Intelligently customizing a cancellation notice for cancellation of a transportation request based on transportation features
The present disclosure relates to systems, non-transitory computer-readable media, and methods for intelligently determining to assess (or not assess) a cancellation penalty and notifying a requestor device of such a penalty in response to either a requestor's selection to cancel a single transportation request for a requestor or a requestor's selection to cancel a shared-transportation request associated with multiple requestors. For example, the disclosed systems can generate conversion probabilities for a requestor reflecting if a cancellation penalty is assessed—and if the cancellation penalty is not assessed—for a cancellation of a transportation request for transport of a lone requestor. Based on the conversion probabilities, the disclosed systems customize a cancellation notice for a requestor device to either include or exclude the cancellation penalty. Additionally, or alternatively, in certain implementations, the disclosed systems generate a transportation efficiency metric for modifying a projected transportation route for a shared transportation to account for a cancellation of a shared-transportation request. Based on the transportation efficiency metric, the disclosed systems customize a cancellation notice for a requestor device to either include or exclude the cancellation penalty.
1. 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, causes the system to:
train a conversion-machine-learning model utilizing ground-truth-conversions and historical transportation features comprising historical cancellation penalties by:
generating predicted conversion probabilities;
comparing the predicted conversion probabilities with ground-truth-conversion probabilities utilizing a loss function to determine a measure of loss; and
modifying parameters of the conversion-machine-learning model based on the measure of loss;
receive a transportation request from a requestor device;
extract a set of transportation features corresponding to the transportation request;
generate, from the set of transportation features utilizing the conversion-machine-learning model, a first conversion probability that a transportation provider transports a requestor associated with the requestor device if a cancellation penalty for cancelling the transportation request is assessed against the requestor;
generate, from the set of transportation features utilizing the conversion-machine-learning model, a second conversion probability that a transportation provider transports the requestor if the cancellation penalty for cancelling the transportation request is not assessed against the requestor; and
based on the first conversion probability and the second conversion probability, provide a customized cancellation notice to the requestor device in response to receiving an indication of a selection to cancel the transportation request from the requestor device.
2. The system of claim 1 , further comprising instructions that, when executed by the at least one processor, cause the system to:
generate a feedback probability that a requestor associated with the transportation request submits feedback in response to the cancellation penalty based on one or more of the set of transportation features corresponding to the transportation request; and
wherein providing the customized cancellation notice is based further on the feedback probability.
3. The system of claim 2 , wherein generating the feedback probability comprises utilizing a requestor-feedback-machine-learning model.
4. The system of claim 2 , further comprising instructions that, when executed by the at least one processor, cause the system to:
apply a transportation-value metric for the requestor to the feedback probability; and
provide the customized cancellation notice based further on the transportation-value metric for the requestor applied to the feedback probability.
5. The system of claim 1 , further comprising instructions that, when executed by the at least one processor, cause the system to provide the customized cancellation notice by:
providing a notification to the requestor device of the cancellation penalty associated with cancelling the transportation request; or
providing a notification to the requestor device that no cancellation penalty will be assessed for cancelling the transportation request.
6. The system of claim 1 , wherein the set of transportation features comprises one or more of:
an estimated time of arrival of a transportation provider at a pickup location of the requestor;
an amount of time between the transportation provider accepting the transportation request and the requestor cancelling the transportation request; or
a distance between the requestor device and the pickup location.
7. The system of claim 1 , wherein the first conversion probability comprises a requestor-session-conversion probability that a transportation provider transports the requestor to a drop-off location based on one of:
the transportation request by the requestor from the requestor device; or
a subsequent transportation request within a threshold time period by the requestor from the requestor device.
8. A computer-implemented method comprising:
training a conversion-machine-learning model utilizing ground-truth-conversions and historical transportation features comprising historical cancellation penalties by:
generating predicted conversion probabilities;
comparing the predicted conversion probabilities with ground-truth-conversion probabilities utilizing a loss function to determine a measure of loss; and
modifying parameters of the conversion-machine-learning model based on the measure of loss;
receiving a transportation request corresponding to a transportation matching system from a requestor device;
extracting a set of transportation features corresponding to the transportation request:
generating, from the set of transportation features utilizing the conversion-machine-learning model, a first conversion probability that a transportation provider transports a requestor associated with the requestor device if a cancellation penalty for cancelling the transportation request is assessed against the requestor;
generating, from the set of transportation features utilizing the conversion-machine-learning model, a second conversion probability that a transportation provider transports the requestor if the cancellation penalty for cancelling the transportation request is not assessed against the requestor; and
based on the first conversion probability and the second conversion probability, providing a customized cancellation notice to the requestor device in response to receiving an indication of a selection to cancel the transportation request from the requestor device.
