Back to results
US 2018/0004718 A1Published application
Fact machine for user generated content
AssigneeYahoo Holdings, Inc.
InventorsAasish Kumar Pappu, Sungjin Lee, Joel Ranjan Tetreault, Amanda Joy Stent
Life of this patent
Estimating the term…
Abstract
A fact machine receives a comment string from a webpage and segments the comments into sub-conversations. The fact machine also identifies entities within comments or draft comments as well as a sentiment and/or attributes associated with the entity. The fact machine then queries a knowledge base for facts relating to the entity, the attributes, and/or the sentiment. The fact machine then provides the facts to a user prior to the user posting a draft comment.
Claims20
A method comprising:
- receiving, by comment segmentation circuitry, a comment string posted on a webpage comprising a plurality of individual comments;
- segmenting, by the comment segmentation circuitry, the plurality of individual comments into at least one sub-conversation;
- determining, by entity extraction and linking circuitry, at least one entity referenced in the sub-conversation;
- querying, by querying circuitry, a knowledge base for at least one fact relating to the at least one entity;
- and
- providing, by user interface circuitry, the at least one fact to a user.
3 more elements
An apparatus comprising:
- comment segmentation circuitry configured to:
- receive a comment string posted on a webpage comprising a plurality of individual comments;
- and
- segment the plurality of individual comments into at least one sub-conversation;
- entity extraction and linking circuitry configured to determine at least one entity referenced in the sub-conversation;
- querying circuitry configured to query a knowledge base for at least one fact relating to the at least one entity;
- and
- user interface circuitry configured to provide the at least one fact to a user.
5 more elements
A method comprising:
- receiving, by communication interface circuitry from a user device, a draft comment to be posted on a webpage prior to the draft comment being posted on the webpage;
- referencing a knowledge graph, by entity extraction and linking circuitry, to determine at least one entity referenced in the draft comment;
- extracting, by sentiment detection circuitry, a polarized sentiment toward the at least one entity within the draft comment;
- extracting, by the entity extraction and linking circuitry, at least one attribute referenced in the draft comment, the at least one attribute associated with the at least one entity and the sentiment;
- querying a knowledge base, by querying circuitry, for at least one fact relating to the at least one entity and the at least one attribute;
- and
- providing, by user interface circuitry, the at least one fact to the user before the draft comment is posted on the webpage.
4 more elements
Dependencies are inferred from the claim text.