• Issues

1. Twitter doesn't allow me to run more than two process with given api-key
2. For Urdu, hindi, People sometimes use their language in english as it sounds out. Google translator is able to detect these languages written in english however it doesn't seem to be perfect
3. If program can't catch up real time data, twitter streaming api raise an error -> You might need to push date to other remote system and let them take cake of it
4. Google translator can't handle emoji. You might need to consider UNICODE Table of all the emoji

• Project Layout

1. Counting country based on certain language
2. Applying data to Machine learning algorithm(RNN, LSTM, ATTENTION..)
3. Visualization with these real-time data
4. Verifying how well a machine learning algorithm works
5. Performance comparison with pyspark if stucked with overwhelming real-time data

• Result(graph..)

• Environment

###Tested python : Python 3.5.2 ###Operating system : Windows