I mentionned in my last post that I have been interested in data mining and data visualization of sumerian administrative texts from the 3rd dynasty of Ur. There are about 65,000 known texts, as cuneiform clay tablets, which record various types of transaction and give insights about the highly structured organization of its society.
In the previous post, I pointed to the GitHub repository in which I maintain the current visualization tools. At the time, there was only one, namely a visualization of the Ur III social network based on single-sender/single-receiver information in the clay tablets. This is presented as a directed graph, where each edge corresponds to a known document.
While this visualization clearly shows interactions between various people, it does not tell anything about the type of interaction. Since the documents record the exchange of goods, one would be interested in knowing which type of goods these interactions concern. This is the objective of the second visualization, which is now available on the same GitHub repository. I opted for another visualization approach, namely a chord diagram, to show the cumulated inflows and outflows of various goods (cattle, barley, etc.) between people during a given (selectable) time-period. This is available here, and here is a snapshot of what it looks like.
It also was an opportunity to work and learn more with d3.js, since this required some tailored code. Interestingly, the standard API for chord diagrams in d3.js does not work for asymmetric flow matrices, so I had to modify and develop some new code, so that inflows and outflows could be represented.
As with the other visualization, it is still not perfect: though administrative sumerian is highly standardized, it is nevertheless difficult to parse automatically, especially when it comes to person names. This is definitely a work in progress, with modifications to come in the future.