Simply put, ChatterTrack identifies trends among a certain user's twitter followers. It allows publishers to track all tweets tweeted by a subset of a twitter user's followers, and store them. It then uses KnightLab's tweet classifier: http://classify.knilab.com/ to classify the tweets, and also performs other analyses on them.
We are building this app for the Spring 2013 edition of Northwestern University's EECS 395 - Technology and Innovation in Journalism, hoster by Professors Larry Birnbaum and Rich Gordon.
Team: Bryan Lowry (Journalist) Liu Liu Khalid Aziz
The application runs off of Django and Apache. You may need to adapt it to get it running off of your favourite webserver. Here is a link to the django documentation that will help you set up your Django installation with Apache: https://docs.djangoproject.com/en/1.2/howto/deployment/modwsgi/
- Make sure you have python (we used version 2.7) and pip installed.
- (recommended but not required) Set up a virtualenv to run the application in.
- Use pip to install the packages in the requirements.txt file:
pip install -r requirements.txtThis will install all python-related libraries required for the application to work. - You may need to install some dependenices that don't come with your default python configuration. We had to install
python-devand the latest version ofgccforpython-nltkon Ubuntu.python-nltkmay also prompt you to download a corpus that it requires to function, in which case, it will guide you through the download. - This project requires an account with the Twitter content provider Datasift: http://www.datasift.com Set up an account with Datasift, and add the username and the license key of your account to the Django settings file under the commented datasift heading.
For any questions about the project, feel free to contact kaziz@u.northwestern.edu