| dc.description.abstract | Scholarship in linguistics, library science, journalism, and classics has long de- pended upon reading and analyzing text in traditional printed formats. Increasingly, manual and automatic text processing techniques are being used to transform text into structured data that can be queried and visualized. Such data often consists of spreadsheets or databases that contain information extracted from the text, such as document metadata, entities and relationships, sentiment, or grammatical structure. Many techniques have been developed to visualize text as structured data. A few display that data in a recognizably traditional format. Representing data within its own text provides a common visual context that allows for more efficient reading and effective interpretation.
In this thesis, we describe a design framework and a reference implementation for visually representing information in traditional text layouts. The framework provides ways to visually represent information about text directly in the text itself, including text styling, highlighting, decoration, and embedding of entire data views. By integrating data graphics into text, readers can see and interact with both the text and its associated data in a unified visual context. The implementation builds upon the declarative language in Improvise to support interactive queries including visual encoding, filtering, and sorting of data for display in the text. We present several application examples and use them to assess the expressiveness, effectiveness, and performance of the framework for navigation and query features desirable to scholars of classical Latin. | en_US |
