by Voyant is an open-source, web-based application for performing text analysis for scholarly reading and interpretation of texts. It is particularly utilized by scholars in the digital humanities to analyze online texts or ones uploaded by users. For one of my classes, I was able to use the program firsthand by going over different interviews of former slaves in different areas across the United States.
To begin, several links provided by the class assignments loaded up different documents that held interviews between the formerly enslaved individuals on what is called the “reader.” There are other key tools as well, including a cirrus, trend, summary, and context. When combined, these tools are organized into what is called a “corpus.” The “cirrus” contains a database of the most used terms in a document, creating a word bubble where the more commonly-used terms are larger than other terms. The “trend” acts similarly to the cirrus, but instead creates a graph utilizing the terms’ relative frequency to the segments of the document. The “summary” is relatively simple, highlighting the basic information of the documentations in the corpus. Finally, the contexts show different terms in the documents, how many there are, and where their locations are in them.
For the assignment, we seemed to focus more on the cirrus, comparing and contrasting the different interviews we had to sort through. Thanks to Voyant, we were able to cover a lot of ground just by observing the terms each interview used, rather than having to read each and every document and keep tally of the terms. Each interview provided a different set of terms and frequencies. For some terms, they appeared in multiple interviews across each corpus. As for others, we see a higher frequency of the terms being used across the different interviews. Combining these ideas, each corpus provided a different subset for observation and data collection, eventually culminating our data into how we answered each of the questions provided for our tasks throughout the module.
To conclude, I find Voyant as a great digital toolset to comb large amounts of data in order to accumulate familiarities of clustered datasets. With the corpus, each tool helped find terms that were used across every interview from every person in every state. I was able to formulate my own opinions when creating conclusions about the former slaves, such as their lifestyles, education, field work, and treatment. After you’re done, each tab across the corpus can be saved and exported for storage and distribution for future use in projects. Voyant’s utilization of the corpus makes sifting through data that much easier for its users.
