US 12,136,112 B2
Recommending the most relevant charity for a news article
Shreyansh Singh, Noida (IN); Gaurav Dhama, Gurgaon (IN); Ankur Arora, New Delhi (IN); Kanishka Kayathwal, Kota (IN); Jessica Carta, Stamford, CT (US); and Ganesh Nagendra Prasad, West New York, NJ (US)
Assigned to MASTERCARD INTERNATIONAL INCORPORATED, Purchase, NY (US)
Filed by MASTERCARD INTERNATIONAL INCORPORATED, Purchase, NY (US)
Filed on Aug. 12, 2021, as Appl. No. 17/401,048.
Prior Publication US 2023/0051764 A1, Feb. 16, 2023
Int. Cl. G06Q 30/02 (2023.01); G06F 16/24 (2019.01); G06F 16/2457 (2019.01); G06F 18/21 (2023.01); G06F 18/22 (2023.01); G06F 40/20 (2020.01); G06F 40/284 (2020.01); G06F 40/30 (2020.01); G06Q 30/0279 (2023.01)
CPC G06Q 30/0279 (2013.01) [G06F 18/2178 (2023.01); G06F 18/22 (2023.01); G06F 40/284 (2020.01); G06F 40/30 (2020.01)] 20 Claims
OG exemplary drawing
 
11. A method of applying machine-learning to determine linguistic similarity between descriptions of one or more charities and an article to identify a charity that is relevant to the article, comprising:
accessing, by a processor, content comprising natural language text;
identifying, by the processor, via a sentence tokenizer, a plurality of sentences of the natural language text;
applying, by the processor, a natural language (NL) model to the plurality of sentences, the NL model being pre-trained on a corpus of documents;
generating, by the processor, as an output of the NL model, a plurality of content sentence embeddings based on the plurality of sentences;
for each candidate charity from among a plurality of charities:
i) accessing a charity sentence embedding generated based a charity query of the candidate charity,
ii) comparing the plurality of content sentence embeddings with the charity sentence embedding, and
iii) determining a level of similarity between the content and the charity query based on the comparison; and
selecting, by the processor, a specific charity from among the plurality of charities that is relevant to the article based on the determined levels of similarity.