Behavioral predictions based on network activity locations

• US 7,958,228 B2
• Filed: 07/11/2007
• Issued: 06/07/2011
• Est. Priority Date: 07/11/2007
• Status: Active Grant

• Current Assignee: R2 Solutions LLC (Acacia Research Corporation)
• Original Assignee: Yahoo! Inc. (Apollo Global Management, Inc.)
• Inventors: Pai, Ramesh G., Law, Mark Nicholas, Patel, Devesh, Riise, Soren
• Primary Examiner(s): Backer; Firmin
• Assistant Examiner(s): ALGIBHAH, HAMZA N
• Application Number:
  US11/776,488
• Publication Number:
  US 20090019182A1
• Time in Patent Office: 1,427 Days
• Field of Search: 709/224
• US Class Current: 709/224
• CPC Class Codes: G06Q 10/00 Administration; Management

Abstract

A computer-implemented method for constructing network activity profiles is provided, which comprises the following: obtaining a plurality of records of network activities from an activity source, each record corresponding to an interaction with a network resource via the network from the activity source, wherein each record comprises at least a network endpoint address from where the interaction originates and an indication of a time of the interaction; for each record, determining a geographical location corresponding to the network endpoint address of that record and associating the determined geographical location with that record; and constructing at least one profile for the activity source based on the plurality of records and at least one geographical location associated with the records, wherein each profile comprises a time-based behavior pattern associated with the at least one geographical location.

Claims

1. A computer-implemented method for constructing network activity profiles, comprising:

obtaining a plurality of records of network activities originating from an activity source, the activity source corresponding to a particular user account, network browser or application program, on the network, and each record corresponding to an interaction by the activity source with a network resource via the network, wherein each record comprises at least a network endpoint address associated with a physical device, the physical device providing access to the network and the network endpoint address used in the interaction between the activity source and the network resource, and wherein each record further comprises an indication of a time of the interaction by the activity source with the network resource;

for each record, determining one or more record geographical locations corresponding to the network endpoint address associated with the physical device by mapping the network endpoint address to the geographical location or locations, wherein the determined record geographical location(s) are based in part on a selected desired level or levels of geographic precision;

associating the determined record geographical location or locations with that record; and

constructing at least one profile for the activity source based on the plurality of records and at least one geographical location associated with the plurality of records, wherein each profile comprises a time-based behavior pattern associated with the at least one geographical location.

2. The computer-implemented method, as recited in claim 1, further comprising:

predicting future behaviors of the activity source based on the at least one profile.

3. The computer-implemented method, as recited in claim 1, wherein the time-based behavior pattern associated with the at least one geographical location represents a time distribution for network activities originated from each of the at least one geographical location during a predetermined period of time.

4. The computer-implemented method, as recited in claim 1, wherein the at least one profile comprises a daily profile comprising a time distribution for network behaviors originated from each of the at least one geographical location during a day.

5. The computer-implemented method, as recited in claim 4, wherein the predicting future behaviors of the activity source based on the at least one profile comprises predicting future daily behaviors of the activity source associated with each of the at least one geographical location based on the daily profile.

6. The computer-implemented method, as recited in claim 1, wherein the at least one profile comprises a weekly profile comprising a time distribution for network activities originated from each of the at least one geographical location for each day of a week during the week.

7. The computer-implemented method, as recited in claim 6, wherein the predicting future behaviors of the activity source based on the at least one profile comprises predicting future weekly behaviors of the activity source associated with each of the at least one geographical location based on the weekly profile.

8. The computer-implemented method, as recited in claim 1, wherein the at least one profile comprises a monthly profile comprising a time distribution for network activities originated from each of the at least one geographical location for each day of a month during the month.

9. The computer-implemented method, as recited in claim 8, wherein the predicting future behaviors of the activity source based on the at least one profile comprises predicting future monthly behaviors of the activity source associated with each of the at least one geographical location based on the monthly profile.

10. The computer-implemented method, as recited in claim 1, wherein the at least one profile comprises a yearly profile comprising a time distribution for network activities originated from each of the at least one geographical location for each month of a year during the year.

11. The computer-implemented method, as recited in claim 10, wherein the predicting future behaviors of the activity source based on the at least one profile comprises predicting future yearly behaviors of the activity source associated with each of the at least one geographical location based on the yearly profile.

12. The computer-implemented method, as recited in claim 1, wherein the activity source is one selected from the group consisting of web browser, and user account.

13. The computer-implemented method, as recited in claim 1, wherein the geographical location or locations is one or more selected from the group consisting of country, state, province, county, city, most granular location, or direct market area.

14. The computer-implemented method, as recited in claim 1, wherein the network endpoint address is an IP (Internet Protocol) address and the indication of the time is a timestamp.

