Cormier commented on the “elusive commonality across the digital humanities landscape.” With new structures and teaching methods, Digital Scholar Lab could help integrate primary source research outside the digital humanities area. He also said that Digital Scholar Lab was not designed merely for those at a faculty or graduate level—even undergraduates can benefit from simplifying the research process.

A Visuals-Heavy Interface

A cloud-based service that incorporates several newer search and discovery technologies, Digital Scholar Lab enables text mining of raw text and uses visual search with graph results. It’s not throwing out Gale’s Artemis interface, with its “rings and tiles” visual display (see the images in the upper-right corner of this article for an example, courtesy of Gale Primary Sources). While Gale uses the terms “rings” and “tiles,” I’ve also seen them referred to, by open source web search engines, as “circles” and “foam trees.” Whatever nomenclature you use, the result is the same—a visual interpretation of your search results when you’re searching very large datasets.

Wendy Kurtz, Gale’s Digital Scholar Lab specialist, explained the elements of the interface and its workflow implications, stressing the importance of unifying content with visualization. In addition to using a number of search filter options, researchers can put search results on a “workbench.” Here, they can analyze results off the platform. Analysis can reveal interconnectivity among the information, objects, and artifacts found. The workbench also provides a community page, which brings researchers together around a particular article or set of articles. Gale is cognizant of the copyright issues surrounding collaboration, particularly when someone has access to a community page, but not to a specific primary source Gale database.

Technology for Analysis

Analysis is hardly a new procedure for digital humanities researchers. But a Gale case study featuring the question “How are feminists depicted in newspapers?” using old-school methods—Python to clean up OCR data and open source visualization tools—was immensely labor- and time-intensive. The same project using Digital Scholar Lab trimmed the time to completion by many weeks.

Another aspect of text analysis is the ability to determine tone. In the beta version of the Digital Scholar Lab, Gale employs IBM Watson’s artificial intelligence (AI) tool for sentiment analysis and personality insights. Watson technology can read, understand, learn, and find patterns in data. Other tools used in Digital Scholar Lab focus on network analysis and clustering, along with text cleaning and term frequency. As the beta testing continues, Gale will refine the choice of tools to reflect the needs of researchers.

Although it’s currently in beta, which means it could change—and Cormier emphasized that what Gale had on display at ALA Midwinter wasn’t the same product that it had shown last summer and that the platform won’t launch until Gale is sure it’s the right approach—I’m confident that the technologies powering Digital Scholar Lab will remain, in one form or another. Putting data in the cloud, using AI, and taking advantage of advanced analysis tools add to the innovative search and discovery components of Digital Scholar Lab.