Scott Weingart, “Demystifying Networks, Parts I & II” (2011).

“A network is a fantastic tool in the digital humanist’s toolbox.” I’m going to take your word for it, Scott, because I hate math and I picked this Module to challenge myself.

I appreciate the warning at the beginning because, as someone skeptical of technology, I often see new tech applied with the idea of limitless possibilities, with little regard for whether it even applies or is useful for a project. Often the positives seem to outweigh the negatives on paper and fill most of the space, but in this case, we start with what could go wrong.

This man really likes metaphors

Is a network in computer science like a molecule in lab science? The basic image representation of the network looks the same in a way… new rabbit hole activated

“I’ll call this assortment of stuff nodes. You’ll also hear them called vertices (mostly from the mathematicians and computer scientists), actors (from the sociologists), agents (from the modelers), or points (not really sure where this one comes from).” Whoever invented this concept has dibs on naming its parts. We all don’t need to reinvent the wheel in our own disciplines.

For successful network analysis, a 2-mode network is ideal. More than that, algorithms struggle to deal with the data

“Most computer scientists working with networks do so with only one or a few types of nodes. Humanities scholars, on the other hand, are often dealing with the interactions of many types of things, and so the algorithms developed for traditional network studies are insufficient for the networks we often have”

Thank you, Scott, for establishing this pitfall

“Humanistic data are almost by definition uncertain, open to interpretation, flexible, and not easily definable. Node types are by definition concrete; your object either is or is not a book. Every book-type thing must share certain unchanging characteristics.”

Does this mean networks are good for establishing connection paths but not necessarily for creating meaning?

“A network of nodes where almost everything is connected to almost everything else is called dense, and dense networks are rarely useful.”

Ok I can relate to this. Thanks, Scott.

A Sparse node group is more efficient than a Dense one

I don’t know enough about networking yet, even with Scott’s help, but I see how this relates to our group’s project and data collection. When adding subjects to our collected data for searching, we chose not to get super specific because it creates too many variables and criteria; limiting the number of tags makes stronger connections to other data sets.

For Example…

Civil Rights

Community & Culture

Community Activism & Self-Determination

Women

Entrepreneurs & Landowners

Education

Churches

I see these acting as the Nodes

“Digital humanities loves RDF (Resource Description Framework), which is essentially a method of storing and embedding structured data. RDF basically works using something called a triple: a subject, a predicate, and an object. “Moretti is an author of Graphs, Maps, and Trees” is an example of a triple, where “Moretti” is the subject, “is an author of” is the predicate, and “Graphs, Maps, and Trees” is the object. As such, nearly all RDF documents can be represented as a directed network. Whether that representation would actually be useful depends on the situation.”

Why was this not mentioned earlier when we were talking about a 2-node, 1-edge system, Scott?

So, the answer to these problems isn’t easy. The solution is complex and requires constructing an algorithm to analyze the complex network dataset. This makes sense because Scott is right that the humanities often try to find as many connections as possible, which leads to the next question rather than solving a finite problem. I think that we treat our research as fact after we have spent numerous pages arguing it, but in reality, there is no way to say it is 100% fact, just that it is a probable outcome based on the data and argument made.

“A node’s degree is, simply, how many edges it is connected to. Generally, this also correlates to how many neighbors a node has, where a node’s neighborhood is those other nodes connected directly to it by an edge. In the network below, each node is labeled by its degree.”

I know this isn’t the gaming module, but this made me think of Minesweeper

“The odd node out on the bottom left, with degree zero, is called an isolate. An isolate is any node with no edges.”

Scott, how do you have a node with no edge connections?

“…in the U.S. electrical grid network, centrality might mean which power station should be removed to cause the most damage to the network.”

Now you are speaking my language, Scott!

Maybe I am wrong, and the more nodes, edges, and relationships, the better. I guess I was thinking about the easiest way to organize the data, but that is where the network comes in. Scott made the point that the more you simplify your data set in terms of the network, the more harmful it is in the long run for something that needs more connections. So, in the end, the complexity of Humanities research makes a custom algorithm necessary to analyze that data. Moreover, you would need people capable of building that algorithm, which adds another level of complexity to the project. Now, when I go back and look at our own project, I wonder whether we are doing more harm by simplifying it to searchable terms found in the data, or whether creating more categories of connections is more beneficial. hmmmmmmmm

 

On a side note… while working on our group project and compiling spreadsheet data, I found a great story that needed to be shared… In Falls Church, Virginia, in the summer of 1913, a heated baseball rivalry was underway.