A Collation Tool for Humanists

Abstract

Comparing a set of similar texts, often called collation, gives a scholar insight into the author's thoughts and motives and into the culture of the time period. Multiple editions can be published by an author for a diverse number of reasons such as political pressure or to further character development. Collating a set of editions can help discover the author's motives, but involves methodically comparing text to find minute differences. This task, which is currently done by hand due to unstandardized printing practices in historic novels, is time consuming, and prone to human error due to fatigue. We propose a Virtual Hinman Collator that automates the comparison process and highlights the differences between a set of pages. This is done through an image processing approach that is less susceptible to the issues of optical character recognition on poor quality images. This tool saves the user an immense amount of time which allows more time to focus on analyzing the text and drawing conclusions instead of finding the differences. With this increased productivity, a better understanding of past cultures can be found through recognizing the motivation and thoughts behind the author's works.