Publication Date

2018

Abstract

Scholars need tools that will allow them to generalize about the fit of themes, events, and rhetorical styles represented in a body of texts. In this article, we introduce the concept of “nested topics,” an approach to topic modeling large-scale textual corpora that highlights implicit ontologies and relationships within the texts themselves. This tool exploits the fact that topic modeling can be used to generalize about topics on an aggregate level as well as a fine-grained level, an approach that has the consequences of revealing overarching themes that appear across all texts as well as more idosyncratic events and rehtorical styles that adhere to only a few documents. The tool’s effectiveness is tested by modeling discussions of property in British parliamentary debates in the nineteenth century. Nested topics help the authors to locate the rhetorical styles engaged in by Irish parliamentarians as they defended tenant rights in the 1880s.

Document Type

Article

Keywords

Current Research in Digital History

Disciplines

Arts and Humanities | Statistics and Probability

DOI

10.31835/crdh.2018.01

Source

Current Research in Digital History

Creative Commons License

Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License.

Share

COinS