8 March 2021
Dr. Geoffrey Rockwell
Professor, Philosophy and Humanities Computing, University of Alberta (Past) Member, Community Planning and Advocacy Council, Compute Canada
Dr. Rockwell specializes in several areas of the humanities, including textual visualization and analysis, multimedia, computer games and humanities computing. He was the project leader for TAPoR (Text Analysis Portal for Research), a network of six Canadian universities, that developed infrastructure for the digital humanities using WestGrid and Compute Canada’s infrastructure and support services. The TAPoR portal (tapor.ca) and Voyant Tools (voyant-tools.org) provide access to online tools for researchers who work with electronic texts.
Dr. Rockwell’s research and the tools he is developing are making it easier for other scholars to apply computational methods to identify patterns in the texts of ‘traditional’ disciplines such as history, philosophy and literature. The research is providing answers to questions that were previously unanswerable.
Big data issues and high-end computing are driving advances in many fields these days, from genomics and bioinformatics to commercial enterprises like Amazon and Netflix. Is this a new area for scholars in the digital humanities?
Not at all. I’m been involved in this field since the late 1980s and there are big data cultural projects going back to the 1970s. It’s only today that you’re now seeing computing and big data techniques going mainstream in the humanities and social sciences.
How have you worked with Compute Canada and the regional consortia to bridge the gap between advanced research computing and the humanities?
I helped organized a workshop with Sharcnet when I was at McMaster University (1994-2008) to find out how people in the arts and humanities were using high performance computing (HPC). Later at the University of Alberta, we held a workshop in 2010 with WestGrid to build and test prototypes aimed at solving practical problems, such finding ways to more efficiently sort through electronic databases and training researchers in the skills needed to use facilities like WestGrid. It helped us understand common computing research needs for humanities researchers across Canada.
You also worked on the TAPoR project which set up servers in labs, as well as text analysis tools, to support digital humanities projects? How is that project continuing to evolve?
TAPoR continues to be a portal where scholars can experiment with various online tools that allow them to manipulate, analyze and visualize electronic text. One outgrowth of that project was a toolset called TAPoRware, which people can use on the Internet to read and analyze digital texts. The next generation of this is Voyant Tools which is a more compute-intensive data-mining technique. When we launched the original TAPoRware tools we were getting about 8,000 tool runs a month; now we’re getting 40,000 to 50,000 tool runs each month from users all over the world.
Why are Voyant Tools so much more popular?
They have several advantages. Voyant is an integrated environment where you can combine many tools. It handles much larger collections and is very forgiving. We now have a version you download and run on your machine which makes it even more robust in situations where you don’t want to use a remote server.
What’s next for this portal?
I’m part of a large multidisciplinary team that received SSHRC (Social Sciences & Humanities Research Council) funding for “Text Mining the Novel” – a project that is developing algorithms and new computational approaches for studying various literary issues, like genre. The successful algorithms will be available as code through TAPoR. We’re also starting the Methods Commons which is online collection of ‘recipes’ on how to mine literary texts.
How is Compute Canada helping to make these big data tools even more accessible for researchers in the humanities?
They have already played a strong support role over the years on the infrastructure side as the TAPoR portal runs on WestGrid. Now Compute Canada is hiring digital humanities specialists whose job it will be to introduce scholars to these tools and platforms. We also have a voice on various boards and committees. Some 40-50% of the professoriate in Canada is in the social sciences and humanities, from literature and language arts to business and law, so it’s great that Compute Canada is finding ways to listen and involve this constituency.