Device identifier similarity models derived from online event signals

• US 9,065,727 B1
• Filed: 08/31/2012
• Issued: 06/23/2015
• Est. Priority Date: 08/31/2012
• Status: Active Grant

• Current Assignee: Google LLC (Alphabet Inc.)
• Original Assignee: Google Inc. (Alphabet Inc.)
• Inventors: Liu, Jia, Bai, Yijian, Patil, Manojav, Ravichandran, Deepak, Jiampojamarn, Sittichai, Ponnekanti, Shankar
• Primary Examiner(s): Dalencourt, Yves
• Assistant Examiner(s): MENDAYE, KIDEST H
• Application Number:
  US13/601,775
• Time in Patent Office: 1,026 Days
• Field of Search: 709/219, 709/223, 709/224, 705/7.29, 705/7.31, 705/26.1, 705/26.7, 705/14.53, 705/14.54, 705/341, 707/3, 707/710, 707/737, 707/749, 707/751, 707/769
• US Class Current: 1/1
• CPC Class Codes: G06Q 30/06 Buying, selling or leasing transactions
• CPC Class Codes: H04L 43/04 Processing captured monitoring data, e.g. for logfile generation
• CPC Class Codes: H04L 67/535 Tracking the activity of the user network monitoring arrangements H04L43/00; recording of computer activity G06F11/34

Abstract

A computerized method and system operable to build a device identifier similarity model with online event signals and determine similar network device identifiers. A processing circuit receives a first set of network device identifiers. The processing circuit represents each network device identifier of the first set by feature data associated with each network device identifier'"'"'s network activity, where the feature data is associated with the content clicked-on or converted-on. The processing circuit applies abstractions on the feature data to form concepts. The processing circuit derives at least one hierarchy of feature data based on the keywords and concepts of the feature data. The processing circuit expands the feature data based on the derived at least one hierarchy of feature data and generates the device identifier similarity model based on the expanded feature data. The processing circuit is also capable of determining long-term and short-term history events.

Claims

1. A computerized method of building a device identifier similarity model with online event signals, the method comprising:

receiving at a processing circuit a first set of network device identifiers;

identifying, by the processing circuit, an online event associated with network activity of each network device identifier of the first set;

identifying, using the processing circuit, for each network device identifier of the first set, one or more long-term browsing history events surrounding the identified online event based on the network device identifier'"'"'s network activity, the long-term browsing history events corresponding to events occurring prior to a first time from the identified online event;

identifying, using the processing circuit, for each network device identifier of the first set, one or more short-term browsing history events surrounding the identified online event based on the network device identifier'"'"'s network activity, the short-term browsing history events corresponding to events occurring after the first time from the identified online event;

representing, using the processing circuit, each device identifier of the first set as a vector based on feature data corresponding to each network device identifier'"'"'s network activity, the feature data comprising keywords corresponding to content associated with the device identifier'"'"'s network activity;

applying, using the processing circuit, abstractions on the feature data to form concepts, wherein each concept represents a category of interest;

deriving, using the processing circuit, at least one hierarchy of the feature data based on the keywords and concepts of the feature data;

expanding, using the processing circuit, the feature data based on the derived at least one hierarchy of the feature data;

applying, using the processing circuit, a clustering algorithm on each of the vectors to identify a plurality of clusters of device identifiers that share a common interest;

providing, using the processing circuit, at least one subset of network device identifiers corresponding to each of the plurality of cluster; and

generating, using the processing circuit, the device identifier similarity model based on the expanded feature data.

2. The computerized method according to claim 1, further comprising:

determining using the processing circuit a first online event signal and a second online event signal based on each network device identifier'"'"'s network activity; and

determining using the processing circuit a time representation of the time between the first online event signal and the second online event signal.

3. The computerized method according to claim 2, further comprising:

comparing using the processing circuit the time representation to a first predetermined value; and

identifying using the processing circuit a long-term time representation, if the time representation is greater than the first predetermined value.

4. The computerized method according to claim 2, further comprising:

comparing using the processing circuit the time representation to a second predetermined value; and

identifying using the processing circuit a short-term time representation, if the time representation is lesser than the second predetermined value.

5. A computerized method for identifying similar network device identifiers, the method comprising:

receiving, at a processing circuit, a first set of network device identifiers;

identifying, by the processing circuit, an online event associated with network activity of each network device identifier of the first set;

identifying, using the processing circuit, for each network device identifier of the first set, one or more long-term browsing history events surrounding the identified online event based on the network device identifier'"'"'s network activity, the long-term browsing history events corresponding to events occurring prior to a first time from the identified online event;

identifying, using the processing circuit, for each network device identifier of the first set, one or more short-term browsing history events surrounding the identified online event based on the network device identifier'"'"'s network activity, the short-term browsing history events corresponding to events occurring after the first time from the identified online event;

representing, using the processing circuit, each device identifier of the first set as a vector based on feature data corresponding to each network device identifier'"'"'s network activity, the feature data comprising keywords corresponding to content associated with the device identifier'"'"'s network activity;

applying, using the processing circuit, abstractions on the feature data to form concepts, wherein each concept represents a category of interest;

deriving, using the processing circuit, at least one hierarchy of the feature data based on the keywords and concepts of the feature data;

expanding, using the processing circuit, the feature data based on the derived at least one hierarchy of the feature data;

applying, using the processing circuit, a clustering algorithm on each of the vectors to identify a plurality of clusters of device identifiers that share a common interest;

providing, using the processing circuit, at least one subset of network device identifiers corresponding to each of the plurality of clusters; and

generating using the processing circuit the set of similar network device identifiers based on the expanded feature data.

