The goal of this project is to have a global idea of how subjects are treated in news media around the world. The initial idea occurred to me after coming across the adbusters campaign #AlltheNewsThatsFitToPrint, where i had the urge to analyse Palestine Vs Israel against the New York Times RSS. Due to lack of immediate results it quickly became more than that, with the possibility of adding a pair of subjects to be analysed against several news channels around the world.
By no means it is meant to be extremely accurate, however it is meant to give a general idea of how certain subjects are treated in different news channels across the world. As of now it analyses the titles of the news post in question and aggregates them by channel source.
The "Sentiment" calculation is based on AFINN which is a "Wordlist-based approach for sentiment analysis", a quick run on how it works can be found here. The Volume stands for the total amount of news in which a certain word appears.
It runs on nodejs, uses mongodb as a database and for the frontend part of things i used Polymer library to make the most use of web components together with the D3 library.
The project is hosted on a free OpenShift account (what a great service they have) and for now it will run until the free account can hold. There is also a GitHub repository although do not expect the best code quality on a weekend pet project :)
Edit: The project no longer exists on OpenShift as they have terminated their free Tier :(
There is also a twitter account that tweets the daily/weekly/monthly/yearly top hits @rssjoust, as well as a yearly mongodb dump if you want to take a look at some data.