Label your data using AI-tags, your own code-frame, or both
What are tags, labels or code-frames?
Tags (or what is sometimes called labels or code-frames) are simply keywords or keyphrases strongly related to an item (e.g. a comment, a description, etc.).
nature are good tags/labels/code-frames for
I travelled to Australia last summer and enjoyed the amazing nature.
Applying tags/labels/code-frames to items in a dataset is referred to as tagging. It is useful in many ways such as categorising items existing in a dataset.
Tag with coding frame
Relevance AI's "Guided Tagging" workflow helps you tag (i.e. apply a coding frame to) your dataset. To use this workflow you **firstly need to extract the tags/ code frames from your dataset using the Generate AI Tags workflow . Then all of those tags are automatically fetched and represented to you on the "Guided Tagging" setup page. Here, you can directly apply those tags or use your domain expertise to modify the list (recommended). See below for more details.
Note: Depending on the size of the dataset, the initial step of automatically extracting the tag list could take rather long (i.e. about an hour).
How to tag data on Relevance AI's platform
Once you have uploaded your data,
- Run one of the tagging workflow such as Generate AI Tags
- Select your dataset and from the menu on the left-hand side select "Guided Tagging" under workflows. If it is not shown on the menu, click on view all and find "Guided Tagging".
- Follow the steps by :
- Specifying which text field in your dataset should be used for fetching automatically extracted tags. Tags are listed under
AI tags -> Tags fields. In our experiment, we selected
dec_tagspreviously generated through the 'Create AI Tags' workflow.
- Type in the desired number of tags you'd like to see (i.e. total number of tags)
- Modify (add/delete) the automatically fetched tag list using your domain knowledge
- Select the text field the tagging should be applied to
- Type in the maximum number of tags to be applied to each document
- Enter a name for the resulting field where the tags will be saved. This means a new field is going to be added to your original dataset where the result of "Guided Tagging" is saved.
- Once finalised, Execute the workflow by clicking on "Run on the Cloud".
- A new window will open in a few second showing you a workflow-progress table. Wait until your task is marked as Completed. Then go to your dataset to see the guided tagging results under your selected field name.
What is a good tag?
Is the tag precise?
This means that the tag properly conveys the intended topic and minimises opportunity for ambiguity. There are several ways that a tag can be not precise. For example:
- “Pricing/dollars, incl. deals”
- Could be split into 2 separate tags: “Pricing” and “Deals”
- “Convenience of Site” → “Close Proximity”
- "sydney" -> "Sydney" (Capital letters matter!) and similarly "Product Condition" -> "product condition" will help the AI
- "Shopping Experience" could instead be broken down into different components that make up the shopping experience such as "checkout experience", "parking lot", "friendly staff".
- “Pricing/dollars, incl. deals”
Is the tag concise?
This means that this is not an incredibly wordy tag. A tag should not have unnecessary words where possible as this can dilute meaning.
- “Good Customer Service” → “Customer Service” (unless you are specifically looking for the good component of the word. This is important because good customer service will not tag customer service but customer service will tag good customer service).
Updated 3 months ago