Whereas previously we have discussed how data can put forth inaccurate depictions or can be misleading when attempting to draw conclusions, this article demonstrates the use of data in uncovering otherwise hidden insights. As the article highlights, text-based records in particular readily provide the opportunity for analysis and visualization of data.

Visualizations allow for an alternative method of “representing significant features…more compactly and more efficiently…in service of drawing attention to … significant aspect[s].”

From a human perspective, there are two sides to this coin: On one hand, text analysis can be overwhelmingly beneficial in that it can help readers to understand texts that they do not have the time, or perhaps the ability, to otherwise comprehend.  On the other hand, how can text analysis help draw a reader’s attention to items of interest they had not previously noticed? How is a computer to know what items are of importance and know better than a human would? (This dilemma is reminiscent of one a peer mentioned in a previous post concerning text mining, computers tend to privilege informational significance over the aesthetic prose of the test, making it extremely difficult to be able to fully automate the understanding of a text).

Essentially, the computer reads the texts as a series of parts and patterns and it is up to humans to write a code that determines what is “important.” Generally, this would require some preconceived notion of what is important in the text- wouldn’t this ultimately defeat the original purpose of discovering items of interest that the reader had not previously considered? In this sense its interesting to think of the variety of perspectives from which text analysis can be approached (e.g. whether the reader is seeking out a general understanding, details of a specific topic, investigating a topic that could shed light on a factor that was not clearly represented in the text overall etc).

In the article “Making Meaning Count” Sinclair and Rockwell explore the differences in word choice for description of black NFL prospects and white NFL prospects. They address that although written words are not the most flashy way to communicate, it still is the most common way, “In  this simple sense, text is already a type of visualization,” and therefore analyzing the text is worthwhile. I think creating the interactive was an interesting way to analyze the word choices and can really demonstrate the differences. They then go on to talk about the logistics of text analysis and talk about how text analysis can synthesize data and how we can then put it into a visual platform such as a wordle. I think that text analysis seems like something that definitely needs to be further developed but could be a huge tool in data analysis. I think that it could be used as a tool to analyze extremely long texts but I also think it can make mistakes and that’s why humans also need to analyze the data to ensure that things are not being missed.

One of my classmates wrote “I think you can be especially creative when the data is in the form of text.” I would agree with this statement it is in such a simple form there are so many different ways that you can go with it and so many different angles you can take when analyzing it.

After reading Text Analysis and Visualization, I was shocked at the power of text. First, finding racism from plain speech is a difficult task, but text analysis shows that it is apparent. It is very easy to forget how much data is really out there in the world and this quote really made a strong point: “It is estimated that every day some 200 billion emails are sent and some 5 billion Google search queries are performed – and they are nearly all text-based.4 The hundred hours of video uploaded to YouTube every minute would remain largely inaccessible were it not for text-based searches of the title, description, and other metadata. “(Jockers and Underwood). Most of that data would be lost without strong text analysis and visualization.

Another topic that stuck out to me was the communication aspect. People still communicate primarily through text, so it would make sense for that to be the biggest data source. The  text analysis of the toy commercials shows these themes.  From the surface the evidence may seem slim, but the word cloud visualizations show certain words can play a major role in advertisements.  It is extremely easy to read and gets the point across quickly.

A fellow students post Text Mining/Language Standardization by BM stood out to me. They mention “computer will never understand the emotional values and ever-changing expressions of human beings”. I agree this is a difficult thing to represent and it should be taken into consideration. The word cloud completely misses all of the meaning and only judges by frequency. That is a loss in information.

One of my classmates mentioned, “When looking at data, not only is it really important to have an good understanding of the raw data and ways to manipulate it, but it is almost equally important to be able to portray this data in a meaningful way. There are many very exciting ways that people have visualized data,  and I think you can be especially creative when the data is in the form of text.” One of the ways that this can be done is through word clouds. The article mentions that word clouds are “static” and not very interactive, which doesn’t allow people to explore it. However, it is still pretty creative. While it is not the most fun, it maximizes your time because you are drawn to the words and topics that are most important throughout whatever article you would’ve had to sit down and read through. Everything is right in front of you, so you can skip over the less important things because it’s obvious which topics deserve your attention more. The article spoke on videos and other forms of relaying information that are entertaining, but with those methods you would still have to listen to and watch information that you don’t need to.

I thought this article was particularly interesting in the realm of data visualization. When looking at data, not only is it really important to have an good understanding of the raw data and ways to manipulate it, but it is almost equally important to be able to portray this data in a meaningful way. There are many very exciting ways that people have visualized data,  and I think you can be especially creative when the data is in the form of text. I also think that data visualization can be especially important given how it can be very difficult to analyze text by itself. I appreciate my classmate’s blog post that emphasizes a balance when it comes to this. While various creative visuals may enhance the reader’s experience, one must also maintain the core messages and conclusions behind the data itself.

Matthew Jocker and Ted Underwood’s paper Text Analysis and Visualization provides a useful introduction to the methods used when working with large quantities of text. The problem proposed is simple. We live in a world with seemingly endless information; how can we eliminate the fluff and get what we want? One can analyze a corpus (body of text(s)) to search for the ubiquity of a single word, or even a pattern of characters. These results can reveal similarities and/or correlations within the corpus that can be useful when researching and annotating works. One could even create a new text solely based upon the analysis. This allows for powerful visualization, resulting in artistic and insightful interpretations of data. Text analysis and visualization is a powerful supplementary research tool that should not be overlooked.

Laura Klein exposed me to a new view on social networks through her tracking of the relationships and interactions that occurred in Thomas Jefferson’s household. The main example she uses to jump start the paper is one of Thomas Jefferson’s slaves and personal chef: James Hemings. It is easier to track the relationships between Jefferson and other politicians or social elites due to documentation of events and access to education. However, one must also consider that there was networking happening in the lower classes as well. Through documents passed between Hemings and other slaves/workers, valuable relational data has been found and added as metadata to a variety of documents. Klein stressed that looking at the top level and creating a clear diagram of relationships is possible, but not necessarily true to history. Accounting for all interactions (especially those of slaves) reveals that the Jefferson household was a complex network, helping reveal the true nature of many people’s lives, including who they relied on and/or trusted the most.

Jeffrey Binder’s piece about text mining had a few points that stood out to me. These were in the discussion on Topic Modeling, its applications as a literary tool, and its common criticisms. By analyzing the different words included in digital documents, one is able to discern the topics that are discussed inside. What is impressive about these programs to me is that they are able to extract the text and then create new ones for further analysis. However, one must consider what documents they are mining from. Are the works creative or quantitative? Are they trustworthy? The data collected must be interpreted to a standard of source type. Having a better grasp on computational ideas allows one to better mine and interpret the resulting text. One must understand that a computer’s standard for language is directly related to the code it runs.

This article discusses the idea of text analysis and visualization and how we use both skills to gain what we need from text. There are ways that we as humans do these things involuntary however we often rely on programs to scan, track and search for us. We are encouraged to rely on these tools but to not forget our human intuition; the only way things produce meaning is when we chose to give something meaning. An example of this was the difference between a text box and a 3D visual that shows the interaction of those words over time. Both provide information but one does it tailored to our needs and motives for application. I think the major takeaway from the article was that with using digital assistance there should always be a balance. One should not rely heavily on tools for analysis and visualization and if one decides to use tools they should use more than one to get the most out of it.

I would like to thank classmate “msomkuti” for their comments, without them I would not have fully understood the article.