Blog Post

Analyzing Political Speeches with Voyant

During the third week of October, we learned the basics of Voyant. I have never heard about this platform for doingd textual analysis. In the past, I used Cat Pack and Concordance, but both applications are old and not very functional for my research; moreover, both applications are not completely free. I started using Voyant as a tool for analyzing Hillary Clinton's speeches. For example, here is a word cloud of a speech that the candidate delivered in Philadelphia. This word cloud shows that Clinton was interested in addressing young people and students. This finding is aligned with what many journalists and pundits have explained: Hillary Clinton has not proved to be an attractive candidate for young people.

Here is the word cloud:




Interesting word cloud! I see "young" and "education" figured prominently, so I think you mean that she was trying to address a problem of here by increasing her appeal to that demographic? Voyant is great, and getting better. You can also look at AntConc, which is free, a very powerful tool for corpus analysis. They have beautiful videos explaining each element of the tool. The big difference is that it sticks with numbers, where Voyant converts them into graphics. Both have their advantages. 


During the presidential elections (2016) of my country (Peru) make a comparison between the plans of government of the first five candidates, as you will observe in the following link:

The results of this brief investigation indicated that the plan of government that maintained a greater balance between its concepts was the one belonging to the grouping of Peruanos Por el Cambio.The opposite happened with the plans of government of the groups of Fuerza Popular and Frente Amplio.

Important challenges for this type of research is the concentration of texts from different sources including the transcriptions of speeches and interviews; And also the gestures of the candidates that are recorded in video and photogallery.