Live-Twitter-Sentiment-Analysis

Sentiment analysis on streaming data of twitter (tweepy api), deep learning Keras.

Datasets:

The public dataset is available at Kaggle.

Environment:

Python 3.7 Tensorflow 1.15.0

Text Editor

Jupyter Notebook

Analysis and Training

Import modules, dataset, preprocess the data, remove non essential words using Inverse Document Frequence (threshold) and training the model using deep neural networks or svm (commented out).

Testing

testing.ipynb works for manual comments. real_time_sentiment_analysis_from_streaming_data.ipynb works for streaming twitter data for hastags and followed by graphical representations.

Thank you.