The site provides broad categories like slavery, military and local news and more specific sub-headings under each, which prove that the settings in the topic modeling tool indeed make a difference. Although some topics had common words in their titles, they belonged under separate headings because of other words in the context of the material. For example, there was a heading named 'Military' but the military orders were placed under the 'Other' and 'Soldiers' categories instead, because there was some sort of distinction that the topic modeling tool may have helped the user discover. The number of topics that the modeling tool looks for is important, as it influences how specific it gets.

As language is so complex and diverse one should really be careful and know what to look for instead of leaving the topic modelling tool to do all the work. As the human, who comprehends the nuances and connotations of the human readable material, it is one's responsibility to have a general grasp of what the topic is about to spot any discrepancies.

From what I understand, with smart data there is a limit to the number of interpretations there can be for the data, as the texts are mostly homogenous and structured similarly. With big data however there is a vast range of interpretations that are possible, because of the numerous overlaps in the data and the differences in structure or lack thereof. I understand why big data is more useful in humanities compared to smart data in that sense.

"Capta" appears to be a better description indeed, as the content which is usually described as data is simply a partial representation of the source object. Taking a movie as an example, the data and metadata can give quantitative and qualitative information about the movie and its aspects, but the person who only read that data and the person who only watched the movie would not have the exact perspective or understanding of the movie in its entirety. Instead, the person who read the data would share the views of the one who captured it, with a specific goal and with specific emphasis on selected parts of the data. The one who watched the movie would pick up on nuance, language, music, coloring, camerawork, details or feelings that only people who watch the movie can truly experience.

Humanists view the material they work with as more than a finite list of discrete information, but as something whose meaning and value cannot be fully encompassed in a strict data set. Miriam Posner used the analogy of calling your family's photo album "data" to bring this into perspective (2015). The nuance of the word does not capture the material's essence or richness wholly.

This example supports the inaccuracy of 'dataset' as a descriptive term even more, as we can see two completely different films checking all the same boxes and categories. The dataset gives fixed points that capture none of the essence of either movie, which may not exactly make it useless, but rather incomplete and being of very little help to understanding the material.

This analogy was actually quite helpful. The material that humanists work with is not properly, wholly expressible in distinct units. There is a more in-depth value that cannot be captured in a spreadsheet, and so the term data would not be the most suitable description.

It makes sense that XML plays an increasingly crucial role, as its structure allows the developers of the data to store context with the data itself. The syntax of XML can be analyzed and it is interpreted by all types of programming languages without errors. This is what makes it so useful for data transfer.

This part brings the traditional library comparison into perspective, as this is what happens there as well. All fictional works intended for adult readers would be clearly separated from non-fictional works, or books intended for children. The organization and classification of resources makes things easier to locate and access.

I am definitely in this position currently. I have been reading the various descriptions of digital humanities but still have no concrete idea in my mind about what it looks like or what the final product should contain.

This is true, as it allows you to truly come to terms with your understanding of what you've done, how it has impacted you and how it will impact others. This benefits the person who did the project as much as the people it is being communicated to.

I agree with this statement. It's like chemists or physicists writing up lab reports; it's a way to document everything that occurred during the practical part of the project and consolidate one's ideas that could explain them.

"One kind of person" is the part of this statement that speaks the loudest to me. This does not only apply to this specific field, but to all jobs that produce a service that is available to a wide and diverse audience. How can I, as a member of a minority, expect my perspective to be reflected in any way in this service/product when none of the producers are anything like me? They are unable to relate to me and thus fail to account for my views. There needs to be someone who can represent me amongst the producers, at least to an extent. A voice that will speak against the discrimination or offensive elements of the service or product that the 'white male' would never notice.

This was actually the first thing that came to mind when the government regulations idea was brought up. Problems like that have to be solved at the root, not just adding existing structures to the equation, but creating a new structure that is yet to be plagued by the long-established societal beliefs. This is what the following sentence speaks of, and I agree with it.

People tend to lose sight of this so I think it is very important. Learning is what should matter the most in education, not just doing the bare minimum to gain a salary.