AI Data Annotation Practice (Sentiment/Intent labeling and LLM response evaluation)
Performed text labeling and classification tasks including sentiment and intent tagging, following structured guidelines to maintain consistency. Evaluated LLM-generated responses using criteria for accuracy, clarity, and helpfulness. Conducted dataset and workflow review to improve overall data quality and reduce inconsistencies. • Labeled 30+ text samples for sentiment (positive, negative, neutral) • Classified user queries into intent categories (purchase, research, support) • Evaluated 15+ AI-generated responses for accuracy, clarity, and helpfulness • Reviewed datasets to identify anomalies and improve quality