Editing and Updating Tags

There are various options available for modifying tags and tagging results.

1. Bulk Editor

Relevance AI's Editor is an advanced tool designed to facilitate large dataset modifications. This is done by applying a variety of user-defined rules. The Bulk Editor is geared towards editing larger datasets, but also allows for single-entry modification. All is done through 3 main steps:

  • Define what you are looking for
  • Define the change(s) you have in mind
  • Apply the change(s)
Relevance AI - Bulk Editor

Relevance AI - Bulk Editor

2. Edit tags on the Review and Refine step of AI-Tagging

AI-tagging workflow takes care of a variety of steps, from keyword extraction to tag modification. Its GUI is specifically designed to ease the tag modification process.

When tags are generated, in the review step, you can remove tags or add new ones and apply the changes.

When tags are applied, in the refine step, you can confirm tags, remove tags or add new ones, either one document at a time.

Or in batch, as shown in the image below:

Note: Refine will be deprecated soon. Please use Editor instead.

3. Edit tags on the Explorer dashboard

Relevance AI's Explorer is a great tool for analysing and understanding your data. One of the available functionalities is the ability to modify field values. Here we explain how to modify values on a Category view.

Scroll down on the page and locate the document whose tags you wish to modify. Click on "Full view" on the top right corner of the cluster card.

Relevance AI - Edit tag on Explorer

Relevance AI - Edit tag on Explorer

Click on "Edit".

Locate the field that you wish to modify. Use 'x' to remove a tag, or type in new values for tagging, and hit "Save document" when done.

4. Workflows for modifying tags

  • Add tags to existing tags receives a new list of tags and retag the dataset (i.e. applies the updated list to tag/label the entries in the dataset)
  • Remove tags receives a tag and removes it if found in the tagging results
  • Rename and combine tags combines the results of different tagging experiments