Abstract
The object of the analysis carried out here is the Old Babylonian archive of Nūr-Šamaš, provenient from the Diyala region, formed by 116 loan contracts conceded by Nūr-Šamaš, and involving some 400 people as debtors and witnesses. Data obtained from the documents of the archive will be used to generate a graph modelling the relations of the attested persons. The graph will then be split into classes maximizing modularity, that is to say, into clusters optimally composed by personal identifications that statistically tend to connect more inside each cluster than with personal identifications of other clusters. Finally, this will be used to raise or lower the probability of stating that pairs of identifications of the type 'ZZZ, son of YYY' and 'YYY' may correspond respectively to a son and his father.
Highlights
This paper is the second in a series concerning digital methods to examine sets of documents from the Diyala region in the Old Babylonian period
It should remain clear that the strategy proposed here is not presented as a definitive and general solution for the problem of pairing fathers and sons
It is suggested that, especially when large sets of documents are under analysis, the study of clusters maximizing modularity might be a useful piece in the toolbox of the researcher
Summary
This paper is the second in a series concerning digital methods to examine sets of documents from the Diyala region in the Old Babylonian period. In what follows, selected data obtained from the documents of the archive of Nūr-Šamaš will be used to generate an abstraction, namely a graph of relations. This graph will be split into clusters maximizing modularity. As explained below in more detail, a partition of the graph into smaller components will be made so that identifications in the same component will be, as much as possible, more frequently connected among themselves than to identifications from other components This will be used to raise or lower the probability of correctly stating that the two occurrences of the name "YYY" such as the ones in the previous paragraph originally referred to the same person or not
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have