What am I reading in ScholarWorks? May 2025

For this month’s What Am I Reading in ScholarWorks I have a partner in reading! My dog, Elsa—a great (and sometimes literal) bibliophage. We were very excited as we were processing Professor Halie Rando’s work to come across "Standardizing Canine Breed Data in Veterinary Records Is Challenging, but Computer Vision Offers an Alternative Perspective on Breed Assignment." 

Rando's paper (written with Professor Nick Howe and other Smith authors) presents research on AI as a tool to identify dog breeds. Elsa and I found the article to not only be illuminating as to how breeds are categorized and why it matters for insight into a dog’s health, but also for the way in which it unpacks how AI training models that use images (computer image studies) for dog breed identification are created while coherently explaining why they are limited in their utility and how they might be improved. 

Reading through this process vis à vis dog breeds is a very fun way to examine the problems with AI datasets in general. We humans are unreliable narrators and it can not be a surprise that the AI models running off of our narrations result in “training models [that] predict what a human would think a dog’s breed is rather than the dog’s actual breed ancestry.” Anyone interested in learning about AI models and/or dogs will enjoy this article!

View in ScholarWorks: Standardizing Canine Breed Data in Veterinary Records Is Challenging, but Computer Vision Offers an Alternative Perspective on Breed Assignment.
 

Smith ScholarWorks is Smith’s institutional repository of digital scholarship and research materials created and curated by faculty, students, and staff at Smith College. Browse by collections, disciplines, and authors.