CELPIP Reading Part 4Reading for Viewpoints Practice

Library Network Tests Algorithmic Reading Suggestions

Evergreen Reach Library Network is asking whether it should generate opt-in reading suggestions from individual borrowing histories. The immediate reason is member requests for help navigating a growing digital and physical collection. Librarians now create general lists and provide individual advice only when members ask. What appears to be one operational choice has implications for library members, librarians, parents, and privacy and literacy advocates.

Elena Kuiper, the network’s reader-services librarian, makes a qualified case that optional suggestions could help members discover overlooked works without replacing human advice. Kuiper reaches this view because many users browse remotely and may not know the terms needed to describe what they want next. “A recommendation can be an invitation, not a judgment about the reader,” Kuiper says, adding that members could reach smaller publishers and older titles that broad popularity lists rarely feature. Kuiper attaches one important qualification, namely that users can reset the profile, hide individual loans, and turn personalization off easily. Kuiper also notes that recommendations should explain why a title appears so users can correct a mistaken pattern.

Omar Velasquez, a library-privacy advocate, challenges the proposed method even while engaging with its goal, arguing that retaining detailed histories changes a temporary transaction into a lasting record of private interests. Underlying Velasquez's position is the claim that people may explore sensitive or unfamiliar topics precisely because library borrowing is normally discreet. “Discovery should not require a permanent portrait of curiosity,” Velasquez says, adding that members could avoid certain materials if they are uncertain who may later access the profile. To preserve the underlying goal, Velasquez recommends creating on-device suggestions from user-selected titles without storing full histories centrally. Velasquez adds that a local-data approach may reduce privacy risk but provide weaker suggestions across several devices.

Before a permanent decision, Louis Morel, the network’s digital-services manager, wants a six-month opt-in pilot using user-selected seed titles, short retention, and curated discovery prompts. That approach would track opt-outs, profile resets, catalogue variety, privacy questions, and use of human advisory services. Kuiper welcomes the broader discovery tools, while Velasquez wants all personalization processed on members’ devices. The library board will publish a data map before deciding whether the pilot can begin.

Keziah Moran, a children’s literacy coordinator, agrees with the goal but broadens the test, arguing that recommendations based mainly on past choices can quietly narrow rather than expand reading. The practical basis for Moran's view is that young readers benefit from surprise, challenge, and exposure to books beyond an established pattern. Moran would implement a plan based on mixing personalized results with unexplained discovery selections and age-appropriate controls, expecting that diversity prompts and librarian-curated interruptions could keep suggestions from becoming repetitive. Ultimately, Moran shares Kuiper’s discovery goal but joins Velasquez in resisting a system that defines readers too narrowly. Moran cautions that parents and young readers need separate controls so a family account does not merge distinct interests.