Tweet Clustering, Cluster LSA Summarization and Sentiment Analysis
Sometimes I got curious about what people think/talk about a topic. To get an intuition it's a good way to search this topic on Twitter. Connecting some basic NLP concepts with curiosity this hobby project of mine helps me:
- pull tweets from Twitter-API,
- cluster them into subtopics with a tuned model,
- make word clouds for each cluster,
- summarize those clusters with the help of a LSA Summarizer,
- give cluster sentiment polarity.
In the output folder you can find an example output, you can check it out to get_an_idea :)