SYSTEM FOR DETERMINING AND OPTIMIZING FOR RELEVANCE IN MATCH-MAKING SYSTEMS

• US 20120005204A1
• Filed: 07/01/2010
• Published: 01/05/2012
• Est. Priority Date: 07/01/2010
• Status: Active Grant

• Current Assignee: Match Group, LLC (Match Group, Inc.)
• Original Assignee: Yahoo! Inc. (Apollo Global Management, Inc.)
• Inventors: Diaz, Fernando, Metzler, Donald, Amer-Yahia, Sihem
• Granted Patent: US 9,449,282 B2
• Time in Patent Office: Days
• US Class Current: 707/733
• CPC Class Codes: G06F 16/337 Profile generation, learning or modification
• CPC Class Codes: G06F 16/9535 Search customisation based on user profiles and personalisation
• CPC Class Codes: G06N 20/00 Machine learning
• CPC Class Codes: G06N 5/048 Fuzzy inferencing
• CPC Class Codes: G06N 7/01 Probabilistic graphical models, e.g. probabilistic networks

Abstract

Disclosed are methods and apparatus for automatically determining the relevance of matches between entities. A set of one or more indicators of relevance for each of a plurality of matches may be detected, where each of the plurality of matches exists between a first entity and a different one of a plurality of entities. Each set of one or more indicators of relevance indicates a degree of two-way interest for a corresponding one of the plurality of matches, the degree of two-way interest indicating both a degree of interest of the first entity in the corresponding one of the plurality of entities and a degree of interest of the corresponding one of the plurality of entities in the first entity. A probability of relevance of each of the plurality of matches may be determined based at least in part upon a corresponding set of one or more indicators of relevance. Each of the plurality of matches may be ranked based at least in part on the corresponding probability of relevance. A ranking function may be trained based upon the probability of relevance of each of the plurality of matches. The ranking function may subsequently be applied to identify and rank matches (e.g., in the absence of indicators of relevance).

Claims

1. A method, comprising:

detecting a set of one or more indicators of relevance for each of a plurality of matches, wherein each of the plurality of matches exists between a first entity and a different one of a plurality of entities, wherein each set of one or more indicators of relevance includes behavioral indicators that indicate a degree of two-way interest for a corresponding one of the plurality of matches, the degree of two-way interest indicating both a degree of interest of the first entity in the corresponding one of the plurality of entities and a degree of interest of the corresponding one of the plurality of entities in the first entity;

determining a probability of relevance of each of the plurality of matches based at least in part upon a corresponding set of one or more indicators of relevance; and

ranking the plurality of matches based at least in part on the corresponding probability of relevance.

2. The method as recited in claim 1, wherein determining a probability of relevance of each of the plurality of matches based at least in part upon a corresponding set of one or more indicators of relevance comprises:

determining a probability of relevance of each of the plurality of matches based in part on a query of the first entity, features of the first entity, and features of a corresponding one of the plurality of entities.

3. The method as recited in claim 2, wherein determining a probability of relevance of each of the plurality of matches based at least in part upon a corresponding set of one or more indicators of relevance comprises:

determining a probability of relevance of each of the plurality of matches based in part on match features for which values of a profile of the corresponding one of the plurality of entities matches the query of the first entity.

4. The method as recited in claim 2, wherein determining a probability of relevance of each of the plurality of matches based at least in part upon a corresponding set of one or more indicators of relevance comprises:

determining a probability of relevance of one of the plurality of matches based in part on match features for which values of a profile of the first entity matches a query of a corresponding one of the plurality of entities.

5. The method as recited in claim 1, wherein the set of one or more indicators of relevance for each of the plurality of matches measures interactions between the first entity and a corresponding one of the plurality of entities.

6. The method as recited in claim 1, wherein 1) the first entity is an individual and each of the plurality of entities is a different website;

2) the first entity is an individual and each of the plurality of entities is a different one of a plurality of potential employers;

     3) the first entity is an employer and each of the plurality of entities is a different one of a plurality of potential employees;

     4) the first entity is an individual and each of the plurality of entities is a different one of a plurality of individuals;

     5) the first entity is a consumer and each of the plurality of entities is a different one of a plurality of consumers;

or

     6) the first entity and each of the plurality of entities are members of a dating website.

7. A computer-readable medium storing thereon computer-readable instructions, comprising:

instructions for detecting a set of one or more indicators of relevance for each of a plurality of matches, wherein each of the plurality of matches exists between a first entity and a different one of a plurality of entities, wherein each set of one or more indicators of relevance indicates a degree of two-way appeal for a corresponding one of the plurality of matches, the degree of two-way appeal indicating a degree to which the corresponding one of the plurality of matches appeals to both the first entity and the corresponding one of the plurality of entities;

instructions for determining a probability of relevance of each of the plurality of matches based at least in part upon a corresponding set of one or more indicators of relevance; and

instructions ranking the plurality of matches based at least in part on the corresponding probability of relevance.

8. The computer-readable medium as recited in claim 7, wherein each set of one or more indicators includes behavioral features that do not provide click feedback from the first entity or a corresponding one of the plurality of entities.

9. The computer-readable medium as recited in claim 7, wherein each set of one or more indicators includes behavioral features pertaining to at least one of transmission of one or more messages or one or more profile views.

