Categorize data

Categorize data uses an AI model to assign each value in a column to a category, writing the result into a new output column. You can optionally guide the model with training examples and a fixed list of allowed categories.

Availability: Requires the WorkflowsAiClassifyValues feature. Contact your OneSchema support representative if you do not see this transform.

What it does

Input: one or more CSV lists.
Output: the same CSV list, with an added column containing the AI-assigned category for each row.

When to use it

Free-text fields (like product descriptions or support ticket subjects) need to be mapped into a fixed set of categories.
You want to standardize inconsistent vendor-supplied category labels into your own taxonomy.
You have a set of labeled examples and want new data classified the same way going forward.

How to configure it

In the MFF builder, add a Categorize data node after the node that produces the CSV list containing the column to classify.

Settings

Setting	Description	Default
Input column name	The column containing the values to categorize.	Empty
Output column name	The name of the new column that will hold the assigned category.	Empty
Training configuration file	An optional file containing example input/output pairs the model uses to learn your categorization logic.	None
Training input column name	The column in the training file containing example input values.	Empty
Training output column name	The column in the training file containing the correct category for each example.	Empty
Enums configuration file	An optional file listing the fixed set of allowed category values. When set, the model is restricted to these categories.	None
Enums column name	The column in the enums file containing the list of allowed category values.	Empty

Example

Input

Description
12oz stainless steel travel mug
wireless bluetooth keyboard

With Input column name = Description, Output column name = Category, and an enums file listing Kitchenware, Electronics, Office Supplies.

Output

Description	Category
12oz stainless steel travel mug	Kitchenware
wireless bluetooth keyboard	Electronics

Troubleshooting

Categories are inconsistent between runs

Provide a training configuration file with clear, representative examples.
Provide an enums configuration file to restrict output to a fixed set of categories.

Values are categorized as "unknown" or left blank

Check that the Input column name exactly matches a column in the incoming file.
Expand your training examples or enums list to cover edge cases in your data.