sentiment_analysis

Using social media data, especially Reddit’s posts, this report explores applying sentiment analysis to derive public opinion on climate change. Python’s PRAW library was leveraged to fetch data from Reddit ‘climatechange’ subreddit, running queries such as ‘global warming’, ‘greenhouse gases’ and ‘carbon footprint.’ Posts were classified as positive, neutral or negative sentiments with the help of VADER Sentiment Analyzer. To support dynamic sentiment monitoring, the project used a hybrid Lambda architecture where both historical (batch) and real time (streaming) data could be processed. Foundational insights came from batch processing, and real time trends were captured using the streaming. Sentiment distributions and how they varied were visualized using histograms and pie charts and finally presented in tables. Results showed that positive sentiments (60%) were preceded by progress in climate action, while negative sentiments (25 – 28%) pointed at policy failure and crisis. Typically the neutral sentiments (12-15%) related to factual or informational content. The analysis emphasized the polarization of climate discourse and its implications for developing targeted awareness campaigns as well as for corporate sustainability strategies. Results of this work show that sentiment study can provide input for strategies aimed at improving public engagement and aligning organizational activities with public perceptions.