Managing Bias When Library Collections Become Data

Abstract

Developments in AI research have dramatically changed what we can do with data and how we can learn from data. At the same time, implementations of AI amplify the prejudices in data often framed as ‘data bias’ and ‘algorithmic bias.’ Libraries, tasked with deciding what is worth keeping, are inherently discriminatory and yet remain trusted sources of information. As libraries begin to systematically approach their collections as data, will they be able to adopt and adapt the AI-driven tools to traditional practices?
 
 Drawing on the work of the AI initiative within Stanford Libraries, the Fantastic Futures conference on AI for libraries, archives, and museums, and recent scholarship on data bias and algorithmic bias, this article encourages libraries to engage critically with AI and help shape applications of the technology to reflect the ethos of libraries for the benefit of libraries themselves and the patrons they serve. A brief examination of two core concepts in machine learning, generalization and unstructured data, provides points of comparison to library practices in order to uncover the theoretical assumptions driving the different domains. The comparison also offers a point of entry for libraries to adopt machine learning methods on their own terms.

The Dataset Nutrition Label (2nd Gen): Leveraging Context to Mitigate Harms in Artificial Intelligence

“…Each decision made during the course of ADS development -from the original business use case to the dataset selection and model training process -further biases the algorithm towards an intended outcome. The goal, therefore, is to understand and manage that bias since complete elimination of bias is not realistic (nor is it desired), and instead to mitigate and minimize harms that may arise from the decisions made by the algorithm or system [5,24]. Through the Dataset Nutrition Label, Data Nutrition Project team is committed to continuing to advance this important area of ADS development through well designed, carefully vetted, and context-aware Labels that help mitigate harm while also improving public understanding of the risks and opportunities of working with ADS, and with particular focus on their underlying data.…”

Det performative samspil mellem digital infrastruktur og kulturelle arrangementer i den professionelle praksis i danske folkebiblioteker

“…Hvor der på det enkelte bibliotek tidligere kunne ophobe sig mange materialer, skal systemet sikre at materialernes cirkulation bestemmes af efterspørgslen. Systemet samler ikke selv data ind, men baseret på maskinlaering anvender det data fra Cicero på nye måder (Liljegren, 2022, s.44;Coleman 2020). IMS centraliserer med andre ord yderligere styringen af materialebeholdningen.…”

Easier material management - at what cost?

“…Technologies like artificial intelligence and machine learning are tools to manage data more efficiently (Boman, 2019) and to provide more generalized decisions and answers (Coleman, 2020). For a long time, algorithms in different search engines have captured the interest of researchers within library and information studies (see for example Bucher, 2018;Haider & Sundin, 2019;Noble, 2018).…”

“…Even though IMMS does not collect any data, it manages data from the library system in a new way through machine learning. The use of machine learning provides a more generalized decision-making (Coleman, 2020). There is a tendency to view this as less biased (Coleman, 2020, p. 11) although it is important to note that the parameters are human made meaning bias will always be present (Haider & Sundin, 2020;Noble, 2018).…”