Customizing mapping interfaces using contemporaneous device request data in response to application interactions

1 . A system comprising:

at least one processor; and

a non-transitory computer readable storage medium comprising instructions that, when executed by the at least one processor, cause the system to:

receive, from a requester device in response to the requester device opening a requester application, an indication of a requester location for a requester;

determine, for a geographic region of the requester location, destination locations indicated in device requests received from a plurality of additional requester devices within a threshold time period of a current reference time for the requester;

determine, utilizing a classification model, a set of trending destination locations from the destination locations based on comparing numbers of device requests for the destination locations received within the threshold time period to a threshold number of requests for the threshold time period, wherein the classification model is trained in part by learning parameters of the classification model based on a device request history of destination locations for the requestor, historical requester destination locations within one or more geographical regions, and a plurality of user segments for a plurality of training time periods by iteratively:

generating a set of predicted trending destination locations based on the device request history, the historical requester destination locations within one or more geographical regions, and the plurality of user segments for the plurality of training time periods;

determining a loss utilizing a loss function that quantifies a difference between the set of predicted trending destination locations and ground-truth trending destination locations; and

adjusting the parameters of the classification model based on the loss;

determine an additional location indicated as sponsored by a third-party system within the geographic region of the requester location; and

provide for display within a mapping interface of the requester device in response to the requester device opening the requester application:

a first portion of the mapping interface comprising a map of the geographic region including an indication of the requester location;

a second portion of the mapping interface comprising a list combining the set of trending destination locations and a location indicator for the additional location; and

in response to an indication of a selected location from the list, a selectable option for initiating a device match request for a trending destination location or the additional location.

2 . The system of claim 1 , further comprising instructions that, when executed by the at least one processor, cause the system to:

receive, from the requester device, an indication of a selection of the selectable option for initiating the device match request via the mapping interface; and

send, to a provider device in response to the indication of the selection of the selectable option, a provider request comprising the trending destination location or the additional location based on the requester location.

3 . The system of claim 1 , further comprising instructions that, when executed by the at least one processor, cause the system to:

identify the device request history of destination locations for the requester.

4 . The system of claim 1 , further comprising instructions that, when executed by the at least one processor, cause the system to:

train, in a first training stage and utilizing the loss function, the classification model based on general destination location data associated with a plurality of requestors; and

train, in a second training stage and utilizing the loss function, the classification model based on specific destination location data corresponding to the requester that comprises the device request history.

5 . The system of claim 3 , further comprising instructions that, when executed by the at least one processor, cause the system to determine the set of trending destination locations by:

determining that the requester and an additional requester associated with the requester within the geographic region of the requester location are in a group;

identifying a device request history of destination locations for the additional requester; and

determining the set of trending destination locations based on the device request history of destination locations for the requester and the device request history of destination locations for the additional requester.

6 . The system of claim 1 , further comprising instructions that, when executed by the at least one processor, cause the system to determine the set of trending destination locations by:

identifying a category of destination locations for the requester; and

determining the set of trending destination locations according to the category of destination locations.

7 . The system of claim 1 , further comprising instructions that, when executed by the at least one processor, cause the system to:

receive an indication of a placement of a user-identified destination location at the requester device; and

generate, for display at an additional requester device of an additional requester associated with the requester, a selectable map marker representing the user-identified destination location within a mapping interface of the additional requester device in response to the indication of the placement of the user-identified destination location at the requester device.

8 . The system of claim 1 , further comprising instructions that, when executed by the at least one processor, cause the system to determine the destination locations indicated in the device requests received from the plurality of additional requester devices by:

receiving, from a plurality of provider devices, additional device request data corresponding to fulfilled device requests including GPS data and wireless location data of the plurality of provider devices; and

verifying the destination locations indicated in the device requests utilizing the additional device request data of the fulfilled device requests.

9 . A non-transitory computer readable storage medium comprising instructions that, when executed by at least one processor, cause a computing device to:

receive, from a requester device in response to the requester device opening a requester application, an indication of a requester location for a requester;

determine, for a geographic region of the requester location, destination locations indicated in device requests received from a plurality of additional requester devices within a threshold time period of a current reference time for the requester;

determine, utilizing a classification model, a set of trending destination locations from the destination locations comparing numbers of device requests for the destination locations received within the threshold time period to a threshold number of requests for the threshold time period, wherein the classification model is trained in part by learning parameters of the classification model based on a device request history of destination locations for the requestor, historical requester destination locations within one or more geographical regions, and a plurality of user segments for a plurality of training time periods by iteratively:

generating a set of predicted trending destination locations based on the device request history, the historical requester destination locations within one or more geographical regions, and the plurality of user segments for the plurality of training time periods;

determining a loss utilizing a loss function that quantifies a difference between the set of predicted trending destination locations and ground-truth trending destination locations; and

adjusting the parameters of the classification model based on the loss;

determine an additional location indicated as sponsored by a third-party system within the geographic region of the requester location; and

provide for display within a mapping interface of the requester device in response to the requester device opening the requester application:

a first portion of the mapping interface comprising a map of the geographic region including an indication of the requester location;

a second portion of the mapping interface comprising a list combining the set of trending destination locations and a location indicator for the additional location; and

in response to an indication of a selected location from the list, a selectable option for initiating a device match request for a trending destination location or the additional location.

