US 12,216,983 B2
Document mark-up and navigation using natural language processing
Ajit Narayanan, Mountain View, CA (US); and Lei Shi, Mountain View, CA (US)
Assigned to Google LLC, Mountain View, CA (US)
Appl. No. 17/594,217
Filed by Google LLC, Mountain View, CA (US)
PCT Filed Sep. 29, 2020, PCT No. PCT/US2020/053208
§ 371(c)(1), (2) Date Oct. 7, 2021,
PCT Pub. No. WO2022/071917, PCT Pub. Date Apr. 7, 2022.
Prior Publication US 2022/0318485 A1, Oct. 6, 2022
Int. Cl. G06F 40/117 (2020.01); G06F 16/34 (2019.01); G06F 40/137 (2020.01); G06F 40/169 (2020.01); G06F 40/205 (2020.01); G06F 40/30 (2020.01)
CPC G06F 40/117 (2020.01) [G06F 16/345 (2019.01); G06F 40/137 (2020.01); G06F 40/169 (2020.01); G06F 40/205 (2020.01); G06F 40/30 (2020.01)] 20 Claims
OG exemplary drawing
 
1. A computer-implemented method comprising:
obtaining a text document;
generating a first plurality of mark-ups of the text document by processing the text document by a first plurality of natural language processing (NLP) models, wherein the first plurality of mark-ups comprises a first mark-up of the text document generated by a first NLP model of the first plurality of NLP models, wherein the first plurality of NLP models comprises a plurality of different types of NLP models configured to generate a plurality of different types of mark-ups each of which provides a corresponding coarse representation of the text document, wherein the first plurality of mark-ups comprises the plurality of different types of mark-ups;
displaying, by way of a user interface, the text document and the first plurality of mark-ups of the text document;
receiving a user selection of a portion of the first mark-up of the text document from the first plurality of mark-ups, wherein the portion of the first mark-up of the text document is associated with a portion of the text document;
in response to the user selection of the portion of the first mark-up of the text document, selecting, from a second plurality of NLP models and using a mapping, a second NLP model by which to process the portion of the text document, wherein the second NLP model is different from the first NLP model, wherein the mapping associates each respective NLP model of the second plurality of NLP models with a corresponding mark-up of the first plurality of mark-ups, wherein each respective NLP model of the second plurality of NLP models is configured to generate a corresponding refined representation of the portion of the text document, and wherein the second NLP model is associated with the first mark-up of the text document;
based on selecting the second NLP model, generating a second mark-up of the text document by processing at least the portion of the text document by the second NLP model, wherein the second mark-up provides the corresponding refined representation of the portion of the text document; and
displaying, by way of the user interface, the text document and the second mark-up of the text document.