trip_advisor_sentiment_analysis

This project uses the dataset provided by Trip Advisor with binary classification label of ‘Happy’ and ‘Not Happy’. For the word embeddings, we used GloVe provided by Stanford University. We used 1D convolution neural networks (CNN) to predict whether the statement was a ‘Happy’ or a ‘Not Happy’ one. GloVe and NLTK Stopwords are used for word embeddings and removing sentence fillers. We used TF-IDF to remove words with least relevance. https://drive.google.com/open?id=1a2-sCADXtXGa43RZsf1OHCOJrte8Y1lN - Dataset link