Lead Data Annotator – Short-Form Video Engagement & Sentiment Analysis
VideoClassification
Categorized and labeled a dataset of 5,000+ short-form videos to train engagement-prediction models. Tasks included identifying "viral hooks," segmenting video transitions, and tagging specific content purposes (educational, entertainment, or promotional). Adhered to a strict 98% inter-annotator agreement (IAA) standard, utilizing multi-pass verification to ensure label consistency across trending audio and visual tropes.
Categorized and labeled a dataset of 5,000+ short-form videos to train engagement-prediction models. Tasks included identifying "viral hooks," segmenting video transitions, and tagging specific content purposes (educational, entertainment, or promotional). Adhered to a strict 98% inter-annotator agreement (IAA) standard, utilizing multi-pass verification to ensure label consistency across trending audio and visual tropes.
2023 - Present