YouTube Political Sustainability Sentiment
Cross-country sentiment analysis of political sustainability discourse on YouTube
Comparative study of how political actors discuss sustainability on YouTube across six countries — USA, UK, Canada, Australia, India, and Pakistan — analyzing both what gets said and how audiences respond.
Approach
- Collection — video links, comments, and engagement stats (views, likes, comment counts, upload dates) pulled per-channel via the YouTube Data API v3, keyed to a keyword list (e.g. “Climate Emergency”, “Renewable Energy Green Jobs”, “Reforestation”) and a channel list tagged by country and party.
- Sentiment — every comment scored with VADER, then aggregated by country, party, and keyword.
- Analysis — keyword frequency over time, engagement distributions by keyword, and cross-country/cross-party sentiment comparison.
Findings
- “Climate Emergency” is the most-used keyword across all six countries since 2019.
- “Renewable Energy Green Jobs” content drew the highest view counts.
- Comments on “Reforestation” content skewed most positive.
- Australian political parties used climate-related keywords most frequently; the USA’s Republican Party had the least content on these topics.
Limitations — disclosed
- YouTube Data API’s 10,000-request daily quota caps large-scale collection; the collector saves partial data on quota exhaustion rather than failing the run.
- Some channels disable comments entirely, limiting sentiment coverage for those videos.
- Scope is bounded to a fixed keyword and channel list — it doesn’t capture sustainability discourse outside those terms.
Stack
Python, YouTube Data API v3, Pandas, VADER, Matplotlib/Seaborn.
Status
Complete.
Code: github.com/awais-de/dsdm