SYSTEMS AND METHODS TO SEARCH RESUMES BASED ON KEYWORDS

• US 20170193089A1
• Filed: 01/04/2016
• Published: 07/06/2017
• Est. Priority Date: 01/04/2016
• Status: Active Grant

• Current Assignee: Meta Platforms, Inc. (f/k/a Facebook, Inc.)
• Original Assignee: Meta Platforms, Inc. (f/k/a Facebook, Inc.)
• Inventors: Fang, Miaoqing
• Granted Patent: US 10,572,519 B2
• Time in Patent Office: Days
• CPC Class Codes: G06F 16/3344 using natural language analysis
• CPC Class Codes: G06F 16/3347 using vector based model
• CPC Class Codes: G06F 16/353 into predefined classes
• CPC Class Codes: G06Q 10/105 Human resources

Abstract

Systems, methods, and non-transitory computer readable media are configured to convert resume text in a resume into an array of values representing a frequency of keywords associated with the resume text. An array of values representing a frequency of search terms associated with a search is generated. The array of values representing a frequency of keywords associated with the resume text and the array of values representing a frequency of search terms associated with a search to generate a score for the resume are combined.

Claims

1. A computer-implemented method comprising:

converting, by a computing system, resume text in a resume into an array of values representing a frequency of keywords associated with the resume text;

generating, by the computing system, an array of values representing a frequency of search terms associated with a search; and

combining, by the computing system, the array of values representing a frequency of keywords associated with the resume text and the array of values representing a frequency of search terms associated with a search to generate a score for the resume.

2. The computer-implemented method of claim 1, wherein the converting resume text further comprises:

training a model based on terms from a resume corpus to create a vector space of vector representations of the terms; and

identifying keywords as anchor points in the vector space for a set of search terms used by recruiters.

3. The computer-implemented method of claim 2, wherein the converting resume text further comprises:

creating a chunk of terms from the resume text based on a selected chunk size and a selected increment value; and

identifying a vector representation in the vector space for each term in the chunk.

4. The computer-implemented method of claim 3, wherein the converting the resume text further comprises:

combining the vector representations for each term in the chunk to generate a vector representation of the chunk.

5. The computer-implemented method of claim 4, wherein the converting the resume text further comprises:

identifying an anchor point constituting a keyword that is nearest to the vector representation of the chunk.

6. The computer-implemented method of claim 5, wherein the array of values representing a frequency of keywords associated with the resume text comprises values associated with a frequency of keywords identified for a plurality of chunks from the resume text.

7. The computer-implemented method of claim 6, wherein the converting the resume text further comprises:

normalizing the array of values representing a frequency of keywords associated with the resume text based on relative importance of the keywords.

8. The computer-implemented method of claim 7, wherein the array of values representing a frequency of keywords associated with the resume text is a row in a resume matrix and the array of values representing a frequency of search terms associated with a search is a column of a search term matrix.

9. The computer-implemented method of claim 8, wherein the combining the array of values comprises:

multiplying the resume matrix and the search term matrix to generate a score matrix.

10. The computer-implemented method of claim 9, wherein each column of the score matrix reflects resume scores for an associated search.

11. A system comprising:

at least one processor; and

a memory storing instructions that, when executed by the at least one processor, cause the system to perform;

converting resume text in a resume into an array of values representing a frequency of keywords associated with the resume text;

generating an array of values representing a frequency of search terms associated with a search; and

combining the array of values representing a frequency of keywords associated with the resume text and the array of values representing a frequency of search terms associated with a search to generate a score for the resume.

12. The system of claim 11, wherein the converting resume text further comprises:

training a model based on terms from a resume corpus to create a vector space of vector representations of the terms; and

identifying keywords as anchor points in the vector space for a set of search terms used by recruiters.

13. The system of claim 12, wherein the converting resume text further comprises:

creating a chunk of terms from the resume text based on a selected chunk size and a selected increment value; and

identifying a vector representation in the vector space for each term in the chunk.

14. The system of claim 13, wherein the converting resume text further comprises:

combining the vector representations for each term in the chunk to generate a vector representation of the chunk.

15. The system of claim 14, wherein the converting resume text further comprises:

identifying an anchor point constituting a keyword that is nearest to the vector representation of the chunk.

16. A non-transitory computer-readable storage medium including instructions that, when executed by at least one processor of a computing system, cause the computing system to perform a method comprising:

converting resume text in a resume into an array of values representing a frequency of keywords associated with the resume text;

generating an array of values representing a frequency of search terms associated with a search; and

combining the array of values representing a frequency of keywords associated with the resume text and the array of values representing a frequency of search terms associated with a search to generate a score for the resume.

17. The non-transitory computer-readable storage medium of claim 16, wherein the converting resume text further comprises:

training a model based on terms from a resume corpus to create a vector space of vector representations of the terms; and

identifying keywords as anchor points in the vector space for a set of search terms used by recruiters.

18. The non-transitory computer-readable storage medium of claim 17, wherein the converting resume text further comprises:

creating a chunk of terms from the resume text based on a selected chunk size and a selected increment value; and

identifying a vector representation in the vector space for each term in the chunk.

19. The non-transitory computer-readable storage medium of claim 18, wherein the converting resume text further comprises:

combining the vector representations for each term in the chunk to generate a vector representation of the chunk.

20. The non-transitory computer-readable storage medium of claim 16, wherein the converting resume text further comprises:

identifying an anchor point constituting a keyword that is nearest to the vector representation of the chunk.