EPFL CS486 Human Computer Interaction Project: AI is Beautiful

• Python 3.x
• Others - See Dependencies (Build With).

Installing dependencies before running using the following:

pip install Flask

And repeat "pip install" for all other dependencies.

• Step 1 - Unzip "HCI Project-Sentiment Analysis Dataset.csv" file and put it into "datas" directory.

• Step 2 - Open Termial Window 1, Move into "datas" diractory from the terminal.

• Step 3 - Run "project.py" and wait for it to be done (instructions show on terminal logs, only needed to run once for analyzing training dataset).

python project.py

• Step 4 - Open Termial Window 2, Move into the top directory and run a Redis server (wait for the connection to establish and keep the window open)

bash redis.sh

• Step 5 - Open Termial Window 3, Move into the top directory and run a Celery worker (wait for the connection to establish and keep the window open)

celery worker -A app.celery --loglevel=info

• Step 6 - Open Termial Window 4, Move into the top directory and start the application (wait for the Debugger to show up)

python app.py

• Step 7 - Open Sentiment Analysis Visualization in your favorate browser

• Final Step - Play with the visualization : )

• Python 3.6.5 - Language used

• Bootstrap
• JQuery - Web
• Flask 1.0.2
• Socket.IO
• flask-socketio - Events
• NVD3 - Visualization

• Redis - Database
• Celery - Task Queues
• Python-Pattern

• scikit-learn - Classifications
• gensim - Word2Vec models

• Yipeng Ji

• Sentiment140 - Dataset used
• scikit-learn - Methods used
• Christopher Potts - Tokenizer used
• John Wittenauer - Analysis Steps applied