Machine learned models for contextual editing of social networking profiles

Patent metadata

Field	Value
Publication number	US-10678997-B2
Application number	US-201715825657-A
Country	US
Kind code	B2
Filing date	Nov 29, 2017
Priority date	Oct 5, 2017
Publication date	Jun 9, 2020
Grant date	Jun 9, 2020

Abstract

In an example, first and second machine learned models corresponding to a particular context of a social networking service are obtained, the first machine learned model trained via a first machine learning algorithm to output an indication of importance of a social networking profile field to obtaining results in the particular context, and the second machine learned model trained via a second machine learning algorithm to output a propensity of the user to edit a social networking profile field if requested. One or more missing fields in a social networking profile for the user are identified. For each of one or more of the one or more missing fields, the field and an identification of the user are passed through the first and second machine learned models, and outputs of the first and second machine learned models are combined to identify one or more top missing profile fields.

First claim

What is claimed is:

1. A system comprising:
a non-transitory computer-readable medium having instructions stored thereon, which, when executed by a processor, cause the system to:
train a first machine learned model, by extracting a first set of features from a plurality of sample member profiles and passing the first set of features along with sample field value labels, to output an indication of importance for a member profile field;
train a second machine learned model, by extracting a second set of features from a plurality of sample member profiles and sample member activity and behavior information and passing the second set of features along with sample edit propensity labels, to output a propensity of users to edit the member profile field if requested, the training of the second machine learned model performed independently of the training of the first machine learned model;
receive, via a user interface of a computer system, one or more operations initiated by a user;
perform the one or more operations on a social networking service to generate a first set of one or more results in a particular context of the social networking service, the particular context based on the one or more operations;
display at least a portion of the first set of the one or more results to the user via the user interface;
identify one or more missing fields in a social networking profile for the user;
for each of one or more of the one or more missing fields, pass the field and an identification of the user through the first machine learned model, producing a first probability as output pass the field and the identification of the user through the second machine learned model, producing a second probability as output and combine the first and second probabilities outputs of the first and second machine learned models;
identify one or more top missing profile fields based on the combined probabilities for each of the one or more missing fields;
prompt the user to add information for the identified one or more top missing profile fields;
in response to receiving information for at least one of the identified one or more top missing profile fields, generate a second set of one or more results in the particular context; and
display at least a portion of the first set of the one or more results to the user via the user interface.

2. The system of claim 1 , wherein the first machine learning algorithm is a logistic regression algorithm.

3. The system of claim 1 , wherein the second machine learning algorithm is a logistic regression algorithm.

4. The system of claim 1 , wherein the first set of features is further extracted from sample social graph and sample member usage and activity information.

5. The system of claim 4 , wherein the first feature data includes member network size and member tenure on the social networking service.

6. The system of claim 5 , wherein the first feature data further includes member recent feed, job search, and messaging activity level and member social network profile completion information.

7. The system of claim 1 , wherein the instructions further cause the system to:
obtain a third machine learned model corresponding to the particular context, the third machine learned model trained via a third machine learning algorithm to output an indication of whether the user is having a positive experience with the first set of the one or more results for the particular context.

8. A computerized method comprising:
training a first machine learned model, by extracting a first set of features from a plurality of sample member profiles and passing the first set of features along with sample field value labels, to output an indication of importance for a member profile field;
training a second machine learned model, by extracting a second set of features from a plurality of sample member profiles and sample member activity and behavior information and passing the second set of features along with sample edit propensity labels, to output a propensity of users to edit the member profile field if requested, the training of the second machine learned model performed independently of the training of the first machine learned model;
receive, via a user interface of a computer system, one or more operations initiated by a user;
performing the one or more operations on a social networking service to generate a first set of one or more results in a particular context of the social networking service, the particular context based on the one or more operations;
displaying at least a portion of the first set of the one or more results to the user via the user interface;
identifying one or more missing fields in a social networking profile for the user; for each of one or more of the one or more missing fields, pass the field and an identification of the user through the first machine learned model, producing a first probability as output, pass the field and the identification of the user through the second machine learned model, producing a second probability as output and combine the first and second probabilities;
identifying one or more top missing profile fields based on the combined probabilities for each of the one or more missing fields;
prompting the user to add information for the identified one or more top missing profile fields;
in response to receiving information for at least one of the identified one or more top missing profile fields, generating a second set of one or more results in the particular context; and
displaying at least a portion of the first set of the one or more results to the user via the user interface.

9. The computerized method of claim 8 , wherein the first machine learning algorithm is a logistic regression algorithm.

10. The computerized method of claim 8 , wherein the second machine learning algorithm is a logistic regression algorithm.

11. The computerized method of claim 8 , wherein the first set of features is further extracted from sample social graph and sample member usage and activity information.

12. The computerized method of claim 11 , wherein the first feature data includes member network size and member tenure on the social networking service.

13. The computerized method of claim 12 , wherein the first feature data further includes member recent feed, job search, and messaging activity level and member social network profile completion information.

14. The computerized method of claim 8 , further comprising:
obtaining a third machine learned model corresponding to the particular context, the third machine learned model trained via a third machine learning algorithm to output an indication of whether the user is having a positive experience with the first set of the one or more results for the particular context.

15. A non-transitory machine storage medium comprising instructions, which when implemented by one or more machines, cause the one or more machines to perform operations comprising:
training a first machine learned model, by extracting a first set of features from a plurality of sample member profiles and passing the first set of features along with sample field value labels, to output an indication of importance for a member profile field;
training a second machine learned model, by extracting a second set of features from a plurality of sample member profiles and sample member activity and behavior information and passing the second set of features along with sample edit propensity labels, to output a propensity of users to edit the member profile field if requested, the training of the second machine learned model performed independently of t

Assignees

• Microsoft Technology Licensing Llc

Inventors

• Ahuja Karan Ashok
• Ayenew Ejigou Befekadu
• Liang Ningfeng
• Bajaj Lokesh P
• Wang Wei
• Fletcher Paul
• Lu Wei
• Chatterjee Shaunak
• Ghosh Souvik
• Li Yang
• Deng Wei
• Wu Qiang

Classifications

• H04L67/22

  Electricity · mapped topic

• G06F40/174 Primary

  Form filling; Merging · CPC title

• G06Q50/01

  Physics · mapped topic

• G06N20/00

  Machine learning · CPC title

• H04L67/306

  User profiles · CPC title