Back to results
US 10,757,062 B2Granted
Analysis of social interaction sentiment
AssigneeInternational Business Machines Corporation
InventorsLiam Harpur, Philip Mullins, Eric Woods
Life of this patent
Estimating the term…
Abstract
Reducing a negative social interaction includes receiving a response to a post from a user, the response includes content to be posted on an activity stream of a social network, analyzing the content of the response to determine a negative response risk to the post, analyzing a profile of the user to determine a tendency of the user to respond negatively in responses, and executing, based on the negative response risk and the tendency, an action for the response to reduce negative responses directed towards the post.
Claims20
A method for promoting a positive social interaction, the method comprising:
- inferring from user preferences for an author of a post, registered trusted entities associated with the author, wherein a registered trusted entity:
- is a user likely to respond positively to the post;
- and
- views the post before other entities view the post;
- receiving, from a user, a response to the post, the response comprising content to be posted on an activity stream of a social network;
- analyzing the content of the response to determine a probability that the response contains negative sentiments directed towards the post;
- analyzing a profile of the user on a same network where the response is to be posted and analyzing a profile of the user on a different network than where the response is to be posted to statistically determine user response tendency based on historical responses of the user;
- executing, based on the probability that the response contains negative sentiments directed towards the post and the user response tendency, an action for the response to promote positive interaction;
- and
- notifying the author when the registered trusted entities do not respond positively to the post.
7 more elements
A system for promoting a positive social interaction, the system comprising:
- a processor coupled to memory;
- a plurality of engines, wherein each engine comprises program instructions stored in memory and executed by the processor, the plurality of engines comprising:
- a presenting engine to:
- receive from a user, user preferences for a post;
- infer from the user preferences, registered trusted entities associated with an author of the post which registered trusted entities:
- are users likely to respond positively to the post;
- and
- view the post before other entities view the post;
- and
- notify the user when the registered trusted entities do not respond positively to the post;
- a receiving engine to receive, from a user, a response to the post, the response comprising content to be posted on an activity stream of a social network;
- a response analyzing engine to:
- analyze the content of the response to determine a probability that the response contains negative sentiments directed towards the post by:
- analyzing facts and sentiments of the response;
- deriving from the facts and the sentiment, via natural language processing, a meaning of the content of the response;
- determining from the meaning of the content of the response, a probability that the response contains negative sentiments directed to the post;
- and
- classifying the probability;
- and
- monitor responses by the registered trusted entities;
- and
- determine from the responses by registered trusted entities, when a registered trusted entity does not respond positively to the post;
- and
- a profile analyzing engine to analyze a profile of the user on the same network where the response is to be posted and analyze a profile of the user on a different network than where the response is to be posted to statistically determine user response tendency based on historical responses of the user;
- an executing engine to reject, based on the probability that the response contains negative sentiments directed towards the post and the user response tendency, the response to promote positive interaction directed towards the post.
22 more elements
A method for promoting a positive social interaction, the method comprising:
- receiving, from a user, user preferences for a post;
- inferring from the user preferences registered trusted entities associated with the author, which registered trusted entities:
- are users likely to respond positively to the post;
- and
- view the post before other entities view the post;
- receiving a response to a post from a user, the response comprises content to be posted on an activity stream of a social network;
- analyzing the content of the response to determine a probability that the response contains negative sentiments directed towards the post by:
- analyzing facts and sentiments of the response;
- deriving from the facts and the sentiments, via natural language processing, a meaning of the content of the response;
- determining from the meaning of the content of the response, a probability that the response contains negative sentiments directed to the post;
- and
- classifying the probability;
- analyzing a profile of the user on a same network where the response is to be posted and analyzing a profile of the user on a different network than where the response is to be posted to statistically determine user response tendency based on historical responses of the user;
- executing, based on the probability that the response contains negative sentiments directed towards the post and the user response tendency, an action for the response to promote positive interaction directed towards the post;
- monitoring responses by the registered trusted entities;
- determining from the responses by registered trusted entities, when a registered trusted entity does not respond positively to the post;
- and
- notifying the user when the registered trusted entity does not respond positively to the post.
15 more elements
Dependencies are inferred from the claim text.