10. The computer-readable medium as recited in claim 7, wherein each set of one or more indicators of relevance:

1) identifies one-way interactions that do not have a corresponding two-way interaction between the first entity and a corresponding one of the plurality of entities;

or

     2) indicates a number of one-way interactions that do not have a corresponding two-way interaction between the first entity and a corresponding one of the plurality of entities.

11. The computer-readable medium as recited in claim 10, wherein the one-way interactions indicate at least one of a group consisting of:

1) the first entity sent a message to the corresponding one of the plurality of entities and the corresponding one of the plurality of entities did not send a reply to the first entity;

     2) the corresponding one of the plurality of entities sent a message to the first entity and the first entity did not send a reply to the corresponding one of the plurality of entities;

     3) the first entity viewed a profile of the corresponding one of the plurality of entities and the corresponding one of the plurality of entities did not view a profile of the first entity; and

     4) the corresponding one of the plurality of entities viewed a profile of the first entity and the first entity did not view a profile of the corresponding one of the plurality of entities.

12. The computer-readable medium as recited in claim 7, wherein each set of one or more indicators of relevance:

1) identifies two-way interactions between the first entity and a corresponding one of the plurality of entities;

or

     2) indicates a number of two-way interactions between the first entity and a corresponding one of the plurality of entities.

13. The computer-readable medium as recited in claim 12, wherein the two-way interactions comprise at least one of an exchange of phone numbers of both the first entity and the corresponding one of the plurality of entities, an exchange of electronic mail addresses of both the first entity and the corresponding one of the plurality of entities, or an exchange of information pertaining to a meeting between the first entity and a corresponding one of the plurality of entities.

14. The computer-readable medium as recited in claim 7, wherein the set of one or more indicators of relevance for each of the plurality of matches indicates a number of exchanges between the first entity and a corresponding one of the plurality of entities, wherein the exchanges include at least one of an exchange of messages or an exchange of profile views.

15. An apparatus, comprising:

a processor; and

a memory, at least one of the processor or the memory being adapted for;

detecting a set of one or more indicators of relevance for each of a plurality of matches, wherein each of the plurality of matches exists between a first entity and a different one of a plurality of entities, wherein each set of one or more indicators of relevance indicates a degree of two-way relevance for a corresponding one of the plurality of matches, the degree of two-way relevance indicating a degree to which the one of the plurality of matches is both relevant to the first entity and relevant to the corresponding one of the plurality of entities;

determining a probability of relevance of each of the plurality of matches based at least in part upon a corresponding set of one or more indicators of relevance; and

ranking the plurality of matches based at least in part on the corresponding probability of relevance.

16. The apparatus as recited in claim 15, wherein the set of one or more indicators of relevance for each of the plurality of matches indicates a difference in at least one of a group consisting of:

1) a number of messages sent by the first entity and a number of messages sent by a corresponding one of the plurality of entities; and

     2) a number of times that the first entity viewed a profile of a corresponding one of the plurality of entities and a number of times that the corresponding one of the plurality of entities viewed a profile of the first entity.

17. The apparatus as recited in claim 15, wherein the set of one or more indicators of relevance for each of the plurality of matches indicates at least one of a group consisting of:

1) a time period during which the first entity and a corresponding one of the plurality of entities exchange messages; and

     2) a time period during which the first entity and the corresponding one of the plurality of entities exchange profile views.

18. The apparatus as recited in claim 15, wherein the set of one or more indicators of relevance for each of the plurality of matches indicates at least one of a group consisting of:

1) a density of messages exchanged between the first entity and a corresponding one of the plurality of entities after a first message has been exchanged between the first entity and the corresponding one of the plurality of entities; and

     2) a density of profile views exchanged by the first entity and the corresponding one of the plurality of entities after the first message has been exchanged between the first entity and the corresponding one of the plurality of entities.

19. The apparatus as recited in claim 15, wherein the set of one or more indicators of relevance for each of the plurality of matches indicates at least one of a group consisting of:

1) the first entity viewed a profile of a corresponding one of the plurality of entities but did not send a message to the one of the plurality of entities; and

     2) the one of the plurality of entities viewed a profile of the first entity but did not send a message to the first entity.

20. The apparatus as recited in claim 15, wherein ranking each of the plurality of matches based in part on the corresponding probability of relevance comprises:

ascertaining one or more parameters of a ranking model using the probability of relevance of each of the plurality of matches.

21. The apparatus as recited in claim 15, at least one of the processor or the memory being further adapted for:

training a ranking model based at least in part on the probability of relevance of each of the plurality of matches to predict features that correlate with relevance and/or identify feature importance of the features that correlate with relevance.

22. The apparatus as recited in claim 21, at least one of the processor or the memory being further adapted for:

applying the trained ranking model to identify and rank matches for a different entity.

23. The apparatus as recited in claim 22, wherein the matches for the different entity do not have corresponding sets of indicators of relevance.

24. The apparatus as recited in claim 15, at least one of the processor or the memory being further adapted for:

training a ranking model based at least in part on the probability of relevance of each of the plurality of matches to predict features that correlate with two-way relevance and/or identify feature importance of the features that correlate with two-way relevance.