Blog post 5: Data mining & quantifying literature - Bella DeMarco
After reading Chapter 7: data mining & analysis, my eyes have been opened to a vast world of different subsets of data categorization and organization. From reading about information visualization, I knew that “visual forms create meaning, they don’t just display it.” (86) Meaning that whatever would be discovered through data mining, would hold meaning to the literature. I learned that by displaying metrics as graphics, there could be interpretations and judgements that are factually and analytically based. This also depends upon which graphic to use for the
metrics present, and there must be dedication to the decision or else the graphic does not prove anything about the data.
This whole chapter was very easy to understand because I pictured Voyant tools while reading. I remembered how easily manipulable the graphs and bubble lines were, that was a graph displaying metrics, AKA information visualization. By applying data mining to Voyant you could recognize traits, patterns, and style in authors, across authors, and even across genres. I think of my group project, working with extremely dated literary classics by Charles Dickens, and picture how data mining could reveal patterns about the text. They have already been showing me what I already knew, classic literature does not necessarily mean good literature. Voyant displays a repetition of words like “said” “ma’am” and “of” which seems dull compared to other analysis I have heard from classmates. His work displays prominent themes on humanity and society, but the plot does not hold enough excitement to pique my interest in the matter. Which is why I am particularly grateful for Voyant tools being able to highlight repeated words in chapters which often highlights characters and relationships.
Overall, I think that data mining tools are extremely helpful in a world where digitization requires distant reading. With the sheer amount of data present in literary works, information visualization and data mining are necessary for making claims.
It can be helpful, but also be skeptical of solely relying on these tools to inform a judgement or thesis about a work or author. And, as you mentioned, yours are dated text, so the tools do not pick up on time period, or historical context, or even language differences, etc. Though, of course, you are starting to see some evidence of "old fashioned" speech here! I wonder how Dickens will fare in comparison to the mystery authors, as far as more narrative themes go.
ReplyDeleteOh, I love the cat. :)
ReplyDeleteI totally agree with you. Unlike the other chapters, this one made the most visual sense to me. Since we were already exploring this sort of data beforehand. I also agree with how these tools need to be used in the modern age. Everything is becoming digital. Without this at our fingertips, understanding the information will be difficult. Mix this with the skills of close and distant reading and anyone will be able to adapt to the future. Especially with how things are today.
ReplyDeleteI really liked how you compared data mining to the Voyant tool and how Voyant uses data mining in order to distant read texts. You also related this to your project and how these tools can be extremely useful when it comes to understanding our own projects. Data mining tools allow us to see certain patterns and language in our reading that you may not notice when reading it yourself. It also allows us to pick out dull filler words in the readings, that shows how dull the story language may be.
ReplyDelete