Tweetwise is my attempt to create an automated system for Bitcoin trading, implementing the techniques described in Colianni et al. to build a predictive model based on Twitter sentiment analysis [1].
- Create a new virtual environment
$ virtualenv -p python3 venv
$ source venv/bin/activate
$ pip install -r requirements.txt
- Download the NLTK corpora:
$ python
>>> import nltk
>>> nltk.download
python collect_tweets.py -o raw_tweets.txt python process_tweets.py -i raw_tweets.txt -o processed_tweets.txt
I spent ~5 months collecting data (10/16 - 2/17), based on the methods described in the paper. You can download the data here.
[1] Colianni, Rosales, Signorotti. Algorithmic Trading of Cryptocurrency Based on Twitter Sentiment Analysis http://cs229.stanford.edu/proj2015/029_report.pdf. Web. 20 September 2016.