable medium of claim 10 , further comprising instructions that, when executed by the at least one processor, cause the computer system to generate the plurality of predicted referee interaction metrics by:
generating, utilizing the trained referee interaction prediction machine learning model, a first plurality of predicted referee interaction metrics; and
generating, utilizing an additional trained referee interaction prediction machine learning model, a second plurality of predicted referee interaction metrics.
12 . The non-transitory computer-readable medium of claim 11 , further comprising instructions that, when executed by the at least one processor, cause the computer system to generate the plurality of predicted referee interaction metrics by combining the first plurality of predicted referee interaction metrics from the trained referee interaction prediction machine learning model and the second plurality of predicted referee interaction metrics from the additional trained referee interaction prediction machine learning model.
13 . The non-transitory computer-readable medium of claim 10 , further comprising instructions that, when executed by the at least one processor, cause the computer system to generate the plurality of referrer client device tiers based on the plurality of predicted referee interaction metrics by:
generating a ranked order of the plurality of predicted referee interaction metrics; and
identifying tier thresholds based on the ranked order of the plurality of predicted referee interaction metrics.
14 . The non-transitory computer-readable medium of claim 10 , further comprising instructions that, when executed by the at least one processor, cause the computer system to:
generate a referrer client device tier database by aggregating measured referee client device interactions from the set of measured referee client device interactions according to the plurality of referrer client device tiers and referrer invitation values transmitted to historical referrer client devices corresponding to the measured referee client device interactions; and
generate, utilizing the optimization model, referrer invitation values corresponding to the plurality of referrer client device tiers based on the referrer client device tier database.
15 . The non-transitory computer-readable medium of claim 10 , further comprising instructions that, when executed by the at least one processor, cause the computer system to generate the set of measured referee client device interactions for the plurality of referrer client device tiers by:
providing a set of digital notifications comprising referrer invitation values to a plurality of test referrer client devices; and
measuring interactions of referee client devices associated with test referrer client devices of the plurality of test referrer client devices.
16 . A system comprising:
at least one processor; and
at least one non-transitory computer-readable storage medium storing instructions that, when executed by the at least one processor, cause the system to:
generate, utilizing a trained referee interaction prediction machine learning model, a plurality of predicted referee interaction metrics from a plurality of referrer client device features;
generate a plurality of referrer client device tiers based on the plurality of predicted referee interaction metrics;
generate, utilizing an optimization model, a digital notification distribution policy for the plurality of referrer client device tiers from a set of measured referee client device interactions for the plurality of referrer client device tiers and a target interaction metric; and
transmit digital notifications to referrer client devices in accordance with the digital notification distribution policy and the plurality of referrer client device tiers.
17 . The system of claim 16 , further comprising instructions that, when executed by the at least one processor, cause the system to generate the plurality of predicted referee interaction metrics by:
generating, utilizing the trained referee interaction prediction machine learning model, a first plurality of predicted referee interaction metrics;
generating, utilizing an additional trained referee interaction prediction machine learning model, a second plurality of predicted referee interaction metrics; and
generating the plurality of predicted referee interaction metrics by combining the first plurality of predicted referee interaction metrics from the trained referee interaction prediction machine learning model and the second plurality of predicted referee interaction metrics from the additional trained referee interaction prediction machine learning model.
18 . The system of claim 16 , further comprising instructions that, when executed by the at least one processor, cause the system to transmit the digital notifications by:
extracting client device features for a candidate referrer client device; and
generating, utilizing the trained referee interaction prediction machine learning model, a predicted referee interaction metric for the candidate referrer client device from the client device features.
19 . The system of claim 18 , further comprising instructions that, when executed by the at least one processor, cause the system to transmit the digital notifications by:
selecting a referrer client device tier for the candidate referrer client device from the plurality of referrer client device tiers based on the predicted referee interaction metric for the candidate referrer client device; and
transmitting to the candidate referrer client device a digital notification comprising a referrer value according to the digital notification distribution policy and the referrer client device tier.
20 . The system of claim 16 , further comprising instructions that, when executed by the at least one processor, cause the system to train the trained referee interaction prediction machine learning model utilizing historical referrer client device features and a set of training referee client device interactions.