WWU student, faculty create AI-assisted tool to navigate oral histories
WWU Computer Science’s Dustin O’Hara and alumnus Jack Sovelove have created an AI-assisted system to make it easier to explore hundreds of hours of long-form oral-history interviews about the United States Civil Rights Movement.
SpokenHistory.org is a public digital-history platform with 140 interviews from civil rights activists, artists, educators, organizers and changemakers.
SpokenHistory uses AI to let users navigate through the Library of Congress’ Civil Rights History Project by topic, person or keyword search to create and save interview playlists. The site also includes lesson plans for K-12 educators to teach students about the movement through the archives’ materials.
Most AI applications focus on automating an existing task, but O’Hara and Sovelove were interested in how AI can augment human capabilities. They developed a framework called interpretive computing, which allows researchers to use AI as a tool to discover connections, compare perspectives and develop interpretation rather than using it as a replacement for human intelligence.
“The role of humans in this case is central. It’s all about how people make sense of things,” said O’Hara, director of Western’s Internet Studies Center. “Otherwise, what’s the point?”
Typically, researchers navigate through hours of interviews beginning to end and manually sort them into general topics. By letting AI do the heavy lifting of organizing interviews into thematic chapters with summaries, keywords and descriptions, users can easily create and navigate curated playlists in minutes.
For example, researchers interested in learning more about how mothers shaped the civil rights movement can choose to explore SpokenHistory’s AI-curated Family, Faith, Culture, and Community playlists or can keyword search “mothers” to ask the AI to identify relevant interviews. Once a custom playlist is created, users can save it to their device or bookmark the link for future use.
“Interpretive computing asks how AI can help people engage more deeply with complex materials,” Sovelove said. “The aim is not to replace human interpretation or produce a definitive answer. It is to help people discover connections, compare perspectives and participate in the ongoing process of making meaning.”
Sovelove and O’Hara met when Sovelove was a student in O’Hara’s dynamic web course, and the two have been collaborating on projects since 2022. They have worked on SpokenHistory for three years.
“Jack has an unusual ability to move between big ideas and working systems,” O’Hara said. “We could begin with an open-ended question about how people might explore an archive, and he could translate that question into something we could build, test and learn from. That combination of technical skill, creativity and intellectual curiosity made Jack an exceptional collaborator.”
SpokenHistory was partially funded by WWU’s Pilot Project Grant program, which awards $5,000 to support early-stage ideas that can lead to new, long-term directions in research, creative work, or scholarship at Western.
O’Hara and Sovelove’s research was published in Digital Scholarship for the Humanities, and the team presented their research at Digital Humanities Today, an international conference hosted by King’s College London, in June.
“Working on this project gave me the chance to take ideas that initially seemed abstract and discover what they could become through design and experimentation,” Sovelove said. “It was especially meaningful to contribute to the work over several years, from our early voice-interface experiments to presenting the research in London.”
Sovelove graduated in Winter 2025 with a B.S. in mathematics/computer science, and he and O’Hara plan to collaborate on more archival projects in the future.
Explore the archive at https://spokenhistory.org/
Mikayla King, ’17, covers the College of Science and Engineering and Woodring College of Education for University Communications. Reach out to her with story ideas at kingm24@wwu.edu.