This module focused on the various forms of text analysis. When I began working on this unit, I was powerfully overwhelmed by the potential that data mining had. As someone who works in the humanities, but in history, the idea of data and vast amounts of it honestly scared me. It was not until I evaluated the various projects, such as America’s Public Bible and my favorite Robots Reading Vogue, that I better understood the importance of data mining. When I specifically worked with the data from the Works Progress Administration’s Slave Narrative Projects data, did I fully understand the critical importance and the essential questions data mining seeks to answer?

Working with Voyant, I learned the key ideas of text mining and the potential a corpus could bring, especially when working on a project with a great amount of data. After uploading the data that I was provided, Voyant created this extremely helpful textual analysis of various aspects of the narratives. Not only could one find the recurring words that appeared throughout many of the narratives, but it also allows you to see where certain words appear in the document and the context surrounding them. With Voyant, there is so much potential to evaluate trends, connections, and context throughout large data portions.

Voyant also allows the user to narrow their data into smaller subsections, such as individual states. The cirrus below is a visualization of the most common words featured in the slave narratives recorded in Missouri. Since I am from Missouri, it was fascinating to evaluate the stories of the formerly enslaved and this data in the larger picture.

This cirrus represents the most common words in Ohio, which differed from those in Missouri. Voyant allows users to compare and contrast critical factors, such as common words, throughout their different data options. This also highlights the differing experiences of the enslaved just a few states away.

Beyond the aforementioned, Voyant allows the user and creator of the data to see vast differences in frequency. In the graph below, I evaluated how the states of Missouri and Ohio differed in their use of the word “Old” in the narratives. Missouri had a much greater relative frequency for the word “old” than Ohio. This, too, highlights the importance of data mining and drawing conclusions based on the data.

Finally, this tool on Voyant summary shows the researcher various comparative issues within the data, whether it be the length of the document, the varying vocabulary, or even distinctive words. The distinctive words section is mainly what I focused on as I drew comparisons between Missouri and Ohio, as well as the larger picture of other states.

When I first began working with text mining and its surrounding process, I was terrified of what it entailed. However, after working with Voyant and seeing how significant data visualization and analysis is, I fully understand not only how important this work is, but with tools like Voyant, the process can be far simpler.