SKILL EXTRACTION SYSTEM

• US 20120197863A1
• Filed: 01/24/2012
• Published: 08/02/2012
• Est. Priority Date: 01/27/2011
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Linkedln Corporation
• Inventors: Skomoroch, Peter N., Hayes, Matthew T., Gupta, Abhishek, Patil, Dhanurjay A.S.
• Granted Patent: US 8,650,177 B2
• Time in Patent Office: Days
• US Class Current: 707/710
• CPC Class Codes: G06F 16/285 Clustering or classification
• CPC Class Codes: G06F 16/335 Filtering based on additional data, e.g. user or group profiles filtering in web context G06F16/9535, G06F16/9536
• CPC Class Codes: G06F 16/35 Clustering; Classification
• CPC Class Codes: G06F 16/355 Class or cluster creation or modification
• CPC Class Codes: G06F 16/9535 Search customisation based on user profiles and personalisation
• CPC Class Codes: G06Q 10/06398 Performance of employee with respect to a job function
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising
• CPC Class Codes: G06Q 30/0269 based on user profile or attribute
• CPC Class Codes: G06Q 50/01 Social networking

Abstract

In an example, disclosed is a machine automated method of identifying a set of skills. In some examples, the method includes extracting a plurality of skill seed phrases from a plurality of member profiles of a social networking site, creating a plurality of disambiguated skill seed phrases by disambiguating the plurality of skill seed phrases using one or more computer processors, and de-duplicating the plurality of disambiguated skill seed phrases to create a plurality of de-duplicated skill seed phrases.

Claims

1. A method comprising:

extracting a plurality of skill seed phrases from a plurality of member profiles of a social networking service;

using one or more computer processors, disambiguating the plurality of skill seed phrases to create a plurality of disambiguated skill seed phrases; and

de-duplicating the plurality of disambiguated skill seed phrases to create a plurality of de-duplicated skill seed phrases, the de-duplicated skill seed phrases identifying a plurality of skills.

2. The method of claim 1, wherein extracting the plurality of skill seed phrases comprises:

tokenizing a portion of the plurality of member profiles into a plurality of tokens; and

analyzing frequency information of the tokens to determine the plurality of skill seed phrases.

3. The method of claim 2, wherein extracting the plurality of skill seed phrases further comprises:

removing skill seed phrases that do not correspond with actual skills.

4. The method of claim 1, wherein creating a plurality of disambiguated skill seed phrases comprises:

clustering the plurality of skill seed phrases based on a frequency of co-occurrence of the plurality of skill seed phrases.

5. The method of claim 4, wherein the clustering is done using an expectation maximum method.

6. The method of claim 1, further comprising:

calculating attributes about the plurality of disambiguated skill seed phrases and storing the attributes in association with the plurality of disambiguated skill seed phrases.

7. The method of claim 6, wherein gathering attributes about the plurality of disambiguated skill phrases comprises computing a top industry of each of the plurality of disambiguated skill phrases based upon an industry listed in member profiles in which each skill seed phrase was found.

8. The method of claim 1, wherein de-duplicating the plurality of disambiguated skill seed phrases comprises:

calculating a plurality of internet search queries, at least one for each of the plurality of disambiguated skill seed phrases, the internet search queries, when executed, returning a list of internet web pages;

for each of the plurality of internet search queries, submitting to a crowd sourcing application, over a network, a job for which the crowd sourcing application returns a specific internet web page selected from the list of internet web pages that is returned from each of the plurality of the internet search queries, the specific internet web page representing an interpretation of a meaning of the disambiguated skill seed phrase corresponding to each internet search query; and

using the plurality of returned specific internet web pages to determine which of the plurality of disambiguated skill seed phrases identify identical skills.

9. The method of claim 8, further comprising:

submitting more than one job for each of the plurality of internet search queries and using an algorithm to select among the returned specific internet web pages to determine which of the returned specific internet web pages for each internet search query is used to determine which of the plurality of disambiguated skill seed phrases identify the identical skills.

10. A system, comprising:

an extraction module configured to extract a plurality of skill seed phrases from a plurality of member profiles of a social networking service;

a disambiguation module configured to disambiguating the plurality of skill seed phrases to create a plurality of disambiguated skill seed phrases; and

a de-duplication module configured to de-duplicate the plurality of disambiguated skill seed phrases to create a plurality of de-duplicated skill seed phrases, the plurality of de-duplicated skill seed phrases identifying a plurality of skills.

