Dynamic embedded topic models and change-point detection for exploring literary-historical hypotheses
This provides a tool for researchers in Comparative Literature and Classics to explore lexical semantic changes over time, though it is incremental as it builds on existing topic modeling techniques.
The authors tackled the problem of analyzing diachronic semantic change in classical and early Christian Latin texts by combining dynamic embedded topic models with change-point detection, resulting in methods for identifying and characterizing patterns that align with traditional scholarship.
We present a novel combination of dynamic embedded topic models and change-point detection to explore diachronic change of lexical semantic modality in classical and early Christian Latin. We demonstrate several methods for finding and characterizing patterns in the output, and relating them to traditional scholarship in Comparative Literature and Classics. This simple approach to unsupervised models of semantic change can be applied to any suitable corpus, and we conclude with future directions and refinements aiming to allow noisier, less-curated materials to meet that threshold.