Final Year Project'16 [B.E.]

###Project Approval Presentation

###Project Review Presentation

• Collect Tweets of particular movie and store it in .csv file
• Run NBayesClassifier.py first
• Run normalplot.py parallely to get sentiment heartbeat graph

Run GetTweets.py or newTwitterStream.py to stream tweets based on topic

• Data Collection Center - Stream ~2000+ Tweets of a given user - Stream ~2000+ Tweets for a given topic

• Training Datsets Used - Huge Tweet Corpus - Classified - University of Michigan Sentiment Analysis competition on Kaggle - Twitter Sentiment Corpus by Niek Sanders

• Preprocessing - Tweet Data Cleaning

• Supervised Learning Model - Naive-Bayes Classifier for Tweets - Naive-Bayes Classifier for Articles - Multinomial Baysian Classifier using Term Frequency - Inverse Document Frequency [TF-IDF] for Articles - Long Short Term Memory [LSTM] with Word2Vector Model [word2vec] for Articles - Long Short Term Memory [LSTM] with Document2Vector Model [doc2vec] for Articles

• Visualization Center - Word Cloud Graph(https://github.com/suraj-jayakumar/sentiment-analysis/blob/master/src/visualization/wordcloud.py) - Stacked Represenation of Postive and Negative Tweets(https://github.com/suraj-jayakumar/sentiment-analysis/blob/master/src/visualization/bar_stacked.py) - Location Bubble Graph

© 2015-2016 Suraj Jayakumar under MIT License