Historical project — early NLP / sentiment-analysis experiments (circa 2010–2011). Kept public for reference. Not actively maintained; Python packaging and Twitter APIs have moved on.
A small collection of algorithms for sentiment analysis of tweets, primarily a Naive Bayes pipeline with train/test pickles.
| Path | Purpose |
|---|---|
naivebayes.py |
Run classification |
naivebayes_train.py |
Train the classifier |
recreate_pickles.py |
Rebuild pickle artifacts from corpora |
pickles/ |
Pre-built train/test data and model |
results/ |
Sample run outputs |
Pickled objects were built from corpora in mohitranka/TwitterSentimentCorpora.
- Educational / archival
- Expect older Python conventions
- Fork freely if you want to modernize it
See LICENSE.