Data Labeling
Labelled data with the disagreement attached
Classification, tagging, entity identification and cleanup — returned with per-item confidence, so you know which labels to trust and which to review.
What workers do
- Assign images or text to a category
- Identify entities in a document
- Label sentiment or intent
- Find duplicates and malformed records
What you upload
- image_url
- text
- reference id
Any format: CSV, TSV, JSON, JSONL, XLSX or TXT. Columns are detected automatically.
Common questions
- Can I supply my own label set?
- Yes. Your categories become the answer options workers choose from.