Program

Hackfest Organizers: Kyla Jemison, Metadata Librarian, University of Toronto Libraries Session Description: Good data organisation is the foundation of much of our day-to-day work in libraries. Most librarians have data or do data entry in spreadsheets. Spreadsheet programs are very useful graphical interfaces for designing data tables and handling very basic data quality control functions. This Library Carpentry session will combine lessons from Computational Thinking, Tidy Data, and Regular Expressions to help participants understand pattern recognition, identify areas where technical skills can help solve problems, and introduce some technical tools readily available to library workers to apply these skills. This will be a hands-on workshop, so participants should be prepared to work on their laptops, ask questions, and collaborate with peers.

• Hackfest

Hackfest Organizers: Tim Ribaric, Scholarly Publishing and Platforms Librarian, Brock University Library John Fink, Digital Scholarship Librairan, McMaster University Libraries Session Description: TBD

• Hackfest

Speakers: Mark Christensen (He/Him), Collections and Content Management Librarian, Simon Fraser University Library Session description: In 2019, SFU Library faced a tight six-month deadline to weed and/or relocate over 75,000 items, and shift about 150,000 items to make space for our Indigenous Curriculum Resource Centre. Traditional spreadsheets were too slow and prone to barcode errors to meet this goal. To solve this, Todd Holbrook and Mark Christensen built DandeLemur, a custom, browser-based tool that connects directly to Alma via an API. This tool and how we use it reduced the time required to deselect items by over 60%. Originally built for our specific project, the tool immediately became an essential tool for stacks maintenance. The DandeLemur was completely rebuilt in late 2025, with considerations for how other libraries using Alma may want to use this tool. Unlike spreadsheets, the DandeLemur provides a clear interface with real-time feedback. By integrating with Alma, it gives staff instant visibility into item status, enabling smarter, immediate decision-making. We will highlight how the tool’s design supports the ergonomics and workflow of library assistants: featuring larger fonts, automatic logic that marks the correct number of copies as "Keep" once items are scanned, and distinct modes for weeding, relocation, and locating. The tool also facilitates teamwork by allowing multiple users to work simultaneously without overlap, using features like ascending/descending sort splits and bookmarks. Attendees will learn how DandeLemur flags problem items, tracks missing books through multi-stage checks, and generates automated nightly reports. We will cover the simple setup process, user role management, and safe list uploading techniques. This presentation shares the story of why we built this purpose-built, user-friendly tool, how it helped us meet our deadline, and how it now serves as a cornerstone of our physical collection maintenance.

• Keynote Presentation

Speakers: Jessica Hymers (She/Her), Scholarly Resources and Metadata Librarian, Scholars Portal Julia Gilmore, Digital Projects Librarian, Scholars Portal Gabby Crowley, Communications & Outreach Librarian, Scholars Portal Alyssa Conlon, Research Data Management Librarian, Borealis and Queen’s University Library Session description: Scholars Portal provides shared digital infrastructure for Canadian university libraries and supports access to a wide range of scholarly resources including books, journals, and data. Increasingly, automated traffic and large-scale bot harvesting are reshaping how these services operate, driving the need for new solutions in response. This session brings together representatives from across Scholars Portal services to discuss how automated harvesting affects different types of scholarly infrastructure in distinct ways. While bot activity may target the same infrastructure ecosystem, the operational, technical, and ethical impacts vary significantly depending on the platform, its users, and the content being accessed. Speakers will discuss how automated traffic influences system reliability, resource allocation, licensing obligations, privacy considerations, and user experience across services. This session will also explore the broader tensions infrastructure providers face as they respond to this rapidly evolving challenge. How do we distinguish beneficial automated use from harmful harvesting? How do we balance openness and global scholarly access with obligations to Canadian institutions and users? What role do data sovereignty, privacy protections, and authentication models play? Rather than presenting a single “correct” approach, this discussion will highlight the values, tradeoffs and uncertainties shaping our current strategies. Speakers will share practical experiences, emerging mitigation approaches, and unresolved questions about the future of scholarly access in an era of increasingly sophisticated automation. The session will include audience discussion, inviting participants to share their own experiences and perspectives on balancing access, openness, security, and sustainability in scholarly infrastructure.

• Break

Speakers: Katya Pereyaslavska, Coordinator, North/Nord & Shared Print Archive Network Milica Ivetic, Data Analyst, North/Nord Erin Bourgard, Data Analyst, North/Nord Session description: The CUP project, delivered between 2023 and 2026, represents a foundational and highly innovative undertaking for North/Nord, marking its first foray into large-scale, in-house collection analysis. Drawing on institutionally supplied title lists from participating partners, the initiative not only enabled coordinated analysis and implementation, but also served as a critical testbed for developing new technical and operational approaches. To address the complexity and national scale of the work, North/Nord engaged a range of subject-matter experts to experiment with and refine emerging workflows, laying the groundwork for a more robust, scalable, and sustainable model for shared print analysis moving forward. The processing and analysis of the title lists provided by each institution relied heavily on the Pymarc Python library (https://pymarc.readthedocs.io/en/stable/#). Pymarc was designed to work with bibliographic data encoded in MARC21. Each institution provided its title lists as MARC binary files. Using Pymarc allowed us to avoid converting the data to different formats which mitigated the risk of information loss, and to streamline processing in Python. The data analysis was conducted using a mix of human- and AI-generated code. The code and workflow were tested and refined several times to achieve a desirable confidence level score. To ensure utmost accuracy levels, all planning, execution, and reviewing were done without the assistance of AI. This session will highlight next steps for North/Nord and adapting this approach to a Government Documents project.

