Right Wing??? Left Wing??? Independent???

Right Wing??? Left Wing??? Independent???

Have you ever wanted to know what party affiliation your favorite news agency had? Have you ever thought that one news agency is slanted to the left or the right? Have you had arguments with coworkers, friends, or family about whether or not their favorite station is biased or not? Well, now you can utilize the analysis of big data and text mining to prove your theory or debunk theirs!

The method of implementation is fairly simple. Firstly collect tweets from the stations that are in question, ensuring that they are from the same time frame. This can be done by copying and pasting it into an Excel spreadsheet. Now using a program to text mine you can count the frequency of individual words and create word associations. I used Rapid Miner to analysis my data. There are several tutorials on the internet that can walk you through that process. When doing this analysis one must be sure to look at the word associations. So that you can see what context the most frequent words were spoken in. For example whether it was a pro-Obama statement or not or whether it was a pro-gun control statement or not. This is key, in understanding the political biased of a news agency.  Now for a quick example!

In this example I looked at CNN, Fox News, BBC International, and NPR. Here are some of the tables I created from the gathered data.

 From the analysis I found many interesting things. Mainly, that there is a large difference in the stories that news agencies report and that they tend to be biased towards the political parties that they are affiliated with. CNN and NPR tend to have more left wing topics with NPR not being quite as far left as CNN. FOX tends to be more right wing in their topics. BBC seems to be the most independent station that I analyzed. There are a finite number of active stories in world news and one would think that all news agencies would report on them relatively equally. This has not been the case with the data I have analyzed.