15. A computer-implemented method for predicting future behaviors, comprising:

constructing at least one profile for an activity source based on a plurality of records of network activities from the activity source and at least one geographical location associated with the records, the activity source corresponding to a particular user account, network browser or application program, on the network, and wherein

each record corresponds to an interaction by the activity source with a network resource via the network originated from a record geographical location corresponding to the network endpoint address in the record, and

each record comprises the record geographical location and an indication of a time of the interaction by the activity source with the network resource, the record geographical location determined based at least in part on a network endpoint address associated with a physical device and on a selected desired level of geographic precision, wherein the physical device provides access to the network and the network endpoint address is used in the interaction between the activity source and the network resource; and

each profile comprises a time-based behavior pattern associated with the at least one geographical location; and

predicting future behaviors of the activity source based on the at least one profile.

16. The computer-implemented method, as recited in claim 15, wherein the predicting future behaviors of the activity source based on the at least one profile comprises predicting future behaviors expected to originate from each of the at least one geographical location based on the time-based behavior pattern associated with that geographical location.

17. The computer-implemented method, as recited in claim 16, further comprising:

providing information associated with each of the at least one geographical location based on the predicted future behaviors expected to originate from that geographical location.

18. The computer-implemented method, as recited in claim 17, wherein the providing information associated with each of the at least one geographical location based on the predicted future behaviors expected to originate from that geographical location comprises for each of the at least one geographical location, providing information based on that geographical location and the predicted future behaviors associated with that geographical location for a time when the predicted future behaviors are expected to originate from that geographical location.

19. The computer-implemented method, as recited in claim 17, wherein the information is at least one selected from the group consisting of advertisement, merchandise offers, service offers, weather information, news, and travel information.

20. The computer-implemented method, as recited in claim 15, wherein the at least one profile comprises

a daily profile comprising a time distribution for network activities originated from each of the at least one geographical location during a day,

a weekly profile comprising a time distribution for network activities originated from each of the at least one geographical location for each day of a week during the week,

a monthly profile comprising a time distribution for network activities originated from each of the at least one geographical location for each day of a month during the month, and

a yearly profile comprising a time distribution for network activities originated from each of the at least one geographical location for each month of a year during the year.

21. The computer-implemented method, as recited in claim 20, wherein the predicting future behaviors of the activity source based on the at least one profile comprises

predicting future daily behaviors of the activity source associated with each of the at least one geographical location based on the daily profile,

predicting future weekly behaviors of the activity source associated with each of the at least one geographical location based on the weekly profile,

predicting future monthly behaviors of the activity source associated with each of the at least one geographical location based on the monthly profile, and

predicting future yearly behaviors of the activity source associated with each of the at least one geographical location based on the yearly profile.

22. The computer-implemented method, as recited in claim 20, further comprising:

providing daily information associated with each of the at least one geographical location based on the predicted daily behaviors expected to originate from that geographical location;

providing weekly information associated with each of the at least one geographical location based on the predicted weekly behaviors expected to originate from that geographical location;

providing monthly information associated with each of the at least one geographical location based on the predicted monthly behaviors expected to originate from that geographical location; and

providing yearly information associated with each of the at least one geographical location based on the predicted yearly behaviors expected to originate from that geographical location.

23. A computer program product for constructing network activity profiles comprising a non-transitory computer-readable medium having a plurality of computer program instructions stored therein, which are operable to cause at least one computing device to:

obtain a plurality of records of network activities originating from an activity source, the activity source corresponding to a particular user account, network browser or application program, on the network, and each record corresponding to an interaction by the activity source with a network resource via the network, wherein each record comprises at least a network endpoint address associated with a physical device, the physical device providing access to the network and the network endpoint address used in the interaction between the activity source and the network resource, and wherein each record further comprises an indication of a time of the interaction by the activity source with the network resource;

for each record, determining determine one or more record geographical locations corresponding to the network endpoint address associated with the physical device by mapping the network endpoint address to the geographical location or locations, wherein the determined record geographical location(s) are based in part on a selected desired level or levels of geographic precision;

associate the determined record geographical location or locations with that record; and

construct at least one profile for the activity source based on the plurality of records and at least one geographical location associated with the plurality of records, wherein each profile comprises a time-based behavior pattern associated with the at least one geographical location.

24. A computer program product predicting future behaviors comprising a non-transitory computer-readable medium having a plurality of computer program instructions stored therein, which are operable to cause at least one computing device to:

construct at least one profile for an activity source based on a plurality of records of network activities from the activity source and at least one geographical location associated with the records, the activity source corresponding to a particular user account, network browser or application program, on the network, and wherein

each record corresponds to an interaction by the activity source with a network resource via the network originated from a record geographical location corresponding to the network endpoint address in the record, and

each record comprises the record geographical location and an indication of a time of the interaction by the activity source with the network resource, the record geographical location determined based at least in part on a network endpoint address associated with a physical device and on a selected desired level of geographic precision, wherein the physical device provides access to the network and the network endpoint address is used in the interaction between the activity source and the network resource, and each profile comprises a time-based behavior pattern associated with the at least one geographical location; and

predict future behaviors of the activity source based on the at least one profile.