Twitter-Sentiment-Analysis

Abstract:

Analysed Tweets and classified it into categories- positive, negative, neutral.

Approach

• Imported dataset
• Analysed data
• Used Count Vectorizer
• Prediction using SVC
• Prediction using Random Forest
• Prediction using Multinomial Naive Bayes
• Applied Grid Search Cross Validation
• Comparison between various classifier

Results

• SVC : 0.633
• Random Forest : 0.723
• Multinomial Naive Bayes : 0.755

Conclusion

Best accuracy comes from Multinomial Naive Bayes.

Clone the Project:

git clone https://github.com/wikiabhi/Twitter-Sentiment-Analysis.git

MIT License Copyright (c) 2018 Abhishek