Visualization of Latin Textual Variants using a Pixel-Based Text Analysis Tool

Abstract

One of the most important activities of Latin scholars is to analyze fragmentary copies of a Classical text and assemble an annotated reconstruction as a conjecture about its original form. We have developed a pixel-based visual text analysis tool to help Latin scholars visualize the evolution of historic copies and analyze the details of alterations and errors introduced in transcription. Coordination of pixel-based visualizations with focus+context navigation across multiple views allows compact representation of text variation across scales of text structure. This approach helps scholars validate the accuracy of variations and assess subtle differences across fragmentary copies as well as past reconstructions. In this paper, we describe the central design features of the tool that help scholars analyze the density and distribution of variants by interacting with text at the granularities of words, lines, and pages simultaneously. We present the results of a user study on our initial multiple view focus+context design and discuss how the results motivate a more visually integrated focus+context design using tiered views.