Tuesday, January 14, 2020

Masked by Trust: Bias in Library Discovery - by Matthew Reidsma



In the spirit of Algorithms of Oppression this book provides insight into how search engine results are generated. Going beyond Google, which may "customize" results for individual users, Reidsma also explains how library databases aren't necessarily without prejudice either, even though all users will see the same results if they enter the exact same search. Auto-suggestions can likewise shape our thinking in ways we may not have anticipated. All search engines are working on algorithms that were created by people (mostly young, white men). Their own biases will necessarily become part of the algorithm, even if they do not intend it.

Reidsma cites Bonnie Nardi and Vicki O'Day's work to suggest that we rethink the metaphor of library tools to a "library ecology".
The thing about the ecology metaphor is that it highlights the interconnectedness of all these different things coming together in one place. It emphasizes the co-evolution of technology and people. Its [sic] about people and tools together.
Perhaps the most  useful reason for dropping the tool metaphor is that tools require convergent thinking. You cannot create a tool with divergent thinking, where many possibilities exist...
...by switching our focus to seeing our technological systems as ecologies, and thus using divergent thinking to address the design and engineering of these systems, we can move beyond the limitations of tool-based thinking and design systems that are made to be used by diverse people.
This gave me a bit to ponder, and will change the way I teach people to think about and use our databases.

No comments:

Post a Comment