When false information spreads online, it can polarize opinions, fuel conspiracy theories, and even influence elections. In times of widespread sharing of misinformation, it can feel overwhelming for those trying to measure the way “fake news” moves through networks and gains followers.

Enter Cody Hennesy, the new Journalism and Digital Media Librarian at the University of Minnesota Libraries. An expert on the use of innovative research tools, Hennesy can make complex research questions — like the spread of misinformation — easier to tackle.

What is text mining?

Text mining, a subset of statistical computing, is a research practice that Hennesy facilitates. Text mining relies on computer software to aggregate and analyze large amounts of textual information — massive sets of data that would be impossible for any person to analyze on their own. Text mining is good for finding patterns and helping researchers to identify cause and effect relationships.

To illustrate the method for those who aren’t familiar with it, Hennesy offered a literary example.

“A classic example would be looking at thousands of novels over a period of 100 years and looking at changes in how gender is represented over the course of that time,” he said. “Using statistical computing, we can actually figure out something as simple as how frequently there are female characters, how much stage time they have, and how that shifts depending on variables like the gender of the authors.”

In this example, a researcher would need access to a large amount of digital text before their software could start its analysis. When Hennesy consults with UMN researchers on a text mining project, much of his work involves helping them gain access to the texts they need — many of them, he noted, are still in copyright.

“I’ve done work around developing legal literacies — helping train researchers to make strong fair use arguments, and understand how contracts and licenses come into play,” Hennesy said. “In terms of library resources, I try to work with vendors to open up some of these texts that we license so that scholars can work with them in new ways.”

Working with DASH

In addition to serving as the liaison librarian for the School of Journalism, Hennesy also works with the Digital Arts, Sciences, & Humanities (DASH) program. DASH supports UMN researchers by helping them use digital tools and methods. As an interdisciplinary program, DASH brings experts together from across the University to consult on research projects.

“I’ve been in meetings with people from LATIS (Liberal Arts Technologies and Innovation Services) and another librarian with government documents expertise, because [the project] involves text mining methods, it involves access, and it involves knowledge of the subject area,” Hennesy said. “The DASH model is a really nice way to bring the right people together and then get someone started on the research.”

During the academic year, the DASH program offers Programming & Pizza, a monthly consultation event geared for graduate students in need of technical guidance.

“Programming & Pizza is a nice access point for people who might not know anyone on campus providing support,” Hennesy said.

What’s next?

Hennesy sees the University Libraries as an exciting place to be. As new technologies emerge and evolve, digital media librarians like Hennesy are leading the way to support the changing research needs of students and faculty.

“The core services of academic libraries are critical and aren’t going anywhere. I really like doing that work,” he said. “And then what’s most exciting to me is figuring out what else we can do as needs are shifting.”

Recently, Hennesy was certified as a Software Carpentry instructor. He joins other certified instructors — like health sciences librarian Caitlin Bakker — to lead Software Carpentry sessions at the Libraries. These two-day workshops for graduate students, faculty, and staff are designed to teach basic computational research skills.

He also has a series of workshops in mind for students to learn text mining methods.

“I’ve been thinking about a web scraping workshop, a workshop on how to use APIs to download social media data and news data, and an introduction to text analysis methods.” Hennesy plans to partner with journalism faculty and with LATIS to develop these workshops and connect with the right audience on campus.

There is plenty of work to be done — whether it’s helping journalism undergraduates with basic research materials or supporting a text mining project to analyze how misinformation spreads through social networks, Cody Hennesy is up to the task.