Horton, Russell, et al. “Mining
Eighteenth Century Ontologies: Machine Learning and Knowledge Classification in the Encyclopédie.” Digital Humanities Quarterly, vol. 3, no. 2, 2009. openresearch-repository.anu.edu.au,
Horton et al. state that the Encyclopédie has three main strategies of organization (as stated in Volume I): alphabetical order, a tree of knowledge, and cross-references. The classification system, based off of the tree of knowledge, is quite complex and while editors Diderot and d’Alembert were able to classify 55,000 articles, their naming system is inconsistent, and some articles are labelled with categories that cannot be found directly in the tree. Using data mining techniques, Horton et al. create a digital classifier and, by “training” it on the existing, pre-classified articles, work to classify the articles that have no category. Additionally, they turn to the pre-classified articles and attempt to re-classify them and then apply the classifier to other 18th century texts. Results, charts, and conclusions are provided in the text.
Country of Publication: United States
Main Classification: Digital Humanities
1. While the goals and results are understood by a general audience, an average understanding of data mining technologies is required for full comprehension of the methodology.
7/13/2020: Created page.