jana93 / TSA-with-SSL

Twitter Sentiment Analysis with Semi Supervised Learning

Twitter Sentimental Analysis with Semi Supervised Learning

Semi-supervised learning can be mainly practized using three commonly used approaches.

1. Wrapper Based Approaches
   1. Self Training
   2. Co Training
2. Topic Based Apporaches
3. Graph Based Approaches

There are also another approaches but they were not much used in recent researches. These are most commonly used approaches.

Key Idea - Getting More Accurate Model over existing Models.