US 12,001,489 B1
Ethics-based multi-modal user post monitoring
Ramya Malur Srinivasan, San Diego, CA (US); and Kanji Uchino, Santa Clara, CA (US)
Assigned to FUJITSU LIMITED, Kawasaki (JP)
Filed by Fujitsu Limited, Kawasaki (JP)
Filed on Jan. 25, 2023, as Appl. No. 18/159,103.
Int. Cl. G06F 16/906 (2019.01); G06F 16/583 (2019.01); G06F 16/783 (2019.01); G06F 16/9035 (2019.01)
CPC G06F 16/906 (2019.01) [G06F 16/583 (2019.01); G06F 16/7834 (2019.01); G06F 16/9035 (2019.01)] 20 Claims
OG exemplary drawing
 
1. A method, executed by a processor, comprising:
receiving a set of user posts and metadata associated with the received set of user posts;
fitting a parameterized logistic function on the metadata associated with the received set of user posts;
clustering each of the received set of user posts associated with corresponding metadata, based on the fitted parameterized logistic function;
determining a category-prior associated with each of the received set of user posts, based on the clustered set of user posts;
determining a set of multi-modal features associated with each of the received set of user posts;
determining a feature likelihood for each of the received set of user posts, based on the determined set of multi-modal features;
determining a final category associated with each of the received set of user posts, based on the determined category-prior and on the determined feature likelihood; and
rendering post categorization information based on the determined final category associated with each of the received set of user posts,
the received set of user posts is filtered based on the determined final category.