Methods and apparatus for user identification via community detection
Methods, apparatus, systems, and articles of manufacture for user identification via community detection are disclosed. Example instructions, when executed, cause at least one processor to at least access personally identifiable information to device links, build a device graph based on the personally identifiable information to device links, split components of the device graph into person clusters using community detection, create a snapshot including a device-to-person link lookup, and prepare a person-level impression measurement report from the snapshot.
1 . A non-transitory computer-readable storage medium, having stored thereon program instructions that, upon execution by a processor, cause performance of operations comprising:
obtaining device logins;
associating at least one device link with the device logins;
identifying each device login as a corresponding device link including the at least one device link;
graphing the at least one device link among device nodes into a device graph; and
preparing an impression report based partly on the device graph, wherein preparing the impression report based partly on the device graph comprises:
deduplicating impression data associated with the device graph to impression data associated with persons, wherein the impression data associated with the device graph is based on impression data from one or more panelists; and
generating the impression report including the persons.
2 . The non-transitory computer-readable storage medium of claim 1 , the operations further comprising:
associating, in the device graph, at least one panelist of the one or more panelists with the at least one device link.
3 . The non-transitory computer-readable storage medium of claim 1 , wherein the device logins are logins, on at least one device, to an email-associated website or app account.
4 . The non-transitory computer-readable storage medium of claim 3 , wherein the email-associated website or app account is associated with the New York Times, CNN.com, Netflix, Hulu, or Amazon Prime Video.
5 . The non-transitory computer-readable storage medium of claim 1 , wherein obtaining device logins comprises:
obtaining, from a plurality of user computing devices, at least one login for each user computing device of the plurality of user computing devices as a corresponding device login of the device logins.
6 . The non-transitory computer-readable storage medium of claim 1 , wherein the at least one device link includes data associated with at least one of an email address, a cookie ID, a mobile ad ID, a UID 2.0, a smart TV ID, an IP addresses, a ZIP code, a name, an address, an Experian ID, an Experian PID, an Experian LUID, or any combination thereof.
7 . A computing system comprising:
a smartphone configured to obtain device logins to an email-associated website or app account;
at least one processor; and
a non-transitory computer-readable storage medium, having stored thereon program instructions that, upon execution by the at least one processor, cause performance of operations comprising:
obtaining, from the smartphone, the device logins;
associating at least one device link with the device logins;
graphing the at least one device link among device nodes into a device graph; and
preparing an impression report based partly on the device graph, wherein preparing the impression report based partly on the device comprises:
deduplicating impression data associated with the device graph to impression data associated with persons, wherein the impression data associated with the device graph is based on impression data from one or more panelists.
8 . The computing system of claim 7 , the operations further comprising:
identifying each device login as a corresponding device link including the at least one device link.
9 . The computing system of claim 7 , wherein the operations further comprising:
associating, in the device graph, at least one panelist of the one or more panelists with the at least one device link.
10 . The computing system of claim 7 , the operations further comprising:
splitting the device graph into person-clusters using community detection, and wherein preparing the impression report based partly on the device graph includes preparing person-level impression estimates based at least party on the person-clusters of the device graph.
11 . A method comprising:
obtaining device logins;
associating at least one device link with the device logins;
identifying each device login as a corresponding device link including the at least one device link;
graphing the at least one device link among device nodes into a device graph; and
preparing an impression report based partly on the device graph, wherein preparing the impression report based partly on the device graph comprises:
deduplicating impression data associated with the device graph to impression data associated with persons, wherein the impression data associated with the device graph is based on impression data from one or more panelists; and
generating the impression report including the persons.
12 . The method of claim 11 , further comprising:
associating, in the device graph, at least one panelist of the one or more panelists with the at least one device link.
13 . The method of claim 11 , wherein obtaining device logins comprises:
obtaining, from a plurality of user computing devices, at least one login for each user computing device of the plurality of user computing devices as a corresponding device login of the device logins.
14 . The method of claim 11 , further comprising:
splitting the device graph into person-clusters using community detection.
15 . The method of claim 14 , wherein preparing the impression report based partly on the device graph includes preparing person-level impression estimates based at least party on the person-clusters of the device graph.