US 11,710,334 B2
Table item information extraction with continuous machine learning through local and global models
Matthias Theodor Middendorf, Constance (DE); Gisela Barbara Cäcilie Hammann, Grasbrunn/München (DE); and Carsten Peust, Constance (DE)
Assigned to OPEN TEXT SA ULC, Halifax (CA)
Filed by Open Text SA ULC, Halifax (CA)
Filed on Aug. 16, 2022, as Appl. No. 17/888,933.
Application 17/888,933 is a continuation of application No. 17/134,964, filed on Dec. 28, 2020, granted, now 11,455,462.
Application 17/134,964 is a continuation of application No. 16/272,692, filed on Feb. 11, 2019, granted, now 10,909,311, issued on Feb. 2, 2021.
Application 16/272,692 is a continuation of application No. 15/964,654, filed on Apr. 27, 2018, granted, now 10,241,992, issued on Mar. 26, 2019.
Prior Publication US 2022/0391580 A1, Dec. 8, 2022
This patent is subject to a terminal disclaimer.
Int. Cl. G06F 17/00 (2019.01); G06V 30/416 (2022.01); G06N 20/00 (2019.01); G06F 16/93 (2019.01); G06F 16/22 (2019.01); G06F 16/25 (2019.01); G06F 40/174 (2020.01); G06F 40/177 (2020.01); G06F 40/186 (2020.01); G06F 40/216 (2020.01); G06F 40/274 (2020.01); G06V 30/412 (2022.01); G06V 30/414 (2022.01); G06F 18/21 (2023.01); G06V 30/19 (2022.01)
CPC G06V 30/416 (2022.01) [G06F 16/2282 (2019.01); G06F 16/258 (2019.01); G06F 16/93 (2019.01); G06F 18/217 (2023.01); G06F 40/174 (2020.01); G06F 40/177 (2020.01); G06F 40/186 (2020.01); G06F 40/216 (2020.01); G06F 40/274 (2020.01); G06N 20/00 (2019.01); G06V 30/1916 (2022.01); G06V 30/412 (2022.01); G06V 30/414 (2022.01)] 20 Claims
OG exemplary drawing
 
1. A method, comprising:
receiving, by a user device, an instruction to auto-complete a database table based on a document image displayed on the user device;
analyzing, by the user device, a portion of a table that is part of the document image, wherein the portion of the table defines a set of initial coordinates over the document image;
determining, by the user device based on the set of initial coordinates over the document image, data points for an initial extraction;
automatically extracting, by the user device, the data points from the portion of the table that is part of the document image;
populating, by the user device, the database table with the data points automatically extracted from the portion of the table that is part of the document image;
storing, by the user device, information about the data points in a local model as positive examples; and
communicating, by the user device, the local model to a server computer having a global model, wherein the local model is utilized in updating the global model, wherein the global model is utilized by the server computer for automatically extracting table information from document images and populating database fields with the table information extracted from the document images.