• Presentation

Speakers: Santiago Gomez, Director of Technology, Edmonton Public Library Session description: Archives and museums hold remarkable collections, but much of that material stays in boxes or behind glass, seen but not always understood. At Edmonton Public Library, we're exploring how to bring some of those collections to life on our digital wall, and we'd like to share where we are in that journey, including the open questions we haven't answered yet. We're currently working with two partners on early-stage concepts: the City of Edmonton Archives, around material documenting the construction of six river valley bridges in the early 1900s, and the University of Alberta Museums, around Qing dynasty silk robes that are rarely displayed. The session covers two things. First, the collaboration playbook we've developed to guide this work. It rests on a simple division of roles: our partners are the experts on the meaning and accuracy of their collections, and we take on the technical and creative work so they never have to understand the technology. Together we define the intent of each exhibit up front, turn that intent into a handful of concept options, and score them openly with a rubric so decisions are fair and explainable. Second, the new generative AI tools we're evaluating for production, including text-to-3D and image-to-3D mesh generation, video, audio, and image tools that could let a small public library team turn archival photos into interactive 3D experiences. We'll show early experiments, like 3D models generated from century-old bridge photographs, and talk honestly about where these tools impress us, where they fall short, and the quality checks we believe are needed for accuracy, cultural appropriateness, and bias. Attendees will leave with a realistic picture of what these tools can do today and a practical framework for partnering with institutions.

• Presentation

Speakers: Jonathan Younker (He/Him), Head, Library Systems & Technologies, Brock University Library Pranjal Chakraborty, Library Systems Analyst, Brock University Library Session description: Instead of paying for a commercial solution like Occuspace to capture and display real-time library occupancy data to our users, we decided to create an in-house solution, using anonymized Wifi statistics, to display on our website and throughout the Library on signage displays. We will cover the idea, the development, how to replicate our solution, and technical details. Having had this solution operational for over a year, we will describe how we are visualizing and using this data to inform our users and our library decision making processes.

• Presentation

Speakers: Amanda Buettner (She/Her), Electronic Resources Librarian, Seton Hall University Libraries Session description: This session will explore how Seton Hall University Library utilized a data-informed methodology to re-examine our big-deal eBook package with ProQuest. Large eBook packages promise bounty and volume, but the evidence to suggest whether this abundance translates into high usage or value is murky. With rising eBook subscription costs and the loss of perpetual access eBook purchasing on the ProQuest platform, SHU decided to analyze the value of our ProQuest Ebook Central package. To weigh the potential loss of resources to our users against the financial benefits of not renewing the package, the project used a three-phase plan of vendor eBook package overlap analysis, print title overlap analysis, and usage analysis with subject librarian input. Our eBook package holdings list, print holdings lists, and publisher supplied holdings lists were obscured with messy and non-standardized metadata, leading to the need for a data cleanup workflow. Key to the project’s success was use of Microsoft Excel’s Power Pivot and Power Query to normalize and analyze holdings metadata. Phase 1 used data cleaning and modeling tools from Excel to compare titles in ProQuest Ebook Central against packages from JSTOR, EBSCO, and Oxford University Press, revealing a 56% overlap. Phase 2 applied a similar method to identify unique ProQuest-only titles and match against our print holdings, resulting in a 4% overlap. Phase 3 pulled in usage data to compare against the unique ProQuest titles and combined these data points with title evaluation by liaison librarians. By revealing considerable title overlap between vendor eBook packages and low usage of the titles relative to the package’s price and volume of titles, the library was able to make an informed decision to drop the expensive and largely unremarkable ProQuest Ebook Central package. This project demonstrates an accessible approach to large-scale eBook assessments, leveraging technology already available and familiar to librarians. The phased data analysis, cleaning, and evaluation offers a replicable workflow for data-informed collection management in academic libraries.

• Presentation

Speakers: Christopher Ross (He/Him), WordPress Developer & Accessibility Consultant, This Is My URL Session description: Every library runs web systems that have to be accessible: the public catalogue, the digital collections, the institutional repository, the WordPress site behind half of it. Increasingly, that accessibility gets audited for real under AODA and the Accessible Canada Act. The work usually shows up too late, after the build, when a compliance review is weeks out and the remediation scramble begins. I build and audit web systems for education, government, and media clients where that review is real and the failures are expensive. This is a technical, practical talk about what those audits actually check, and where library web systems reliably fail them: generated markup that looks fine and tests broken, third-party discovery layers and widgets that ship inaccessible, dynamic content that passes a manual review and fails a keyboard one, and the PDFs nobody tagged. In twenty minutes you'll get the specific failure modes an auditor will find, how to test for them yourself before they do (the keyboard pass, the screen-reader pass, and the automated tooling that genuinely helps versus the kind that lies to you), and the one build practice that turns an accessibility audit from a fire into a formality. No "why accessibility matters" preamble. You work in libraries; you already know. This is the how, from someone who has sat on the wrong side of the audit and learned to build so it doesn't happen twice.

• Presentation