Content Moderation Labeling (text Classification)
Worked on text based content moderation datasets used to train machine learning models for detecting unsafe, offensive, or policy violating content. Tasks included classifying user generated text into categories such as safe, spam, hate speech, harassment, or sensitive content. Applied labeling guidelines consistently across large volumes of short form and long form text data. Ensured annotation accuracy through guideline adherence, contextual interpretation of ambiguous content, and review of borderline cases.