US 10,891,339 B2
Categorized social opinions as answers to questions
Susan M. Cox, Rochester, MN (US); Janani Janakiraman, Austin, TX (US); Nadiya Kochura, Bolton, MA (US); and Fang Lu, Billerica, MA (US)
Assigned to INTERNATIONAL BUSINESS MACHINES CORPORATION, Armonk, NY (US)
Filed by International Business Machines Corporation, Armonk, NY (US)
Filed on Nov. 15, 2017, as Appl. No. 15/814,286.
Application 15/814,286 is a continuation of application No. 15/582,922, filed on May 1, 2017, granted, now 10,540,409.
Prior Publication US 2018/0314760 A1, Nov. 1, 2018
This patent is subject to a terminal disclaimer.
Int. Cl. G06F 16/95 (2019.01); G06F 16/951 (2019.01); G06Q 50/00 (2012.01)
CPC G06F 16/951 (2019.01) [G06Q 50/01 (2013.01)] 14 Claims
OG exemplary drawing
 
1. A method of computing, by a processor and a memory device, an opinion as an answer to a question, comprising:
analyzing, by the processor, the question to determine a set of categories, a category corresponding to a type of possible answers responsive to the question, wherein the type of possible answers responsive to the question is determined using a Natural Language Processing analysis to identify a characteristic of the question;
extracting, from the memory device, a set of opinions from social media, wherein each opinion is from a corresponding responder on a social media platform to which the question is sent;
computing, by the processor, a responder-specific language model corresponding to a responder of the opinion, the responder-specific language model constructed using an analysis of historical social data of the responder, wherein the responder-specific language model establishes a threshold degree of sentiment used by the responder in the historical social data;
computing, by the processor, a value of a responder-specific language strength parameter for the opinion, wherein the responder-specific language strength parameter is indicative of a degree of sentiment of the opinion relative to the threshold degree of sentiment used by the responder;
computing, by the processor, a strength of the opinion using a subset of a set of strength parameters, a strength parameter in the subset comprising the responder-specific language strength parameter for the opinion;
ranking, by the processor, the opinion according to the strength of the opinion;
storing the opinion, the strength of the opinion, and the category into the memory device; and
presenting the opinion, the strength of the opinion, and the category in a filterable presentation.