US 11,886,800 B1
Transformer model architecture for readability
Jing Wang, Atlanta, GA (US); John Matthew Mastin, Jr., Atlanta, GA (US); Sowmyanka Andalam, Cumming, GA (US); Piyasa Molly Paul, Decatur, GA (US); Dallas Leigh Taylor, Atlanta, GA (US); and Andres Castro, Atlanta, GA (US)
Assigned to INTUIT INC., Mountain View, CA (US)
Filed by INTUIT INC., Mountain View, CA (US)
Filed on Jan. 31, 2023, as Appl. No. 18/104,258.
Int. Cl. G06F 40/151 (2020.01); G06F 40/166 (2020.01); G06F 40/253 (2020.01); G06F 40/284 (2020.01)
CPC G06F 40/151 (2020.01) [G06F 40/166 (2020.01); G06F 40/253 (2020.01); G06F 40/284 (2020.01)] 20 Claims
OG exemplary drawing
 
1. A method comprising:
detecting, in a written electronic communication, a first input sentence satisfying a readability metric threshold;
processing, by a sentence transformer model responsive to the first input sentence satisfying the readability metric threshold, the first input sentence to output a first suggested set of sentences, wherein the sentence transformer model is trained with a set of documents comprising a set of input documents and a simplified natural language version of the input documents;
evaluating the first suggested set of sentences along a set of acceptability criteria;
determining, based on the evaluating, that the set of acceptability criteria is satisfied;
modifying, based on determining that the set of acceptability criteria is satisfied, the written electronic communication with the first suggested set of sentences to obtain a modified written electronic communication; and
storing the modified written electronic communication.