US 11,921,808 B2
Auto-evolving of online posting based on analyzed discussion thread
Shikhar Kwatra, Raleigh, NC (US); Paul Krystek, Highland, NY (US); John D. Wilson, League City, TX (US); and Sarbajit K. Rakshit, Kolkata (IN)
Assigned to International Business Machines Corporation, Armonk, NY (US)
Filed by INTERNATIONAL BUSINESS MACHINES CORPORATION, Armonk, NY (US)
Filed on Aug. 3, 2020, as Appl. No. 16/947,454.
Prior Publication US 2022/0035884 A1, Feb. 3, 2022
Int. Cl. G06F 7/00 (2006.01); G06F 16/9536 (2019.01); G06F 16/955 (2019.01); G06N 5/04 (2023.01); G06N 20/00 (2019.01)
CPC G06F 16/9558 (2019.01) [G06F 16/9536 (2019.01); G06N 5/04 (2013.01); G06N 20/00 (2019.01)] 20 Claims
OG exemplary drawing
 
1. A method for online posting annotation, the method comprising:
identifying a topic in an online posting using an ensemble model consisting of more than one base model, wherein at least one of the more than one base model utilizes natural language processing (NLP) and semantic understanding techniques, and wherein the ensemble model aggregates a result of each of the more than one base model into a final result;
summarizing contents from a conversation thread of the online posting, wherein the contents comprise one or more comments and questions;
determining that the one or more comments and questions to the online posting relates to the identified topic of the online posting using at least text analysis techniques to establish a correlation between the comment and the online posting;
updating the online posting based on additional information from the conversation thread;
amending, dynamically, at least one detail from the one or more comments and questions to the identified topic of the online posting until a threshold is reached which indicates that all information regarding the online posting has been included in an amendment to the online posting;
tagging the at least one added detail with at least one piece of metadata;
validating a viewing access based on the at least one piece of metadata;
viewing dynamically appended content to the online posting, based on relationship type, strength, and profile of friend to user who posted the online posting; and
displaying an amended online posting.