YouTube × Sentiment

5 datasets that are both from YouTube and tagged Sentiment.

5 datasets.

YTCommentVerse: Multi-Category YouTube Comment Corpus Analyzed

A YouTube (multi-category comment corpus) dataset of 40,824,926 comments from 20,568,637 unique commenters across 178,027 videos, collected by YTCommentVerse project team via YouTube Data API. Each record contains hashed video ID, hashed commenter channel ID, comment text, vote count, video category (entertainment/gaming/news/music/etc.). YTCommentVerse contains 40,824,926 YouTube comments spanning 25 categories with privacy-preserving hashed identifiers, enabling large-scale multilingual comment analysis. Suitable for YouTube comment analysis, multilingual NLP, engagement research, content moderation, commenter behavior modeling, cross-category platform analysis.

40.8M records20.7M nodes40.8M edges2025

rtGender — All Platforms (Cumulative) Analyzed

A Multiple Platforms dataset of 2,469,836 posts, 26,496,468 responses across 5 platforms, collected by Peixian and Zhang. Each record contains platform, posts, responses, op_gender, responder_gender, sentiment, relevance. Cumulative rtGender dataset combining all platforms for cross-platform gender analysis. Suitable for cross-platform gender analysis, sentiment comparison, response prediction.

29M records21.8M nodes23.9M edges2018

Wish 107.5 YouTube Comments Dataset Analyzed

A YouTube dataset of 2,004,605 comments from 635 videos, collected by Bwan Wando via YouTube API. Each record contains comment text, like count, reply count, video ID, timestamp, anonymized author ID. A large-scale dataset of 2M+ comments from the Wish 107.5 YouTube channel, a popular Philippine FM radio station known for its live music performances. Suitable for sentiment analysis, topic modeling, fan engagement, cultural studies, natural language processing.

2M records2024

rtGender — TED Analyzed

A TED dataset of 0 posts, 190,414 responses, collected by Peixian and Zhang. Each record contains posts, responses, op_gender, responder_gender, sentiment, relevance. rtGender dataset for TED — studies gendered responses in online discourse. Suitable for gender studies, sentiment analysis, response prediction.

190.4K records409.2K nodes346.5K edges2018

YouTube Comment Insights Analyzed

23,992 YouTube comments with sentiment and tone labels and short pros/cons notes, published on Hugging Face by AnandforU, in train (19,992), validation (2,000) and test (2,000) splits. Each record holds the comment text, sentiment, tone, pros and cons. Suitable for sentiment analysis and comment moderation.

24K records2024