The first step in our project is to train a sentiment analysis and classification Amazon ML model for Tweets. Fortunately, we have access to a rather large twitter sentiment dataset composed of over 1.5M Tweets tagged 0/1 for negative/positive sentiments. The dataset is available at http://thinknook.com/twitter-sentiment-analysis-training-corpus-dataset-2012-09-22/. This dataset is an aggregation of two twitter sentiment analysis datasets:
- University of Michigan Sentiment Analysis competition on Kaggle: https://inclass.kaggle.com/c/si650winter11
- Twitter Sentiment Corpus by Niek Sanders: http://www.sananalytics.com/lab/twitter-sentiment/
This Twitter Sentiment Analysis Dataset contains 1,578,627 classified ...