Inferring attributes of organizations using member graph

Patent metadata

Field	Value
Publication number	US-10769426-B2
Application number	US-201514929128-A
Country	US
Kind code	B2
Filing date	Oct 30, 2015
Priority date	Sep 30, 2015
Publication date	Sep 8, 2020
Grant date	Sep 8, 2020

Abstract

In an example embodiment, a member profile corresponding to a member of a social networking service is obtained. Usage information for the member is then obtained, and one or more member metrics are calculated based on the member profile and usage information for the corresponding member. A plurality of features are extracted from the member profile and the one or more member metrics. The plurality of features is inserted into an organization name confidence score model to obtain a confidence score for an organization name in the member profile.

First claim

What is claimed is:

1. A computer-implemented method comprising;
obtaining, by a processor, a plurality of sample member profiles, each sample member profile corresponding to a sample member of a social networking service;
for each of the plurality of sample member profiles:
obtaining, by the processor, usage information for a corresponding sample member;
calculating, by the processor, one or more member metrics based on the sample member profile and usage information for the corresponding sample member;
extracting, by the processor, a first plurality of features from the sample member profile and the one or more member metrics;
feeding, by the processor, the first plurality of features into a supervised machine learning organization confidence score algorithm to train an organization name confidence score model to calculate a confidence score for a particular member profile indicating a probability that an organization name in the particular member profile is accurate, the supervised machine learning organization confidence score algorithm executed by the computer to implement a supervised machine learning classifier;
obtaining, by the processor, a member profile corresponding to a member of a social networking service;
obtaining, by the processor, usage information for the member;
calculating, by the processor, one or more member metrics based on the member profile and usage information for the corresponding member;
extracting, by the processor, a second plurality of features from the member profile and the one or more member metrics;
inputting, by the processor, the second plurality of features into the organization name confidence score model to obtain a confidence score for an organization name in the member profile.

2. The method of claim 1 , further comprising:
based on the confidence score for the organization name in the member profile, combining a first organization record corresponding to the organization name with a second organization record.

3. The method of claim 1 , further comprising:
based on the confidence score for the organization name in the member profile, adding the organization name to an organization record that is missing an organization name, the organization record having at least one field matching a field in the member profile.

4. The method of claim 1 , wherein the one or more member metrics include how frequently a member accesses the social networking service.

5. The method of claim 1 , wherein the one or more member metrics include how frequently a member updates a member profile on the social networking service.

6. The method of claim 1 , wherein the one or more member metrics include how frequently a member communicates with other members via the social networking service.

7. The method of claim 1 , wherein the first plurality of features and the second plurality of features include a power user score calculated based on the one or more member metrics.

8. The method of claim 1 , wherein the first plurality of features and the second plurality of features include a connection density score calculated on based on connections in the member profile, the connection density score indicating a level at which the member is connected to other members having member profiles in a group with the member profile.

9. The method of claim 1 , wherein the first plurality of features and the second plurality of features include an email connection density score calculated based on an email address in the member profile, the connection density score indicating a level at which the email address contains a domain that is shared with email addresses of other members having member profiles in a group with the member profile.

10. A system comprising:
a processor;
a computer-readable medium having instructions stored there on, which, when executed by the processor, cause the system to perform operations comprising:
obtaining a plurality of sample member profiles, each sample member profile corresponding to a sample member of a social networking service;
for each of the plurality of sample member profiles:
obtaining usage information for a corresponding sample member;
calculating one or more member metrics based on the sample member profile and usage information for the corresponding sample member;
extracting a first plurality of features from the sample member profile and the one or more member metrics;
feeding, using a computer, the first plurality of features into a supervised machine learning organization confidence score algorithm to train an organization name confidence score model to calculate a confidence score for a particular member profile indicating a probability that an organization name in the particular member profile is accurate, the supervised machine learning organization confidence score algorithm executed by the computer to implement a supervised machine learning classifier;
obtaining a member profile corresponding to a member of a social networking service;
obtaining usage information for the member;
calculating one or more member metrics based on the member profile and usage information for the corresponding member;
extracting a second plurality of features from the member profile and the one or more member metrics;
inputting the second plurality of features into the organization name confidence score model to obtain a confidence score for an organization name in the member profile.

11. The system of claim 10 , wherein the operations further comprise:
based on the confidence score for the organization name in the member profile, combining a first organization record corresponding to the organization name with a second organization record.

12. The system of claim 10 , wherein the operations further comprise:
based on the confidence score for the organization name in the member profile, adding the organization name to an organization record that is missing an organization name, the organization record having at least one field matching a field in the member profile.

13. The system of claim 10 , wherein the one or more member metrics include how frequently a member accesses the social networking service.

14. The system of claim 10 , wherein the one or more member metrics include how frequently a member updates a member profile on the social networking service.

15. The system of claim 10 , wherein the one or more member metrics include how frequently a member communicates with other members via the social networking service.

16. The system of claim 10 , wherein the first plurality of features and the second plurality of features include a power user score calculated based on the one or more member metrics.

17. The system of claim 10 , wherein the first plurality of features and the second plurality of features include a connection density score calculated based on connections in the member profile, the connection density score indicating a level at which the member is connected to other members having member profiles in a group with the member profile.

18. The system of claim 10 , wherein the first plurality of features and the second plurality of features include an email connection density score calculated on based on an email address in the member profile, the connection density score indicating a level at which the email address contains a domain that is shared with email addresses of other members having member profiles in a group with the member profile.

19. A non-transitory machine-readable storage medium comprising instructions, which when implemented by one or more ma

Assignees

• Microsoft Technology Licensing Llc

Inventors

• Guo Songtao
• Wang Ke
• Lai Alex Ching
• Kumar Aarti
• Tsang Keith Wai Kit
• Thakur Rekha
• Lin Song
• Degiere Christopher Matthew

Classifications

• G06F16/215 Primary

  Improving data quality; Data cleansing, e.g. de-duplication, removing invalid entries or correcting typographical errors · CPC title

• G06V30/19167

  Active pattern learning · CPC title

• G06V30/19173

  Classification techniques · CPC title

• G06V30/19147

  Obtaining sets of training patterns; Bootstrap methods, e.g. bagging or boosting · CPC title

• G06V10/761

  Proximity, similarity or dissimilarity measures · CPC title