Systems and methods for statistically associating mobile devices to households
Embodiments of the invention relate to methods and systems for associating a mobile device to a household. In various embodiments, a plurality of latitude-longitude pairs is received for a mobile device during a time period. The latitude-longitude pairs are organized into a plurality of clusters corresponding to geographic regions visited by the mobile device during the time period. For each cluster, a score is calculated that represents a likelihood that a user of the mobile device resides in a household within the cluster. The cluster with the highest score is identified as being the location of the user's household. The computation is preferably conducted recursively over time periods. The mobile device is then associated with the user's household.
1. A method of associating a mobile device to a household, the method comprising:
receiving a plurality of latitude-longitude pairs for a mobile device, the latitude-longitude pairs corresponding to a location of the mobile device during a time period;
organizing the plurality of latitude-longitude pairs into a plurality of clusters, the clusters corresponding to geographic regions visited by the mobile device during the time period;
calculating a score for each cluster, the score representing a likelihood that a user of the mobile device resides in a household within the cluster;
identifying a location of the user's household, the location corresponding to the cluster having the highest score;
associating the mobile device with the user's household;
receiving, during a second time period, a plurality of second latitude-longitude pairs for the mobile device, the second latitude-longitude pairs corresponding to a location of the mobile device during the second time period;
organizing the plurality of second latitude-longitude pairs into a plurality of second clusters, the second clusters corresponding to geographic regions visited by the mobile device during the second time period;
calculating a second score for each second cluster, the second score representing a likelihood that the user of the mobile device resides in a household within the second cluster; and
identifying a second location of the user's household, the second location corresponding to the second cluster having the highest second score.
2. The method of claim 1 , wherein the latitude-longitude pairs are obtained using a technique selected from the group consisting of global positioning with a global positioning system, cellular triangulation, and/or WiFi identification.
3. The method of claim 1 , wherein the time period is selected from an hour, a day, a week, or a month.
4. The method of claim 1 , wherein the mobile device is selected from a cellular phone, a portable computing device, and a tablet computer.
5. The method of claim 1 , wherein the latitude-longitude pairs are associated with an ad request sent by the mobile device.
6. The method of claim 1 , wherein calculating the score comprises applying a weight based on at least one of a time and a location associated with a longitude-latitude pair.
7. The method of claim 6 , wherein the weight is based on the time associated with the latitude-longitude pair, according to a likelihood that the mobile device was in the household at the time.
8. The method of claim 6 , wherein the weight is based on the location associated with the latitude-longitude pair, according to a likelihood that the location corresponds to a residential location.
9. The method of claim 1 , further comprising rejecting a latitude-longitude pair based on an indication that the latitude-longitude pair is erroneous.
10. The method of claim 1 , further comprising performing a state update, the state update comprising:
determining whether the location of the user's household and the second location of the user's household are consistent;
if the location and the second location are consistent, determining a revised location of the user's household as a function of the location and the second location;
if (i) the location and the second location are not consistent, and (ii) the highest second score is less than or equal to the highest score, associating the mobile device with the location; and
if (i) the location and the second location are not consistent, and (ii) the highest second score is greater than the highest score, associating the mobile device with the second location of the user's household.
11. The method of claim 1 , further comprising assigning statistical metrics to measure fitness of the device ID and household ID association.
12. A system for associating a mobile device to a household, the system comprising:
a data storage device operating on a server computer, the data storage device storing a plurality of latitude-longitude pairs for a mobile device, the latitude-longitude pairs corresponding to a location of the mobile device during a time period; and
a processor for executing computer-readable instructions that, when executed:
organize the plurality of latitude-longitude pairs into a plurality of clusters, the clusters corresponding to geographic regions visited by the mobile device during the time period;
calculate a score for each cluster, the score representing a likelihood that a user of the mobile device resides in a household within the cluster;
identify a location of the user's household, the location corresponding to the cluster having the highest score;
associate the mobile device with the user's household;
receive, during a second time period, a plurality of second latitude-longitude pairs for the mobile device, the second latitude-longitude pairs corresponding to a location of the mobile device during the second time period;
organize the plurality of second latitude-longitude pairs into a plurality of second clusters, the second clusters corresponding to geographic regions visited by the mobile device during the second time period;
calculate a second score for each second cluster, the second score representing a likelihood that the user of the mobile device resides in a household within the second cluster; and
identify a second location of the user's household, the second location corresponding to the second cluster having the highest second score.
13. The system of claim 12 , wherein the latitude-longitude pairs are obtained using a technique selected from the group consisting of global positioning with a global positioning system, cellular triangulation, and/or WiFi identification.
14. The system of claim 12 , wherein the time period is selected from an hour, a day, a week, or a month.
15. The system of claim 12 , wherein the mobile device is selected from a cellular phone, a portable computing device, and a tablet computer.
16. The system of claim 12 , wherein the latitude-longitude pairs are associated with an ad request sent by the mobile device.
17. The system of claim 12 , wherein calculating the score comprises applying a weight based on at least one of a time and a location associated with a longitude-latitude pair.
18. The system of claim 17 , wherein the weight is based on the time associated with the latitude-longitude pair, according to a likelihood that the mobile device was in the household at the time.
19. The system of claim 17 , wherein the weight is based on the location associated with the latitude-longitude pair, according to a likelihood that the location corresponds to a residential location.
20. The system of claim 12 , wherein the computer-readable instructions, when executed, further reject a latitude-longitude pair based on an indication that the latitude-longitude pair is erroneous.
21. The system of claim 12 , wherein the computer-readable instructions, when executed, perform a state update comprising:
determining whether the location of the user's household and the second location of the user's household are consistent;
if the location and the second location are consistent, determining a revised location of the user's household as a function of the location and the second location;
if (i) the location and the second location are not consistent, and (ii) the highest second score is less than or equal to the highest score, associating the mobile device with the location; and
if (i) the location and the second location are not consistent, and (ii) the highest second score is greater than the highest score, associating the mobile device with the second location of the user's household.
22. The system of claim 12 , wherein the computer-readable instructions, when executed, assign statistical metrics to measure fitness of the device ID and household ID association.
23. A non-transitory computer program product embodied on a computer-readable medium and comprising computer code for associating a mobile device to a household, the code comprising instructions for:
receiving a plurality of latitude-longitude pairs for a mobile device, the latitude-longitude pairs corresponding to a location of the mobile device during a time period;
organizing the plurality of latitude-longitude pairs into a plurality of clusters, the clusters corresponding to geographic regions visited by the mobile device during the time period;
calculating a score for each cluster, the score representing a likelihood that a user of the mobile device resides in a household within the cluster;
identifying a location of the user's household, the location corresponding to the cluster having the highest score;
associating the mobile device with the user's household;
receiving, during a second time period, a plurality of second latitude-longitude pairs for the mobile device, the second latitude-longitude pairs corresponding to a location of the mobile device during the second time period;
organizing the plurality of second latitude-longitude pairs into a plurality of second clusters, the second clusters corresponding to geographic regions visited by the mobile device during the second time period;
calculating a second score for each second cluster, the second score representing a likelihood that the user of the mobile device resides in a household within the second cluster; and
identifying a second location of the user's household, the second location corresponding to the second cluster having the highest second score.