Suggesting tags on online social networks

Patent metadata

Field	Value
Publication number	US-10467282-B2
Application number	US-201514973673-A
Country	US
Kind code	B2
Filing date	Dec 17, 2015
Priority date	Dec 17, 2015
Publication date	Nov 5, 2019
Grant date	Nov 5, 2019

Abstract

In one embodiment, a method includes accessing one or more content objects associated with a first user of an online social network; generating a set of tag candidates by extracting one or more n-grams from the content objects, wherein each tag candidate comprises one or more text strings or ideograms from the extracted n-grams; calculating, for each tag candidate of the set of tag candidates, a quality-score for the tag candidate based on an affinity coefficient between the first user and a respective content object from which the n-grams or ideograms comprising the tag candidate were extracted; and sending, to a client system for display, one or more of the tag candidates as suggested tags for the first user, wherein each suggested tag has a quality-score greater than a threshold quality-score.

First claim

What is claimed is:

1. A method comprising, by one or more computing systems of an online social network:
accessing, by one or more of the computing systems, one or more content objects posted to the online social network that a first user of the online social network has interacted with on the online social network, wherein each content object is associated with one or more concepts;
generating, by one or more of the computing systems, a set of tag candidates for a profile page of the first user by extracting one or more n-grams from the content objects, wherein each tag candidate comprises one or more text strings or ideograms from the extracted n-grams, wherein the profile page of the first user identifies the first user on the online social network and comprises a name of the first user and one or more of the content objects that the first user has interacted with;
calculating, by one or more of the computing systems, for each tag candidate of the set of tag candidates, a quality-score for the tag candidate based on an affinity coefficient between the first user and a respective content object from which the n-grams or ideograms comprising the tag candidate were extracted, wherein the affinity coefficient between the first user and the respective content object is based on one or more user actions taken by the first user with respect to the one or more concepts associated with the respective content object;
sending, by one or more of the computing systems, to a client system of a second user of the online social network for display, one or more of the tag candidates as suggested tags for the profile page of the first user, wherein each suggested tag has a quality-score greater than a threshold quality-score;
receiving, by one or more of the computing systems, an indication that the second user has selected one or more of the suggested tags to tag the profile page of the first user; and
sending, by one or more of the computing systems to a client system of the first user, a notification to the first user that the first user's profile page has pending tags for review.

2. The method of claim 1 , further comprising:
accessing a social graph comprising a plurality of nodes and a plurality of edges connecting the nodes, each of the edges between two of the nodes representing a single degree of separation between them, the nodes comprising:
a first node corresponding to a first user associated with the online social network; and
a plurality of second nodes corresponding to a plurality of content objects associated with the online social network, respectively.

3. The method of claim 2 , wherein the quality-score for each tag candidate is further based on a degree of separation between the first node corresponding to the first user and the second node corresponding to the content object from which the n-grams or ideogram comprising the tag candidate were extracted.

4. The method of claim 1 , wherein generating the set of tag candidates comprises extracting one or more n-grams or ideograms from a set of posts authored by the first user.

5. The method of claim 1 , wherein generating the set of tag candidates comprises extracting one or more n-grams or ideograms from a set of posts accessed by the first user.

6. The method of claim 5 , wherein the set of posts accessed by the first user comprises posts authored by one or more second users within a threshold degree of separation of the first user in a social graph and viewed by the first user.

7. The method of claim 1 , wherein the suggested tags for the first user comprise one or more object-tags associated with one or more friends of the first user within the online social network.

8. The method of claim 1 , further comprising:
for one or more of the suggested tags, sending, to the client system of the second user of the online social network for display, references to one or more content objects that are tagged with the suggested tag.

9. The method of claim 8 , further comprising:
receiving a selection of a reference to one of the content objects; and
sending, to the client system of the second user of the online social network for display, a profile page associated with the content object corresponding to the selected reference.

10. The method of claim 1 , further comprising:
indexing the one or more n-grams extracted from the content objects;
identifying n-grams having the same root element; and
categorizing n-grams having the same root element under the same tag candidate.

11. The method of claim 1 , further comprising:
compiling a list of blacklisted tags, wherein tag candidates on the list of blacklisted tags are not sent as suggested tags.

12. The method of claim 1 , further comprising:
receiving, from the client system, a text query;
parsing the text query to identify one or more n-grams;
searching the set of tag candidates to identify one or more tag candidates that match the text query; and
sending, to the client system, one or more of the matched tag candidates in ranked order based on the affinity coefficients between a user and the content object associated with each matched tag candidate.

13. The method of claim 12 , further comprising:
suggesting an alternative query if the text query results in a number of matched tag candidates that is below a predetermined number of matches.

14. The method of claim 1 , wherein generating the set of tag candidates comprises:
calculating, for each extracted n-gram, a term-score based on a term-frequency inverse document frequency (TF-IDF) analysis of the n-gram;
filtering each n-gram having a term-score below a predetermined value;
accessing a tag database of user-generated tags that have been generated by any user of the online social network;
comparing each n-gram having a term-score greater than or equal to the predetermined value to the user-generated tags in the tag database; and
selecting as tag candidates one or more of the n-grams matching the user-generated tags in the tag database.

15. The method of claim 14 , wherein generating the set of tag candidates further comprises:
filtering each n-gram matching a tag listed in a list of blacklisted tags from the set of tag candidates.

16. The method of claim 1 , further comprising:
filtering the set of tag candidates by comparing each tag candidate to a set of user-generated tags that have been generated by any user of the online social network to identify tag candidates that match the user-generated tags, wherein tag candidates that do not match at least one of the user-generated tags are filtered from the set of tag candidates.

17. The method of claim 1 , wherein the quality-score for a given tag candidate is further based on one or more user interactions with suggested tags that were previously sent to the first user.

18. The method of claim 1 , further comprising:
generating a tag dashboard, wherein the tag dashboard comprises one or more pending tags associated with the first user, wherein each pending tag is a tag candidate selected by a user of the online social network; and
receiving, from the first user, an indication of an approval or a rejection of the one or more pending tags.

19. One or more computer-readable non-transitory storage media of an online social network embodying software that is operable when executed to:
access one or more content objects posted to the online social network that a first user of the online social network has interacted with on the online social network, wherein each content object is associated with one or

Assignees

• Facebook Inc

Inventors

• Shorman Aigerim
• Yang Jiawen Connie
• Shrewsberry Eric Andrew
• Yeh Wei Zhong
• Fotinatos Jason
• Ho Dustin S
• Murray Jack
• Lin Jeffrey
• Outridge Alisha S

Classifications

• G06Q10/40

  Business processes related to social networking or social networking services · CPC title

• H04L63/101 Primary

  Access control lists [ACL] · CPC title

• H04L67/306

  User profiles · CPC title

• G06F16/48 Primary

  Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually · CPC title

• H04L51/32

  Electricity · mapped topic