The following steps need to be followed to achieve data classification:
- Data preparation: Upload a CSV/TSV file containing the text to be categorized, or generate a new dataset directly.
- Adding a Category Column: Click the + button to create a new column, and enter a prompt such as "Main Topic for Category {{text}}" (replace {{text}} with the actual column name)
- Model Selection: Select the appropriate model for the classification task from the Hugging Face Hub (e.g. meta-llama/Llama-3.3-70B-Instruct)
- Optimization of results::
- Manually fix misclassified cells
- Click the like button for quality results
- Click "Regenerate" to apply feedback to the entire column.
- model comparison(Optional): create multiple columns to categorize using different models, add judgment columns to assess differences in quality of results using LLM
This process is particularly suited to scenarios such as analyzing user comments and categorizing question topics, and accuracy can be significantly improved through iterative feedback.
This answer comes from the articleAI Sheets: building and processing datasets using AI models in tables without codeThe
May not be reproduced without permission:AI productivity tools " How to implement data categorization tasks in AI Sheets? What are the exact steps?
English 
简体中文 
日本語 
Deutsch 
Português do Brasil