Sentiment-modules on online social networks

Patent metadata

Field	Value
Publication number	US-10216850-B2
Application number	US-201615014846-A
Country	US
Kind code	B2
Filing date	Feb 3, 2016
Priority date	Feb 3, 2016
Publication date	Feb 26, 2019
Grant date	Feb 26, 2019

Abstract

In one embodiment, a method includes accessing a plurality of communications, each communication being associated with a particular content item and including a text of the communication; calculating, for each of the communications, sentiment-scores corresponding to sentiments, wherein each sentiment-score is based on a degree to which n-grams of the text of the communication match sentiment-words associated with the sentiments; determining, for each of the communications, an overall sentiment for the communication based on the calculated sentiment-scores for the communication; calculating sentiment levels for the particular content item corresponding sentiments, each sentiment level being based on a total number of communications determined to have the overall sentiment of the sentiment level; and generating a sentiments-module including sentiment-representations corresponding to overall sentiments having sentiment levels greater than a threshold sentiment level.

First claim

What is claimed is:

1. A method comprising, by one or more computing devices:
accessing, by the one or more computing devices, a plurality of communications authored by one or more users of an online social network, each communication being associated with a particular content item and comprising a text of the communication;
calculating, for each of the plurality of communications, one or more sentiment-scores corresponding to one or more sentiments, respectively, wherein at least one of the sentiment-scores is based on an output of a first classifier function, wherein the output of the first classifier function is calculated based on:
a degree to which one or more n-grams of the text of the communication match one or more sentiment-words associated with the one or more sentiments, and
a context determined to be associated with the particular content item, wherein the context is determined based on one or more n-grams associated with the particular content item;

determining, for each of the plurality of communications, an overall sentiment for the communication based on the calculated sentiment-scores for the communication;
calculating, by the one or more computing devices, one or more sentiment levels for the particular content item corresponding to one or more sentiments, respectively, each sentiment level being based on a total number of communications determined to have the overall sentiment of the sentiment level; and
generating, by the one or more computing devices, a sentiments-module comprising one or more sentiment-representations corresponding to one or more overall sentiments having sentiment levels greater than a threshold sentiment level.

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 communications of the online social network.

3. The method of claim 1 , wherein a particular sentiment-word is determined to be associated with a particular sentiment based on a usage of the particular sentiment-word in communications on the online social network that are determined to be associated with the particular sentiment.

4. The method of claim 1 , wherein the at least one of the sentiment-scores is further based on the particular content item.

5. The method of claim 1 , wherein the at least one of the sentiment-scores is further based on a content-distributor associated with the particular content item.

6. The method of claim 1 , further comprising sending the sentiments-module to a client system of a first user of the online social network.

7. The method of claim 6 , further comprising:
receiving an input from the client system, the input comprising a selection of an interactive element associated with a particular overall sentiment; and
sending, to the client system for display, one or more communications determined to be associated with the particular overall sentiment, the one or more communications being viewable by the first user as determined by privacy settings associated with the one or more communications.

8. The method of claim 7 , wherein each of the one or more sent communications is authored by a social connection of the first user.

9. The method of claim 6 , wherein, for each represented overall sentiment of the sentiments-module, a numerical representation of the total number of communications determined to have the represented overall sentiment is included in the sentiments-module.

10. The method of claim 6 , wherein the total number of communications determined to have the overall sentiment comprises a total number of communications authored by one or more second users having one or more user-attributes corresponding to a particular user-attribute of the first user.

11. The method of claim 10 , wherein the particular user-attribute of the first user is demographic information.

12. The method of claim 10 , wherein the particular user-attribute of the first user is location information associated with the client system of the first user.

13. The method of claim 1 , wherein the calculation of the sentiment-scores for the each of the plurality of communications further comprises calculating two or more sentiment-scores based on two or more classifier functions.

14. The method of claim 13 , wherein the determination of the overall sentiment further comprises summing the sentiment-scores calculated from the two or more classifier functions.

15. The method of claim 14 , wherein the summing of the sentiment-scores further comprises weighting the sentiment-scores from the two or more classifier functions.

16. The method of claim 13 , wherein the calculation of the overall sentiment for the each of the plurality of communications further comprises applying a threshold filter, wherein the overall sentiment for the communication is determined to be a null overall sentiment if there is a threshold level of contradicting sentiment-scores for the communication.

17. The method of claim 13 , wherein each of the two or more classifier functions calculates its respective sentiment-scores based on different sets of information, the different sets of information being selected from a group consisting of: the text of the communication, a title associated with the particular content item, and a text of one or more replies made in response to the communication.

18. The method of claim 1 , wherein the sentiments-module is generated for the particular content item if there are a threshold number of communications for which an overall sentiment is determined.

19. One or more computer-readable non-transitory storage media embodying software that is operable when executed to:
access a plurality of communications authored by one or more users of an online social network, each communication being associated with a particular content item and comprising a text of the communication;
calculate, for each of the plurality of communications, one or more sentiment-scores corresponding to one or more sentiments, respectively, wherein at least one of the sentiment-scores is based on an output of a first classifier function, wherein the output of the first classifier function is calculated based on:
a degree to which one or more n-grams of the text of the communication match one or more sentiment-words associated with the one or more sentiments, and
a context determined to be associated with the particular content item, wherein the context is determined based on one or more n-grams associated with the particular content item;

determine, for each of the plurality of communications, an overall sentiment for the communication based on the calculated sentiment-scores for the communication;
calculate one or more sentiment levels for the particular content item corresponding to one or more sentiments, respectively, each sentiment level being based on a total number of communications determined to have the overall sentiment of the sentiment level; and
generate a sentiments-module comprising one or more sentiment-representations corresponding to one or more overall sentiments having sentiment levels greater than a threshold sentiment level.

20. A system comprising: one or more processors; and a non-transitory memory coupled to the processors comprising instruct

Assignees

• Facebook Inc

Inventors

• Kazi Rousseau Newaz
• Rich Mark Andrew
• Sauper Christina Joan
• Herda{Hacek Over (G)}Delen Amaç
• Tanikella Soorya Vamsi Mohan
• Westervelt Brett Matthew
• Loomans Maykel Andreas Louisa Jozef Anna
• Bussing Adam Eugene
• Zheng Shuyi

Classifications

• G06Q10/40

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

• G06F16/9535 Primary

  Search customisation based on user profiles and personalisation · CPC title

• G06F40/30

  Semantic analysis · CPC title

• G06Q30/0251

  Targeted advertisements · CPC title

• G06F16/9536 Primary

  Search customisation based on social or collaborative filtering · CPC title