10 . The non-transitory computer readable storage medium of claim 9 , further comprising instructions that, when executed by the at least one processor, cause the computing device to:

receive, from the requester device, an indication of a selection of the selectable option for initiating the device match request via the mapping interface; and

send, to a provider device in response to the indication of the selection of the selectable option, a provider request comprising the trending destination location or the additional location based on the requester location.

11 . The non-transitory computer readable storage medium of claim 9 , further comprising instructions that, when executed by the at least one processor, cause the computing device to:

identify the device request history of destination locations for the requester.

12 . The non-transitory computer readable storage medium of claim 9 , further comprising instructions that, when executed by the at least one processor, cause the computing device to:

train, in a first training stage and utilizing the loss function, the classification model based on general destination location data associated with a plurality of requestors; and

train, in a second training stage and utilizing the loss function, the classification model based on specific destination location data corresponding to the requester that comprises the device request history.

13 . The non-transitory computer readable storage medium of claim 11 , further comprising instructions that, when executed by the at least one processor, cause the computing device to determine the set of trending destination locations by:

determining that the requester and an additional requester associated with the requester within the geographic region of the requester location are in a group;

identifying a device request history of destination locations for the additional requester; and

determining the set of trending destination locations based on the device request history of destination locations for the requester and the device request history of destination locations for the additional requester.

14 . The non-transitory computer readable storage medium of claim 9 , further comprising instructions that, when executed by the at least one processor, cause the computing device to determine the current reference time for the requester as a time at which the requester device opens the requester application.

15 . A method comprising:

receiving, from a requester device in response to the requester device opening a requester application, an indication of a requester location for a requester;

determining, for a geographic region of the requester location, destination locations indicated in devices requests received from a plurality of additional requester devices within a threshold time period of a current reference time for the requester;

determining, utilizing a classification model, a set of trending destination locations from the destination locations based on comparing numbers of device requests for the destination locations received within the threshold time period to a threshold number of requests for the threshold time period, wherein the classification model is trained in part by learning parameters of the classification model based on a device request history of destination locations for the requestor, historical requester destination locations within one or more geographical regions, and a plurality of user segments for a plurality of training time periods by iteratively:

generating a set of predicted trending destination locations based on the device request history, the historical requester destination locations within one or more geographical regions, and the plurality of user segments for the plurality of training time periods;

determining a loss utilizing a loss function that quantifies a difference between the set of predicted trending destination locations and ground-truth trending destination locations; and

adjusting the parameters of the classification model based on the loss;

determining an additional location indicated as sponsored by a third-party system within the geographic region of the requester location; and

providing for display within a mapping interface of the requester device in response to the requester device opening the requester application:

a first portion of the mapping interface comprising a map of the geographic region including an indication of the requester location;

a second portion of the mapping interface comprising a list combining the set of trending destination locations and a location indicator for the additional location; and

in response to an indication of a selected location from the list, a selectable option for initiating a device match request for a trending destination location or the additional location.

16 . The method of claim 15 , further comprising:

receiving, from the requester device, an indication of a selection of the selectable option for initiating the device match request via the mapping interface; and

sending, to a provider device in response to the indication of the selection of the selectable option, a provider request comprising the trending destination location or the additional location based on the requester location.

17 . The method of claim 15 , further comprising

identifying the device request history of destination locations for the requester.

18 . The method of claim 17 , further comprising:

training, in a first training stage and utilizing the loss function, the classification model based on general destination location data associated with a plurality of requestors; and

training, in a second training stage and utilizing the loss function, the classification model based on specific destination location data corresponding to the requester that comprises the device request history.

19 . The method of claim 17 , wherein determining the set of trending destination locations comprises:

determining that the requester and an additional requester associated with the requester within the geographic region of the requester location are in a group;

identifying a device request history of destination locations for the additional requester; and

determining the set of trending destination locations based on the device request history of destination locations for the requester and the device request history of destination locations for the additional requester.

20 . The method of claim 17 , further comprising determining the current reference time for the requester as a time at which the requester device opens the requester application.