Blog Post 4 - Information Visualization and Data Mining

 Johanna Drucker's chapter 6 on Information Visualization was important because it notes that we see visualizations as interpretive things rather than transparent reflections of reality. She explains that the person making the graphs can really change the way data appears when they decided to use a bar chart vs. line graph and changing the scale of the graph. Making an axis look smaller could make a huge difference, same with choosing a different type of chart. This is important to look for in digital humanities work, like Six Degrees of Francis Bacon. It shows so many diagrams of early social connections to show who was important but Drucker reminds us that these things are interpretive and shaped by editing decisions. It also applies to Yesterday, Today, Tomorrow because it shows the personal narratives of past,present and future into charts which can show shared themes between people, but these experiences can't be quantified so information is left out.

I really enjoyed reading this because I've never considered that the way we put data together can change the way people interpret it. When making our graphs and charts from data in lab, we pick it either based on the best fit/how the instructor tells us to, so that could change the way viewers see it.

In Chapter 7, Drucker discusses data mining and analysis. She talks about frequency counts and topic modeling and how they make "distant reading" possible because there are patters. Drucker does make it known that these aren't exactly neutral either. Frequency counts could suggest a thematic difference, but they ignore context and nuance. She notes that it's important to step back from the individual text to try to see patterns instead. 

I really liked the idea of data mining, because it highlights the data rather than making people sift through it. I think sometimes it's easier to grasp the main idea than the gritty details of an experiment/study. I think it was cool to see how people felt through the pandemic and honestly the four words could've made it easier for people to relate. However, all of this made me realize that data seems to always have some sort of bias. I think it can be good when people interpret things differently just to open up conversation, but it can also be bad having people interpret factual data differently than the way it's known to be true. 

Comments

Popular posts from this blog

blog post 1: what is digital humanities?

What is DH?

RM ☆ Blog Post 1