Granted Patent B2

US 12,646,083 · App. 14/926,852 · Granted Jun 2, 2026

Inventors: Stephanie Yang (New York, NY); Blake Shaw (New York, NY)

Assignee: Foursquare Labs, Inc.

G06Q30/02G06N20/00G06Q10/109

Abstract

Claims (59)

1 . A computer-implemented method comprising:

determining one or more new venues, wherein the determination is based upon identifying venue data existing for a period of time less than or equal to a predetermined time threshold value, wherein the venue data comprises active signal data and passive signal data received from a plurality of client devices, the passive signal data comprising wireless scan data collected by one or more client devices;

for the one or more new venues:

generating a total recent popularity using the venue data, wherein the total recent popularity is associated with a recent period of time that is less than or equal to the predetermined time threshold value, wherein the recent popularity is based at least in part upon a decayed sum of an aggregate count for one or more of a check-in to a venue, passive data indicating a visit to a venue, and saving information about a venue to a list;

generating a change in recent popularity using the venue data, wherein the change in recent popularity is associated with the recent period of time, wherein the change in recent popularity is based upon a linear trend line derived based upon a recent time series of the aggregate count for one or more of a second check-in to a venue, second passive data indicating a visit to a venue, and second saving information about a venue to a list; and

generating, using a machine-learning process, a predicted popularity value based on the total recent popularity and the change in recent popularity;

determining, based on predicted popularity values for the one or more new venues, a normalized predicted popularity value for each venue, wherein the normalized predicted popularity value is determined based at least upon a transformation of one or more features used to generate the predicted popularity value, wherein the transformation normalizes each feature value of the one or more features against comparable feature values of comparable venues having a similar category to the one or more new venues to approximate a standard distribution, the comparable venues being located in the geographic region of the one or more new venues;

determining, using a machine learning model, one or more scores for the one or more new venues based at least on the normalized predicted popularity values and user preferences, wherein the machine learning model generates the one or more scores by weighting collected venue data based upon a type associated with the collected venue data;

ranking the one or more new venues based at least on the one or more scores to generate a customized ranked list of venues; and

providing, to a processing device, at least a part of the customized ranked list of venues for display by the processing device, wherein the customized ranked list comprises summation data of current popularity for ranked venues included in the customized ranked list.

2 . The computer-implemented method according to claim 1 , wherein the venue data is collected data that is further selected from active signal data corresponding with the venue and received from the plurality of client devices.

3 . The computer-implemented method according to claim 2 , wherein the passive signal data is collected signal data selected from a group consisting of: collected wireless scan data, geocoding information, application information, or device information; and

wherein the active signal data is collected signal data that is selected from a group consisting of: venue check-in data, venue visit data, tips including the one or more new venues, recommendations including the one or more new venues, reviews of the one or more new venues, saved data corresponding with the one or more new venues, and likes/dislikes for the one or more new venues.

4 . The computer-implemented method according to claim 1 , wherein determining a normalized predicted popularity value further comprises:

evaluating the total recent popularity and the change in recent popularity using a weighted model.

5 . The computer-implemented method according to claim 4 , wherein machine-learning processing operations are applied for evaluating the total recent popularity and the change in recent popularity of the one or more new venues.

6 . The computer-implemented method according to claim 1 , wherein the customized ranked list comprises a display of the one or more new venues and normalized predicted popularity values.

7 . The computer-implemented method according to claim 6 , wherein the one or more new venues are located in a geographical region other than a geographical region associated with the processing device.

8 . The computer-implemented method according to claim 1 , further comprising updating the customized ranked list, and providing the updated customized ranked list to the processing device.

9 . A system comprising:

at least one processor; and

a memory operatively connected with the processor, wherein the memory comprises computer-executable instructions that, when executed by the at least one processor; cause the at least one processor to perform a method comprising:

evaluating a total recent popularity using the venue data, wherein the total recent popularity is associated with a recent period of time that is less than or equal to the predetermined time threshold value, wherein the recent popularity is based at least in part upon a decayed sum of an aggregate count for one or more of a check-in to a venue, passive data indicating a visit to a venue, and saving information about a venue to a list;

evaluating a change in recent popularity using the venue data, wherein the change in recent popularity is associated with the recent period of time, and wherein the change in recent popularity is based at least in part upon a linear trend line derived based upon a recent time series of the aggregate count for one or more of a check-in to a venue, passive data indicating a visit to a venue, and saving information about a venue to a list;

determining one or more new venues, wherein the determination is based upon identifying venue data existing for a period of time less than or equal to a predetermined time threshold value, wherein the venue data comprises active signal data and from passive signal data received from a plurality of client devices, the passive signal data comprising wireless scan data collected by one or more client devices;

evaluating the venue data associated with each venue of the one or more new venues to generate, using a machine-learning process, a predicted popularity value for each venue;

