Twitter Sentiment

• Siddhesh Gunjal

Objective of this project was to understand and predict the Sentiment for a particular Sentence. For the initial training and testing tweeter data was used for 2 different case studies.

CASE STUDY 1: Sentiment analysis for tweets in the year 2008 (When tweeter actually got famous)
CASE STUDY 2: Sentiment analysis and prediction for tweet in the period of Trump election in US.

For both case studies tweets were analyzed for most common words in +ve & -ve tweets also Hashtag analysis was done to understand the most common hashtag trends in the period. Predictions were done using different Methods including CNN, LSTM, CNN with word2vec, LSTM with word2vec.