6. The computerized method according to claim 5, further comprising:

identifying using the processing circuit the one or more long-term browsing history events and the one or more the short-term browsing history events by;

identifying using the processing circuit a time representation between a first online event and a second online event; and

comparing using the processing circuit the time representation with a predetermined value, wherein if the time representation is greater than the predetermined value the time representation will indicate a long-term browsing history event, and wherein if the time representation is lesser than the predetermined value the time representation will indicate a short-term browsing history event.

7. The computerized method according to claim 6, wherein the predetermined value is adjustable.

8. The computerized method according to claim 6, further comprising:

incrementing, by the processing circuit, the predetermined value by a unit of measurement.

9. A system for building a device identifier similarity model with online event signals comprising a processing circuit including a processor and a memory coupled thereto, the processing circuit operable to:

receive a first set of network device identifiers;

identify an online event associated with network activity of each network device identifier of the first set;

identify, for each network device identifier of the first set, one or more long-term browsing history events surrounding the identified online event based on the network device identifier'"'"'s network activity, the long-term browsing history events corresponding to events occurring prior to a first time from the identified online event;

identify, for each network device identifier of the first set, one or more short-term browsing history events surrounding the identified online event based on the network device identifier'"'"'s network activity, the short-term browsing history events corresponding to events occurring after the first time from the identified online event;

represent each device identifier of the first set as a vector based on feature data corresponding to each network device identifier'"'"'s network activity, the feature data comprising keywords corresponding to content associated with the device identifier'"'"'s network activity;

apply abstractions on the feature data to form concepts, wherein each concept represents a category of interest;

derive at least one hierarchy of the feature data based on the keywords and concepts of the feature data;

expand the feature data based on the derived at least one hierarchy of the feature data;

apply a clustering algorithm on each of the vectors to identify a plurality of clusters of device identifiers that share a common interest;

provide at least one subset of network device identifiers corresponding to each of the plurality of clusters; and

generate the device identifier similarity model based on the expanded feature data.

10. The system of claim 9, wherein the processing circuit is further operable to:

determine a first online event signal and a second online event signal based on each network device identifier'"'"'s network activity; and

determine a time representation of the time between the first online event signal and the second online event signal.

11. The system of claim 10, wherein the processing circuit is further operable to:

compare the time representation to a first predetermined value; and

identify a long-term time representation, if the time representation is greater than the first predetermined value.

12. The system of claim 10, wherein the processing circuit is further operable to:

compare the time representation to a second predetermined value; and

identify a short-term time representation, if the time representation is lesser than the second predetermined value.

13. A system for identifying similar network device identifiers comprising a processing circuit including a processor and a memory coupled thereto, the processing circuit operable to:

receive a set of network device identifiers;

identify an online event associated with network activity of each network device identifier of the first set;

identify, for each network device identifier of the first set, one or more long-term browsing history events surrounding the identified online event based on the network device identifier'"'"'s network activity, the long-term browsing history events corresponding to events occurring prior to a first time from the identified online event;

identify, for each network device identifier of the first set, one or more short-term browsing history events surrounding the identified online event based on the network device identifier'"'"'s network activity, the short-term browsing history events corresponding to events occurring after the first time from the identified online event;

represent each device identifier of the first set as a vector based on feature data corresponding to each network device identifier'"'"'s network activity, the feature data comprising keywords corresponding to content associated with the device identifier'"'"'s network activity;

apply abstractions on the feature data to form concepts, wherein each concept represents a category of interest;

derive at least one hierarchy of the feature data based on the keywords and concepts of the feature data;

expand the feature data based on the derived at least one hierarchy of the feature data;

apply a clustering algorithm on each of the vectors to identify a plurality of clusters of device identifiers that share a common interest;

provide at least one subset of network device identifiers corresponding to each of the plurality of clusters; and

generate the set of similar network device identifiers based on the expanded feature data.

14. The system of claim 13, wherein the processing circuit is further operable to:

identify the one or more long-term browsing history events and the one or more the short-term browsing history events by;

identify a time representation between a first online event and a second online event; and

compare the time representation with a predetermined value, wherein if the time representation is greater than the predetermined value the time representation will indicate a long-term browsing history event, and wherein if the time representation is lesser than the predetermined value the time representation will indicate a short-term browsing history event.

15. The system of claim 14, wherein the predetermined value is adjustable.

16. The system of claim 14, wherein the processing circuit is further operable to:

increment the predetermined value by a unit of measurement.