4d. Textual analysis tools
Several studies have indicated that people generally prefer reading materials in print rather than on a screen (for reasons of comfort, portability and cost, among others). Advances in display technologies have narrowed the gap since earlier days of flickering CRT monitors. Still, it remains that computer screens generally provide content at a much lower resolution than print. While ongoing efforts by several companies to produce electronic paper are promising, one does not have to wait to experience several of the benefits of working with electronic texts rather than print texts. Electronic texts can be readily indexed, searched, browsed dynamically, reconfigured and analyzed. Those are some of the benefits that the HCI-Book project will strive to fully exploit in the design and development of electronic text interfaces.
Despite some preliminary research on user perceptions of text analysis tools, much work remains to be done in studying the practices of readers of electronic texts, particularly of humanities scholars. The design and conception of analytic tools for electronic editions can be approached from two directions: 1) the needs and expectations of users (based in large part on current models), and 2) the innovation of new forms of analytic tools (which would be difficult for most users to imagine, because they would represent an entirely new paradigm of interacting with texts).
There are several ways in which analytic tools can be associated with texts:
- as visual cues embedded in the text itself (typographic indications, etc.; for example, colour coding of documents in HyperPo);
- as links or buttons in a text's margins that spawn results (text pushing, for example "Reading Tools" in Open Journal System); and
- through a tool's ability to access a given text (text pulling, for example TaporWare Tools).
It should be emphasized that the HCI-Book project is chiefly concerned with conducting research on how to integrate analytic tools and texts, and which tools in particular are most appropriate for users of electronic texts. The focus will be on developing prototypes that could subsequently be expanded and/or commercialized in a different context.
What is text analysis and why is it relevant to the interface of the book?
This section explains how people read through using analytical methods and computing tools that support them. Text analysis tools are, in effect, an interface to the book — an unusual type of interface. Some of the questions we want to answer are:
- How do people use text analysis tools in reading or studying a book? What informal ways of analyizing a text have emerged, whether in popular culture or the scholarly community?
- What are the expected and broadly understood text analysis fuctions? What types of tools do most scholars understand, and how do these functions appear in reading environments? For example, we can hypothesize that most scholars understand searching and KWIC (keyword in context) displays; what other analytic functions do scholars understand?
- What could the analytical interface to the book be? What new types of analytical interfaces are emerging? In particular we want to look at visual interfaces to the text and emerging paradigms.
To pursue these questions we imagine the following activities:
- A visual survey of text analysis tools and their interfaces. The survey's result would be a visual thesaurus and a topology of interactive analytical interface elements.
- A survey to find out what sorts of text analysis needs currently exist in the scholarly community. This would be a online survey, possibly followed up by phone interviews. It will try to tease out emerging needs and functions, which should also draw out emerging reading methods independent from tools that can be computer supported.
- Prototyping of new analytical interfaces. Working within an interdisciplinary context that includes computational linguists, interface designers, and text analysis tool developers, the project could develop imagined interfaces that respond to activities i and ii