Joint embedding of corpus pairs for domain mapping

• US 10,642,919 B2
• Filed: 08/18/2016
• Issued: 05/05/2020
• Est. Priority Date: 08/18/2016
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Jagmohan, Ashish, Khabiri, Elham, Segal, Richard B., Vaculin, Roman
• Primary Examiner(s): Phan, Tuankhanh D
• Application Number:
  US15/240,664
• Publication Number:
  US 20180052857A1
• Time in Patent Office: 1,356 Days
• Field of Search: None
• CPC Class Codes: G06F 16/24578 using ranking
• CPC Class Codes: G06F 16/958 Organisation or management of web site content, e.g. publishing, maintaining pages or automatic linking
• CPC Class Codes: G06N 20/00 Machine learning
• CPC Class Codes: G06N 3/045 Combinations of networks
• CPC Class Codes: G06N 3/084 Backpropagation, e.g. using gradient descent

Abstract

Techniques for outside-in mapping for corpus pairs are provided. In one example, a computer-implemented method comprises: inputting first keywords associated with a first domain corpus; extracting a first keyword of the first keywords; inputting second keywords associated with a second domain corpus; generating an embedded representation of the first keyword via a trained model and generating an embedded representation of the second keywords via the trained model; and scoring a joint embedding affinity associated with a joint embedding. The scoring the joint embedding affinity comprises: transforming the embedded representation of the first keyword and the embedded representation of the second keywords via the trained model; determining an affinity value based on comparing the first keyword to the second keywords; and based on the affinity value, aggregating the joint embedding of the embedded representation of the first keyword and the embedded representation of the second keywords within the second domain corpus.

Claims

1. A system, comprising:

a memory that stores computer executable components; and

a processor that executes the computer executable components stored in the memory, wherein the computer executable components comprise;

a first learning component that;

generates a first embedded representation of a profile term of first profile terms associated with a user identity profile, via a trained model, wherein the first terms are associated with the user identity profile of a first domain corpus;

a second learning component that;

generates a second embedded representation of second terms, associated with a second domain corpus, via the trained model;

an identification component that;

compares the profile term to the second terms to determine an affinity value based on a joint embedding of the first embedded representation and the second embedded representation, resulting in a comparison data;

maps the profile term to the second terms resulting in mapped data; and

outputs the mapped data to user identity profiles to facilitate aggregation, across the user identity profiles, of a scoring value associated with specific terms, wherein the scoring value comprises a term collection ranking in descending or ascending order, wherein the terms of the term collection ranking are ranked based on the scoring value;

a joint embedding component that;

based on the affinity value, generates display data associated with the comparison data for display via a webpage.

2. The system of claim 1, wherein the second terms are associated with a plurality of user identity profiles of the second domain corpus.

3. The system of claim 1, wherein the profile term of the first profile terms is associated with the web page.

4. The system of claim 1, wherein the affinity value is based on a frequency of an occurrence of the profile term within the second domain corpus.

5. The system of claim 1, wherein the affinity value is related to a scoring of the joint embedding of the comparison data.

6. The system of claim 2, wherein the computer executable components further comprise:

an aggregation component that;

outputs the first embedded representation and the second embedded representation to a display screen.

7. A computer program product for generating display data for a web page, the computer program product comprising a computer readable storage medium having program instructions embodied therewith, the program instructions executable to:

generate a first embedded representation of a profile term of the first terms, associated with a user identity profile of user identity profiles, via a trained model, wherein the first terms are associated with the user identity profile of a first domain corpus;

generate a second embedded representation of second terms, associated a second domain corpus, via the trained model;

compare the profile term to the second terms to determine an affinity value based on a joint embedding of the first embedded representation and the second embedded representation, resulting in a comparison data;

map the profile term to the second terms resulting in mapped data;

output the mapped data to user identity profiles to facilitate aggregation, across the user identity profiles, of a scoring value associated with specific terms, wherein the scoring value comprises a term collection ranking in descending or ascending order, wherein the terms of the term collection ranking is ranked based on the scoring value; and

based on the affinity value, generate display data associated with the comparison data for display by a webpage.

8. The computer program product of claim 7, wherein the program instructions are further executable to:

identify equivalent terms between the first domain corpus and the second domain corpus.

9. The computer program product of claim 7, wherein the program instructions are further executable to:

identify non-equivalent terms between the first domain corpus and the second domain corpus.

10. The computer program product of claim 8, wherein the program instructions are further executable to:

rank the equivalent terms in descending order based on a frequency of occurrence associated with the equivalent terms.

11. The computer program product of claim 10, wherein the comparison data comprises taxonomy tag data associated with a taxonomy tag of the equivalent terms.

12. The computer program product of claim 10, wherein the program instructions are further executable to:

output the display data to a display screen to facilitate marketing related to a user of the user identity profile.

13. The computer program product of claim 7, wherein the second terms are associated with a plurality of user identity profiles within the second domain corpus.

14. The computer program product of claim 7, wherein the affinity value is determined in response to an unsupervised estimation between the first domain corpus and the second domain corpus.

15. The computer program product of claim 7, wherein the program instructions are further executable to:

perform a cluster analysis of the first domain corpus relative to the second domain corpus.

16. A computer-implemented method, comprising:

generating, by a device operatively coupled to a processor, a first embedded representation of a profile term of the first terms, associated with a user identity profile, via a trained model, wherein the first terms are associated with the user identity profile of a first domain corpus;

generating, by the device, a second embedded representation of second terms, associated a second domain corpus, via the trained model;

comparing, by the device, the profile term to the second terms to determine an affinity value based on a joint embedding of the first embedded representation and the second embedded representation, resulting in a comparison data;

mapping, by the device, the profile term to the second terms resulting in mapped data;

outputting, by the device, the mapped data to user identity profiles to facilitate aggregation, across the user identity profiles, of a scoring value associated with specific terms, wherein the scoring value comprises a term collection ranking in ascending order, wherein the terms of the term collection ranking is ranked based on the scoring value;

based on the affinity value, generating, by the device, display data associated with the comparison data for display by a webpage; and

determining, by the device, affinity values for respective terms that occur at greater than a defined frequency across the user identity profile and one or more other user identity profiles.

17. The computer-implemented method of claim 16, wherein determining the affinity value comprises scoring, by the device, the first embedded representation and the second embedded representation.

18. The computer-implemented method of claim 16, further comprising:

transmitting, by the device, the display data to a display of a mobile device.

19. The computer-implemented method of claim 16, further comprising:

outputting, by the device, the first embedded representation and the second embedded representation to the second domain corpus.

20. The computer-implemented method of claim 16, further comprising:

generating, by the device, a profile term ranking associated with the first embedded representation in relation to the second embedded representation.