Read and Review Automation Anywhere Documentation

Automation 360

Close Contents


Open Contents

Build extraction model

  • Updated: 2022/02/08

    Build extraction model

    Data extraction model typically leverages Artificial Intelligence (AI) and Machine Learning (ML) technologies to accurately extract text, key-value pairs, and tables from documents.

    IQ Bot extraction service

    Requires no separate AI model as this will be done as part of creating and training a learning instance in IQ Bot.

    See Create a learning instance.

    Microsoft Azure Form Recognizer service

    During your use case evaluation, Microsoft Azure Form Recognizer models can be created by Sales Engineering. But for any production deployment, contact Professional Services, or Partner Enablement Manager (PEM) at Automation Anywhere.

    How to Create and train Azure Form Recognizer model (A-People login required)

    The benchmarks for the time taken for labeling fields in the Microsoft Azure Form Recognizer service are as follows:
    • Up to 30 fields: 1 hour
    • 30 to 100 fields: 2 hours
    • 100 to 400 fields: 4 hours
    Note: While creating a model, ensure there are no duplicate model names. And while creating a table, ensure that the table name does not contain a space. For example, Sales chart.xls.
    Send Feedback