Amutheezan / TSAwithSSL

:zap: Twitter Sentiment Analysis with Semi-Supervised Learning

Twitter Sentiment Analysis

Pre-requisites

Before get into the project, please setup these pre-requisites.

1. Python 2.7
2. Scikit-Learn
3. NLTK
4. PyQt4 (for GUI part)

Guide (Ubuntu)

1. Clone the repository by entering following command in terminal

git clone https://github.com/TSAwithSSL/TSAwithSSL.git

2. Go to repository, and go to src folder and type the following commands

cd ..{REPOSITORY}/src
python Trainer.py

then following GUI will come

You can specify the values for following

1. Label Limit
2. Un Label Limit
3. Test Limit
4. No of Iteration
5. Training Type
6. Train Data set
7. Test Data set
8. Confidence
9. Confidence Difference

and click the button for make model; you can also predict the tweet once the model generated.