I used Voyant to analyze two speeches: Michelle Obama’s final speech as First Lady of the United States, and President Barack Obama’s final speech as President of the United States.
Voyant takes in text which becomes a corpus that is then analyzed for things such as word frequency, total word count, unique word count, etc.
The Voyant tool is pretty user friendly, but I did have a few technical difficulties. The graph for one of my speeches disappeared, and even after clicking many different buttons, I could not seem to recover it.
The word clouds that the Voyant tool creates are in a section called “Cirrus,” where the top frequency words of a corpus are visualized. The most frequent words in Michelle Obama’s speech were: “people,” “young,” “college,” and “education.” President Obama’s most frequent words were: “democracy,” “america,” “change,” “people,” and “years.” When analyzing this, I did remove words such as “the,” and “it’s” so that I could focus on the heart of the speeches.
The word frequencies for these speeches seemed to pull out words that were foundational for each person as they navigated their time in the White House. Both Obamas ended by summing up the things they stood for while in office, and the tool highlighted that.
Overall, my experience with the Voyant tool was good. It seems like it is very useful for text analysis!
One response to “Voyant Exploration”
Hi Amara, this is a great start to analysis using Voyant. Using the Cirrus function, you found that
[t]he most frequent words in Michelle Obama’s speech were: “people,” “young,” “college,” and “education.” President Obama’s most frequent words were: “democracy,” “america,” “change,” “people,” and “years.”
And you go on to say that
The word frequencies for these speeches seemed to pull out words that were foundational for each person as they navigated their time in the White House. Both Obamas ended by summing up the things they stood for while in office, and the tool highlighted that.
So, if you were going to use the data provided only by the cirrus tool, what would be your next step for using the results you cite to create an argument about one or both of these speeches? What you provide is a summary of the data that Voyant gives you, but as you reminded us in class, what the computer visualizes is not the analysis. We use that to craft our analysis. Where do we go from here?