Viral Analytics: Embedding Voyeur into Common Web Frameworks

Team

• Stéfan Sinclair (PI)
• Geoffrey Rockwell
• Stan Ruecker

Summary Description

Mashups (the recombination of content and functionality from multiple online sources) are one of the defining characteristics of Web 2.0 (O'Reilly 2005). The underlying principle is clear: the value of content can be greatly enhanced by allowing it to interact directly with content from elsewhere, especially when the user is able to drive the process (Cañas et al. 2007; Bitzer & Shumann 2009). Mashups are so prevalent on the web today that we hardly notice them, from a site with aggregated Facebook and Twitter feeds to travel information sites with real-time traffic, weather and points of interest data (see for example weatherbonk.com).

Voyeur (http://voyeurtools.org/) is a web-based text analysis environment that has been designed to be mashed or integrated into remote sites (the same way YouTube^? videos can be embedded into just about any web page). The ability to embed Voyeur panels represents a significant advancement in the development of online text analysis tools such as HyperPo and TAPoRware (Rockwell et al. 2009). It is currently possible to drop a Voyeur panel into an arbitrary page and to access the dynamic analytic features of Voyeur. The primary purpose of this project is to move beyond a simple ad-hoc integration of Voyeur tools and to test the full integration of Voyeur into various web frameworks and online journals that are commonly used by digital humanists.

In order to accomplish integration of Voyeur with web frameworks, we need to better understand what frameworks are used by digital humanists and why. We know anecdotally of several projects using WordPress^? (the prevalent open source blogging platform), Drupal (a widely-used open source content management system), and OJS (an open source journal management and publishing system), but we lack a clear understanding of just how wide- spread each of the major frameworks is as well as which secondary frameworks are used. We also lack knowledge about how users might interact with such tools in a web framework. As such, Viral Analytics has the following research goals:

• Environmental Scan to determine which web frameworks are being used by digital humanities projects.
• Technical Scan to identify the technical challenges of building customized modules for various web frameworks.
• Experiments with digital humanities blogs, journals, and sites to understand how users will interact with tools when embedded in different types of content (journal articles, text collections, library databases, blogs, etc.).
• User Study to understand how analytic tools can be presented in staged levels of complexity for different types of users.

In addition to the survey of web frameworks and a report on technical specificities, we will prototype integration with three strategic partners: Digital Humanities Quarterly (DHQ), Digital Studies / Champs numérique (DSCN), and the McMaster Library (that is interested in experimenting with integration of Voyeur into its front-end interface). By embedding Voyeur modules within these academic frameworks we will test an alternative way of reaching a wide audience of potential users with accessible analytics. We also anticipate that the availability of modules for widely-used open source frameworks will generate additional interest in text analytics by other providers.