I have finally managed to motivate myself to learn and use R to conduct some text analysis and mine Twitter data. Thanks to my friends in the MSU EPET program (Josh Rosenberg, Jon Good, Spencer Greenhalgh, Alex Lishinski, and Dr. Chin-Hsi Lin). If you don’t know about R, it is a free software that you can use to do all sorts of qualitative and quantitative analysis. Check it out.
At present, I am working on analyzing tweets around the Miss Representation project, that use the following hashtags: #askhermore, #mediawelike, and #notbuyingit. Learn more about the Miss Representation project here. I will post more on that soon.
Meanwhile, I am playing with wordclouds and sentiment analysis, and tried my hand at it using 5000 tweets that included the word Modi in them. Narendra Modi is the current Prime Minister of India, who is under some grilling lately for different reasons. I am not going into the political discussion here, so, I am just sharing the word clouds that I made using R.
I found a 3:2 ratio between negative and positive sentiments around Narendra Modi as a topic on Twitter. If you make a word cloud of these sentiments separately, this is how their word cloud look.
It is interesting to look at these word clouds and notice what kind of words people are using when they share a certain set of emotions on Twitter. There is talk of enthusiasm, achievements, cooperation, etc. when people talk about Modi in positive sentiment. However, when talking about him in negative sentiment, people tend to use words like credibility, ashamed, losing, rejection, intolerance, etc.
If you are following the news from India, you can get a sense of these emotions,
I will keep playing with R, and post more cool stuff like this.