A few months ago Sinclair launched an AI pilot project across the Academic Affairs division. Faculty and staff could apply with project ideas for their departments and if accepted, would get access to institutional paid accounts to ChatGPT or Claude to work on those projects.
I’d long been thinking about potentially using AI to help with library assessment, so it seemed like a good fit.
We have always collected a ton of data related to library instruction, research help services, e-resource statistics, room/space use, tutorial views, and much more. But that data sits in multiple places, is in a wide range of formats, and doesn’t easily tell a unified narrative about the library without a significant amount of analysis and work to combine it all.
And with all the “hats” we typically wear, plus the fact that we don’t have anyone on our staff with a significant amount of data analysis experience, this seemed like a great project to
My goal was to have Claude identify trends, link data to core library learning outcomes, and help suggest what’s working well and what we might be able to improve or change to better serve the needs of our campus community.
I think ultimately this will explore library impact on student success and faculty/staff research in a way that would be extremely difficult and time-consuming without the help of AI.
It also would involve uploading library data to a Claude Project, which did immediately made me think about privacy. But ultimately almost none of that would be personally identifiable, and for the few things that do have email addresses I can easily delete that column before uploading.
Anyway, my proposal got accepted and I’ve been experimenting with Claude for a few weeks now. So far I’ve been trying to give it as much context about our library resources, services, and what we’ve tried so far. When Fall semester starts I’m excited to start to think about how to structure the project for its full launch.
I’ll update again in a few weeks with how the project is going.