Requirements: The client approached DNJ Indiego to analyze the sentiments from tweets. The requirement was to understand peoples emotions.

Solution: We trained and developed a Twitter Sentiment Analysis supervised learning model using python, TensorFlow and NLP libraries. Key implementation steps included:

• Extraction – Fetch tweets using Twitter API keys.
• Preparation – Prepare dataframe for preprocessing the tweets to remove non-contextual words
• Parsing – The tweet messages were processed and we generated the training dataset.
• Feature Extraction – Using pre trained embedding’s using Golve.
• Training – Using LSTM network to train the data.
• Prediction – Trained models was used to analyze the sentiment of new tweets