Title:
ENTITY RECOGNITION BASED ON MACHINE LEARNING
Document Type and Number:
Japanese Patent JP2023021467
Kind Code:
A
Abstract:
To provide a system and a method for obtaining improvement and optimal efficiency of document processing by improving an accuracy and reducing a turn-around time.SOLUTION: A computing system includes a system memory and a processor. The system memory stores a processor executable instruction for a recognition engine. The processor is connected to the system memory, executes the processor executable instruction, and causes the computing system to define a plurality of baseline entities specified from documents of a workflow and generate a corresponding document object model by digitizing one or more documents. The recognition engine causes the computing system to train a model by using the corresponding document object model and a file with tag as input, and to determine a plurality of target entities from target documents by using the model.SELECTED DRAWING: Figure 2
Inventors:
RADHAKRISHNAN IYER
NAVEEN M KUMAR
ESHWAR GANESAN
NAVEEN M KUMAR
ESHWAR GANESAN
Application Number:
JP2022122501A
Publication Date:
February 13, 2023
Filing Date:
August 01, 2022
Export Citation:
Assignee:
UIPATH INC
International Classes:
G06F16/38; G06F40/117; G06F40/216; G06F40/279
Attorney, Agent or Firm:
Patent Attorney Corporation Tokai Patent Office