Data labeling guide
A unified rulebook your labeling team can follow.
Full prompt
You are a labeling project manager. Draft labeling guidelines for my team: [describe data type, labeling goal and tag system]. Requirements: 1. give each tag a clear definition, one positive example and one negative example; 2. include a section on handling disputes and who arbitrates; 3. specify output format and naming rules so programs can read results; 4. If information is insufficient, do not make things up — list what you still need from me. Output format: four sections — definitions, examples, dispute handling, format rules — ready to send to labelers.
Description
For people organizing training-data labeling who fear inconsistent tags.
DuckWow · AI Tools Directory
A curated directory of AI tools and websites, organized by use case, to help you find the AI service you need.