11. The system of claim 10, wherein the extraction module:

tokenizes a portion of the plurality of member profiles into a plurality of tokens and analyzes frequency information of the tokens to determine the plurality of skill seed phrases.

12. The system of claim 10, wherein the extraction module removes skill seed phrases that do not correspond with actual skills.

13. The system of claim 10, wherein the disambiguation comprises:

clustering the plurality of skill seed phrases based on a frequency of co-occurrence of the plurality of skill seed phrases.

14. The system of claim 13, wherein the clustering is performed by an expectation maximum operation.

15. The system of claim 10, wherein the disambiguation module is further configured to calculate attributes about the plurality of disambiguated skill seed phrases and store the attributes in association with the plurality of disambiguated skill seed phrases.

16. The system of claim 15, wherein the attributes about the plurality of disambiguated skill seed phrases includes a top industry of at least one of the plurality of disambiguated skill phrases computed based upon an industry listed in the member profiles in which the at least one of the plurality of disambiguated skill seed phrases was found.

17. The system of claim 10, wherein the de-duplication module:

calculates a plurality of internet search queries, at least one for each of the disambiguated skill seed phrases, the internet search queries, when executed, returning a list of internet web pages;

for each of the plurality of internet search queries, submits to a crowd sourcing application over a network, a job for which the crowd sourcing application returns a specific internet web page selected from the list of internet web pages that is returned from each of the plurality of the internet search queries, the specific internet web page representing an interpretation of a meaning of the disambiguated skill seed phrase corresponding to each internet search query; and

using the plurality of returned specific internet web pages, determines which of the plurality of disambiguated skill seed phrases identify identical skills.

18. The system of claim 17, wherein the de-duplication module submits multiple jobs for each of the plurality of disambiguated skill seed phrases and uses an algorithm to select among the returned specific internet web pages to determine which of the returned specific internet web pages for that internet search query is used to determine which of the plurality of disambiguated skill seed phrases identify identical skills.

19. A machine readable storage medium including instructions, which when executed on the machine, cause the machine to:

extract a plurality of skill seed phrases from a plurality of member profiles of a social networking service;

disambiguate the plurality of skill seed phrases to create a plurality of disambiguated skill seed phrases; and

de-duplicate the plurality of disambiguated skill seed phrases to create a plurality of de-duplicated skill seed phrases, the de-duplicated skill seed phrases identifying a plurality of skills.

20. The machine readable medium of claim 19, wherein extracting the plurality of skill seed phrases comprises:

tokenizing a portion of the plurality of member profiles into a plurality of tokens; and

analyzing frequency information of the tokens to determine the plurality of skill seed phrases.

21. The machine readable medium of claim 20, wherein extracting the plurality of skill seed phrases further comprises:

removing skill seed phrases that do not correspond with actual skills.

22. The machine readable medium of claim 19, wherein creating a plurality of disambiguated skill seed phrases comprises:

clustering the plurality of skill seed phrases based on a frequency of co-occurrence of the plurality of skill seed phrases.

23. The machine readable medium of claim 22, wherein the clustering is done using an expectation maximum method.

24. The machine readable medium of claim 19, wherein the instructions further comprise instructions which when executed by the machine cause the machine to:

calculate attributes about the plurality of disambiguated skill seed phrases and store the attributes in association with the plurality of disambiguated skill seed phrases.

25. The machine readable medium of claim 24, wherein gathering attributes about the plurality of disambiguated skill phrases comprises computing a top industry of each of the plurality disambiguated skill phrases based upon an industry listed in member profiles in which each skill seed phrase was found.

26. The machine readable medium of claim 19, wherein de-duplication comprises:

calculating a plurality of internet search queries, at least one for each of the plurality of disambiguated skill seed phrases, the internet search queries, when executed returning a list of internet web pages;

for each of the plurality of internet search queries, submitting to a crowd sourcing application, over a network, a job for which the crowd sourcing application returns a specific internet web page selected from the list of internet web pages that is returned from each of the plurality of the internet search queries, the specific internet web page representing an interpretation of a meaning of the disambiguated skill seed phrase corresponding to each internet search query; and

using the plurality of returned specific internet web pages to determine which of the plurality of disambiguated skill seed phrases identify identical skills.

27. The machine readable medium of claim 26, wherein de-duplication further comprises:

submitting more than one job for each of the plurality of internet search queries and using an algorithm to select among the returned specific internet web pages to determine which of the returned specific internet web pages for each internet search query is used to determine which of the plurality of disambiguated skill seed phrases identify the identical skills.