9. The computer-implemented method of claim 8 , further comprising:
generating a feedback probability that a requestor associated with the transportation request submits feedback in response to the cancellation penalty based on one or more of the set of transportation features corresponding to the transportation request; and
wherein providing the customized cancellation notice is based further on the feedback probability.
10. The computer-implemented method of claim 9 , wherein generating the feedback probability comprises utilizing a requestor-feedback-machine-learning model.
11. The computer-implemented method of claim 9 , further comprising:
applying a transportation-value metric for the requestor to the feedback probability; and
providing the customized cancellation notice based further on the transportation-value metric for the requestor applied to the feedback probability.
12. The computer-implemented method of claim 8 , wherein providing the customized cancellation notice comprises:
providing a notification to the requestor device of the cancellation penalty associated with cancelling the transportation request; or
providing a notification to the requestor device that no cancellation penalty will be assessed for cancelling the transportation request.
13. The computer-implemented method of claim 8 , wherein the set of transportation features comprises one or more of:
an estimated time of arrival of a transportation provider at a pickup location of the requestor;
an amount of time between the transportation provider accepting the transportation request and the requestor cancelling the transportation request; or
a distance between the requestor device and the pickup location.
14. The computer-implemented method of claim 8 , wherein the first conversion probability comprises a requestor-session-conversion probability that a transportation provider transports the requestor to a drop-off location based on one of:
the transportation request by the requestor from the requestor device; or
a subsequent transportation request within a threshold time period by the requestor from the requestor device.
15. A non-transitory computer-readable medium comprising instructions that, when executed by at least one processor, causes a computer system to:
train a conversion-machine-learning model utilizing ground-truth-conversions and historical transportation features comprising historical cancellation penalties by:
generating predicted conversion probabilities;
comparing the predicted conversion probabilities with ground-truth-conversion probabilities utilizing a loss function to determine a measure of loss; and
modifying parameters of the conversion-machine-learning model based on the measure of loss;
receive a transportation request corresponding to a transportation matching system from a requestor device;
extract a set of transportation features corresponding to the transportation request:
generate, from the set of transportation features utilizing the-conversion-machine-learning model, a first conversion probability a transportation provider transports a requestor associated with the requestor device if a cancellation penalty for cancelling the transportation request is assessed against the requestor;
generate, utilizing the conversion-machine-learning model from the set of transportation features, a second conversion probability that a transportation provider transports the requestor if the cancellation penalty for cancelling the transportation request is not assessed against the requestor; and
based on the first conversion probability and the second conversion probability, provide a customized cancellation notice to the requestor device in response to receiving an indication of a selection to cancel the transportation request from the requestor device.
16. The non-transitory computer-readable medium of claim 15 , further comprising instructions that, when executed by the at least one processor, cause the computer system to:
generate a feedback probability that a requestor associated with the transportation request submits feedback in response to the cancellation penalty based on one or more of the set of transportation features corresponding to the transportation request; and
wherein providing the customized cancellation notice is based further on the feedback probability.
17. The non-transitory computer-readable medium of claim 16 , wherein generating the feedback probability comprises utilizing a requestor-feedback-machine-learning model.
18. The non-transitory computer-readable medium of claim 16 , further comprising instructions that, when executed by the at least one processor, cause the computer system to:
apply a transportation-value metric for the requestor to the feedback probability; and
provide the customized cancellation notice based further on the transportation-value metric for the requestor applied to the feedback probability.
19. The non-transitory computer-readable medium of claim 15 , further comprising instructions that, when executed by the at least one processor, cause the computer system to provide the customized cancellation notice by:
providing a notification to the requestor device of the cancellation penalty associated with cancelling the transportation request; or
providing a notification to the requestor device that no cancellation penalty will be assessed for cancelling the transportation request.
20. The non-transitory computer-readable medium of claim 15 , wherein the set of transportation features comprises one or more of:
an estimated time of arrival of a transportation provider at a pickup location of the requestor;
an amount of time between the transportation provider accepting the transportation request and the requestor cancelling the transportation request; or
a distance between the requestor device and the pickup location.