System and method for disambiguating non-unique identifiers using information obtained from disparate communication channels

• US 8,166,168 B2
• Filed: 12/17/2007
• Issued: 04/24/2012
• Est. Priority Date: 12/17/2007
• Status: Active Grant

• Current Assignee: Verizon Patent and Licensing Incorporated (Verizon Communications Inc.)
• Original Assignee: Yahoo! Inc. (Apollo Global Management, Inc.)
• Inventors: Hayashi, Nathaniel Joseph, Boerries, Marco, Davis, Marc Eliot, Higgins, Christopher William, Martinez, Ronald, O'Sullivan, Joseph James, Trout, Robert Carter
• Primary Examiner(s): Walsh, John B.
• Application Number:
  US11/958,157
• Publication Number:
  US 20090157593A1
• Time in Patent Office: 1,590 Days
• Field of Search: None
• US Class Current: 709/226
• CPC Class Codes: G06Q 10/10 Office automation; Time management

Abstract

The disclosure describes systems and methods for disambiguating a non-unique identifiers of real world entities (RWEs) detected by a computing network by dynamically identifying relationships between RWEs known to the network. The relationships are determined based on social, spatial, temporal and logical information known about the RWEs based on the RWEs previous interactions with the network. These relationships are then used to generate a probability for each RWE that the non-unique identifier identifies that RWE. Based on the probabilities an RWE is selected and the proper network identifier may be used in place of the non-unique identifier.

Claims

1. A method comprising:

determining, via a computing device, that a non-unique username contains insufficient information to identify a first real-world entity (RWE), the non-unique username being associated with an information object (IO);

retrieving, via the computing device, data associated with the IO, the IO being associated with the first RWE, wherein retrieving data associated with the IO comprises;

retrieving at least one of social data, spatial data, temporal data and logical data associated with the IO; and

retrieving information identifying a second RWE associated with the IO;

generating, via the computing device, a probability for each of a plurality of RWEs including the first RWE and second RWE based on the data associated with the IO and data associated with the RWEs; and

selecting, via the computing device based on the generated probabilities, the first RWE as the RWE identified by the non-unique username, wherein said selection based upon a determination from the information that the non-unique username does not identify the second RWE.

2. The method of claim 1 further comprising:

for each of the plurality of RWEs, determining a relationship between each RWE and the second RWE based on the retrieved social data, spatial data, temporal data and logical data; and

generating the probability for each RWE based on the relationship between that RWE and the second RWE.

3. The method of claim 2, further comprising:

assigning each relationship a relative weight; and

generating the probability for each RWE based at least in part on the relative weight assigned to the relationship between that RWE and the second RWE.

4. The method of claim 2, wherein determining a relationship between one of the plurality of RWEs and the second RWE includes at least one of:

comparing social data, spatial data, temporal data and logical data associated with each RWE;

retrieving at least one known relationship between the one of the plurality of RWEs and the second RWE; and

identifying messages between the RWEs.

5. The method of claim 1 wherein the second RWE is a creator of the IO.

6. A method for identifying a real-world entity (RWE) associated with an information object (IO), the method comprising:

retrieving, via a computing device, IO data associated with the IO, the IO data including at least one of social data, spatial data, temporal data and logical data;

retrieving, via the computing device, RWE data for each of a plurality of RWEs, the RWE data including at least one of social data, spatial data, temporal data and logical data;

comparing, via the computing device, the RWE data and IO data, wherein comparing comprises;

generating a signature for each RWE, the signature including a set of time and location combinations associated with each RWE; and

comparing the IO data to the signature of each RWE;

generating, via the computing device, a probability for each of the plurality of RWEs based on results of the comparing operation; and

selecting, via the computing device based on the generated probabilities, at least one first RWE as the RWE associated with the IO.

7. The method of claim 6 wherein retrieving IO data further comprises:

identifying a second RWE as a creator of the IO; and

retrieving information identifying at least one of a time and location the IO was created.

8. The method of claim 7 further comprising:

for each of the plurality of RWEs, determining a relationship between that RWE and the second RWE based on the retrieved social data, spatial data, temporal data and logical data; and

generating that RWE'"'"'s probability based on the relationship between that RWE and the second RWE.

9. The method of claim 8, further comprising:

assigning each relationship a relative weight; and

generating each RWE'"'"'s probability based at least in part on the relative weight assigned to the relationship between that RWE and the second RWE.

10. The method of claim 8, wherein determining a relationship between one of the plurality of RWEs and the second RWE includes at least one of:

comparing social data, spatial data, temporal data and logical data associated with each RWE;

retrieving at least one previously-defined relationship between the one of the plurality of RWEs and the second RWE; and

identifying messages between the RWEs.

11. The method of claim 8 further comprising:

comparing the signatures of each RWE to the retrieved time and location the IO was created to identify third RWEs having signatures with time and location combinations near the retrieved time and location the IO was created, and

determining a relationship between each of the third RWEs and the second RWE based on the retrieved social data, spatial data, temporal data and logical data.

12. A system comprising:

a plurality of processors;

a correlation engine, implemented by at least one of said plurality of processors, connected via at least one communication network to a plurality of computing devices transmitting information objects (IOs) over the at least one communication network;

computer-readable media connected to the correlation engine storing at least one of social data, spatial data, temporal data and logical data associated with a plurality of real-world entities (RWEs);

wherein the correlation engine, based on the detection of a first IO containing a non-unique entity identifier, identifies one or more relationships between the first IO and the plurality of RWEs; and

wherein the social data, spatial data, temporal data and logical data associated with the plurality of RWEs are derived from the IOs transmitted over the at least one communication network and the relationships include at least one explicit relationship provided by either an owner of the first IO or the RWE assigning a relationship between the owner and the RWE; and

a disambiguation engine, implemented by at least one of said plurality of processors, that generates a probability for each of the plurality of RWEs based on the identified relationships, selects one of the plurality of RWEs as being identified by the non-unique entity identifier based on the probabilities generated, and associates the first IO with the selected one of the plurality of RWEs via associating a unique network identifier for the selected one of the plurality of RWEs with the first IO.

13. A computer-readable storage medium tangibly encoded with computer-executable instructions, that when executed by a computing device, perform a method, comprising:

dynamically identifying relationships between physical entities known to a communication network in response to a first receipt, from a first source, of a first non-unique identifier;

based on the identified relationships between the first source, a first physical entity and a second physical entity, using a first unique network identifier associated with the first physical entity in place of the first non-unique identifier; and

in response to a second receipt, from a second source, of the first non-unique identifier, using a second unique network identifier associated with a second physical entity in place of the first non-unique identifier based on the identified relationships between the second source, the first physical entity and the second physical entity.