Allen, Colin and Murdock, Jaimie
(2020)
LDA Topic Modeling: Contexts for the History & Philosophy of Science.
[Preprint]
Abstract
In this paper we discuss the application of LDA topic modeling to questions that interest historians & philosophers of science, which we illustrate primarily through our own work on modeling Charles Darwin's reading and writing behavior. We discuss the need to go beyond simplistic presentations of topic models that tend to give scholars the idea that the algorithms produce results that are superficial and perhaps unreliable. The ways in which topic models are often misrepresented and misunderstood frame our attempt to convince readers that, despite appearances, topic modeling provides a lot more of value to HPS research than merely providing for enhanced search and information retrieval from large sets of documents. Rather than "topics" we prefer to think of these topic models as revealing contexts for individual reading and wrote, leading us to ask questions about the individual exploration and exploitation of the materials to which a scientist such as Darwin had access. We discuss the use of topic models as tools for identifying influence and measuring creativity within those contexts and conclude that the interplay between human intelligence and sophisticated algorithms will expand the range of questions about science that HPS scholars will ask, and can answer.
Monthly Views for the past 3 years
Monthly Downloads for the past 3 years
Plum Analytics
Actions (login required)
 |
View Item |
![[feed]](http://philsci-archive.pitt.edu/style/images/feed-icon-32x32.png){kind=icon}