Emma Booth (she/her)

https://tinyurl.com/yc4vzcpw

@cilipmdg.bsky.social invites participants to a WorldCafe Event @wellcomecollection.bsky.social on the topic of Subject Analysis & LCSH within the current metadata landscape.

Wed 26th November 2025 11:00-16:00
Booking essential: tinyurl.com/yc4vzcpw
#cataloguing #metadata #event #critcat #ethics

Library & Info Studies/Science students & early-career professionals can apply for a #sponsoredplace at the MDG Conference #MetaFutures
Sponsorship covers the cost of a Full Programme Ticket & reimbursement for travel & accommodation costs (max £500)
Full details at: www.cilip.org.uk/members/grou...

https://www.cilip.org.uk/members/group_content_view.asp?group=201298&id=1113783

📢 Registration is now open for the MDG Conference 2026!
#MetaFutures will take place on 11th-12th March 2026 at Engineers' House in Bristol followed by the UKCoR RDA Day on 13th March 2026. 📅
Early Bird rates available until 03/11/25
Full details and registration at: www.cilip.org.uk/members/grou...

Our next Practical Cataloguing Webinar is "Linked data, Ontologies & Entity reconciliation" with Fran Frenzel, 13:00-14:00 (GMT+1) on Thurs 16/10/25.

Register at: www.cilip.org.uk/events/Event...
#metadata #linkeddata #cataloguing #training

I've never done 'live cataloguing' over Teams to 1 person, let alone 90 people! Kudos to Will for giving this live demo.

We are now working through cataloguing another example publication into a template MARC21 record.
This is the 'practical' element of the webinar; attendees are encouraged to create their own record using the tools and information provided at the outset.

A fully useful and descriptive MARC21 record for a modern print book will provide sufficient data points for discovery and access. It should enable library users to search for, find, identify, select and obtain/access the resource. Plus explore other related resources (via authorised access points).

As we work through the practical example, Will is explaining the different MARC fields, and when the cataloguer must transcribe from the resource, and when they record information following metadata standards such as RDA www.rdatoolkit.org/about

Controlled headings enable the catalogue to link together publications by the same individual.
Name disambiguation is particularly important and cataloguers in the UK often use Library of Congress Name Authority File to apply Name Headings. authorities.loc.gov

Will has also referenced the external sources that cataloguers use - the tools in the cataloguer toolbox - including using controlled vocabularies and authority files in order to create Authorised Access Points.

Will has been working through an example MARC record for a particular publication; working field by field he is explaining the notation and entry conventions. The what goes where and how it is recorded of library cataloguing with MARC21!

BlueSky is glitching today from the MDG account! - I'll complete the thread using my personal account...

I'll be doing some selective live-posting for this webinar over on the @cilipmdg.bsky.social account. Follow this 🧵if you're interested in some insights into the first of MDG's Practical Cataloguing webinar series: "Introduction to MARC21 Cataloguing"
#metadata #cataloguing #MARC21 #libraries 📚🐈

Pages 1-2: Editorial by The Editors. Pages 3-11: ‘We are here not because we are metadata-breakers; we are here in our efforts to become metadata-makers’ by Helen K. R. Williams. Pages 12-25: Envisioning Dante by Ourania Karapasia. Pages 26-33: Decolonising heritage collections by Jessica Roberts. Pages 34-44: Unchartered cells by Carol Hunter. Pages 45-48: The challenges of data ingest, transformation and aggregation at the National Bibliographic Knowledgebase by Jennie-Claire Crate. Pages 49-58: Non-MARC cataloguing by Anne Welsh. Pages 59-62: Book review: Ethics in Linked Data by Elizabeth Cooper. Pages 63-65: Book review: Records and information management by Sarah Henning.

The latest issue of our OA Journal 'Catalogue & Index' (ISSN 2399-9667) has now been published. Issue 212 (Sept 2025) contains articles on the topic of "Non-MARC Cataloguing and Metadata Practices."

Read the full issue at: journals.cilip.org.uk/catalogue-an...
#cataloguing #metadata #nonMARC

Librarians *should* be exercising more judgement and objective expertise, looking at the whole picture, thinking about long term & bibliodiversity. Libraries do not receive budget to "buy content" they receive budget to "disseminate knowledge".

Demmy name-checked @openlibhums.org and @openbookcollective.bsky.social as reasons to be hopeful in scholarly publishing. He finished by arguing that the future belongs to those who can discover new ways of publishing research outside of commercial models (I paraphrase). #OASPA25

Fab news Caroline!🎉

Practical Cataloguing Webinar 1: Introduction to MARC21 Cataloguing is this Wednesday 24th Sept 13:00-14:00 (GMT+1).
Free to all. Registration required: www.cilip.org.uk/events/Event...
#cataloguing #cataloging #metadata #training #event #webinar #libraries

OpenAI admits AI hallucinations are mathematically inevitable, not just engineering flaws In a landmark study, OpenAI researchers reveal that large language models will always produce plausible but false outputs, even with perfect data, due to fundamental statistical and computational limi...

"In a landmark study, OpenAI researchers reveal that large language models will always produce plausible but false outputs, even with perfect data, due to fundamental statistical and computational limits."

www.computerworld.com/article/4059...

Any #librarians following me should give @libraryleo.bsky.social a follow too - I know Elle from my time on committee with @uklibrariesnag.bsky.social - let's grow our BlueSky librarian community #Skybrarians #libraries 📚

How is the presumably vast energy increase from embedding AI into every single product and task and search being measured and offset, particularly considering Clarivate's goal to be net zero by 2040? #igelu2025

"...we use the Mistral LLM hosted via Amazon Web Services (AWS), which has pledged to be carbon neutral by 2040."

I am still frustrated at this being part of a response to staff over concerns about the carbon costs of AI at the library.

Are we really supposed to trust a pledge from Amazon?

On the topic of AI in libraries...

This could potentially be harmful as well as misleading so metadata professionals add important value here in assessing AI outputs to mitigate against these harms.

Even using the AI tool for 'factual based' tasks such as generating a contents note from a table of contents can lead to inaccurate results - the AI can format the information you extract and feed into it, but it does also make stuff up! Especially if you're asking it to do bulk processing.

Part of the rationale for the testing that Leeds undertook was to give feedback to Ex Libris about improving the AI Metadata Assistant for others across the sector. The tool may not be that useful for Leeds' cataloguing needs, but it could be useful for smaller libraries with fewer staff.

The environmental impact of AI and LLMs has also been considered by engaging with Ex Libris about the AI tool. - It is a complex balancing act to weigh up functionality/utility and time/cost/sustainability.

Leeds have considered the ethical considerations of using AI for metadata work - they have approached with caution using AI for subject analysis & ensure that trained human cataloguers take the care to ethically apply subject headings. The AI is used for the factual data rather than the subjective.

So far testing at Leeds has shown that the limitations of the content that this AI assistant can search and return to users means that it is not yet fit-for-purpose. Leeds have therefore elected to not turn on this functionality for their Library users.

Alison speaks briefly now about the Primo Research Assistant - an AI tool that works within the Library's discovery layer to support users with searching for appropriate resources for their research enquiry.