AI Data Transformers Workflows Exercise

This article provides workflow exercises for utilizing common Reveal AI Data Transformers.


The purpose of this document is to provide workflow exercises for Reveal's common data transform features such as Image Labeling and Translation.  These exercises are designed to help customers become more familiar with these key features.

<*Return to Workflow Exercises*>

Product Area

Workflow

Relevant Processes

Image Labeling

Find image documents and have image labels created for them

  1. Filter to all documents with a “jpg” extension
  2. Move to the grid view and create a new image labeling job making sure to overwrite existing labels
  3. Monitor progress of image labeling on the jobs tab
  4. Once the job completes, confirm the image label field populates correctly
  5. View a document with image labels and confirm the image labels are showing in the appropriate document viewer widget

KB Articles

  1. Dashboard
  2. Grid
  3. How to Create Image Labels (Image Labeling)
  4. Jobs Management
  5. Image Labeling – Supported File Types

Translation

Translate foreign language documents

  1. Filter to all documents containing a foreign language from the “BD Primary Language” field (recommended to use Japanese)
  2. Create a new translate job converting these documents to English, choosing to create a new destination text set called “Translation” for the results (not shown in gif)
  3. Monitor the translation progress on the jobs tab
  4. Confirm that a new indexing job is completed upon translation finishing
  5. Confirm that the translation text is now available for these documents in the document viewer

KB Articles

  1. Metadata Filtering
  2. Grid
  3. Jobs Management
  4. How to Index Data
  5. Document Review Toolbar

<*Return to Workflow Exercises*>

 

Last Updated 9/15/2023