determining, based on predicted popularity values, a normalized predicted popularity value for each venue, wherein the normalized predicted popularity value is determined based at least upon a transformation of one or more features used to generate the predicted popularity value, wherein the transformation normalizes each feature value of the one or more features against comparable feature values of comparable venues having a similar category to the one or more new venues to approximate a standard distribution, the comparable venues being located the geographic region of the one or more new venues;

determining, using a machine learning model, one or more scores for the one or more new venues based at least on the normalized predicted popularity values and user preferences, wherein the machine learning model generates the one or more scores by weighting collected venue data based upon a type associated with the collected venue data;

ranking the one or more new venues based at least on the one or more scores to generate a customized ranked list of venues; and

providing at least a part of the customized ranked list of venues for display by a processing device, wherein the customized ranked list comprises summation data of current popularity for ranked venues included in the customized ranked list.

10 . The system according to claim 9 , wherein the venue data is collected data that is further selected from active signal data corresponding with the venue and received from the plurality of client devices;

wherein the passive signal data is collected signal data selected from a group consisting of: collected wireless scan data, geocoding information, application information, or device information; and

wherein the active signal data is collected signal data that is selected from a group consisting of: venue check-in data, venue visit data, tips including the one or more new venues, recommendations including the one or more new venues, reviews of the one or more new venues, saved data corresponding with the one or more new venues, or likes/dislikes for the one or more new venues.

11 . The system according to claim 10 , wherein generating the predicted popularity value further comprises:

generating the predicted popularity for the venue based on the total recent popularity and the change in recent popularity of the venue.

12 . The system according to claim 11 , wherein machine-learning processing operations are applied for evaluating the total recent popularity and the change in recent popularity of the one or more new venues.

13 . The system according to claim 9 , wherein the customized ranked list comprises a display of the one or more new venues and normalized predicted popularity values in a ranked order according to the generated normalized predicted popularity values.

14 . The system according to claim 13 , wherein the determined one or more new venues are located in a geographical region other than a geographical region associated with the processing device.

15 . The system according to claim 9 , wherein providing at least a part of the customized ranked list further comprises transmitting at least a part of the customized ranked list to another processing device.

16 . The system according to claim 9 , wherein the memory further comprises computer-executable instructions that, when executed by the at least one processor, perform operations comprising:

updating the customized ranked list; and

providing the updated customized ranked list to one or more processing devices.

17 . The system according to claim 9 , wherein normalized predicted popularity values are determined using a weighted model.

18 . A system comprising:

at least one processor; and

a memory operatively connected with the processor, wherein the memory comprises computer-executable instructions that, when executed by the at least one processor, perform a method comprising:

collecting venue data for a plurality of new venues, wherein the plurality of new venues are determined based upon identifying venue data existing for a period of time less than or equal to a predetermined time threshold value, wherein the venue data comprises active signal data and from passive signal data received from a plurality of client devices, the passive signal data comprising wireless scan data collected by one or more client devices;

evaluating the venue data for the plurality of new venues, wherein the evaluating further comprises, for the plurality of new venues:

determining a total recent popularity using the venue data, wherein the recent popularity is based at least in part upon a decayed sum of an aggregate count for one or more of a check-in to a venue, passive data indicating a visit to a venue, and saving information about a venue to a list;

determining a change in recent popularity using the venue data, wherein the total recent popularity and the change in recent popularity are each associated with a recent period of time that is less than or equal to the predetermined time threshold value, wherein the change in recent popularity is based upon a linear trend line derived based upon a recent time series of the aggregate count for the one or more of a check-in to a venue, passive data indicating a visit to a venue, and saving information about a venue to a list;

generating, using a machine-learning process, a predicted popularity value based on the total recent popularity and the change in recent popularity; and

determining, based on predicted popularity values for the one or more new venues, a normalized predicted popularity value for each venue, wherein the normalized predicted popularity value is determined based at least upon a transformation of one or more features used to generate the predicted popularity value, wherein the transformation normalizes the one or more features against one or more comparable features of comparable venues having a similar category to the one or more new venues in order to approximate a standard distribution, the comparable venues being located in the geographic region of the one or more new venues;

determining, using a machine learning model, one or more scores for the one or more new venues based at least on the normalized predicted popularity values and user preferences, wherein the machine learning model generates the one or more scores by weighting collected venue data based upon a type associated with the collected venue data;

generating a ranked list for the plurality of new venues based on the one or more scores;

generating a customized ranked list based upon user preferences; and

providing at least a part of the customized ranked list to a processing device for display, wherein the customized ranked list comprises summation data of current popularity for ranked venues included in the customized ranked list.

19 . The system according to claim 18 , wherein the memory further comprises computer-executable instructions that, when executed by the at least one processor, perform operations comprising:

updating the customized ranked list; and

transmitting the updated customized ranked list to one or more processing devices.

Assignments (12)

Continuity (1)

References Cited (16)