A Government Librarian at the 2024 UKSG Conference: Part 2, by Naeem Yar

 

Naeem Yar is a Librarian with the Welsh Government and serves as GIG’s Events Co-ordinator. This is the second in a three-part series of reflections on the 2024 UKSG Conference in Glasgow, which are written in a personal capacity.

Both sentimental keepsake and practical utensil - as far as prizes go, I've had worse (Naeem Yar)

After having a merry – but fruitless – time at the UKSG quiz on Monday night, I woke up the following morning slightly the worse for wear: not in a bad enough state to justify ingesting painkillers, but rough enough to wolf down a cooked breakfast without much guilt and guzzle plenty of water. Then, I trudged to the SEC Centre alongside one of my quiz team-mates, who was staying at the same hotel. Whilst making myself decent in my room, I gazed at the wooden spoon which was the prize for our joint last place finish and reflected on the evening. Our cheerful attitude to our lack of success seemed a good metaphor for a willingness to take a risk on experiments that do not quite work out and to learn from them – a theme that seemed to recur at various times throughout the conference.

 

I made sure to drag myself to the main hall in time for the second plenary, focused on open access publishing and in particular the contours of change that have occurred in the prevalence of different OA models in recent years. This is of particular importance for our library: like many libraries, we need to spend our money carefully and open access publishing is an important channel for enabling our user base to access the publications they need to ensure that the Welsh Government’s activities are informed by high quality peer reviewed evidence. It is therefore important for us to understand where the winds are blowing with regards to OA and which models are becoming more widespread. In part so that we know where to look for OA publications and can pass this knowledge on to our users, and to be able to understand the threats to the viability of OA publishing so we can factor this in when planning our acquisitions of resources.

 

The panel at the Open Access plenary shortly before it kicked off, flanked by Conference Planning Subcommittee chair Josh Sendall to the right (Simon Williams)

The panel featured speakers from the US and the UK, and discussed the OA publishing situation internationally, but there seemed to be some broad trends which were discernible across contexts. Overall, OA publications seem to have increased, primarily driven by an increase in hybrid OA with gold OA stagnating and green OA declining.

 

The increase in open access publications is positive for the availability of research, however some of the shifts between the different OA models have implications for the findability of publications. As noted by the Francis Crick Institute’s Beth Montague-Hellen and the University of Nottingham’s Katie Fraser in their talk on green OA (slides available via the University of Nottingham repository), repository-archived have issues with regards to discoverability, so a decline in their prevalence in favour of publisher-hosted forms of OA may seem to be an advance in terms of ease of finding publications. However, the growth of hybrid rather than gold OA still means that users without institutional or personal access to a journal would need to sift through papers published in it to find the articles that they would be able to access. This is particularly the case for a small library like ours. It does not have a catalogue with thousands of article-level records provided by external suppliers and can integrate metadata indicating the open access status of articles, and in some cases tools such as Unpaywall, which can provide links to OA articles, including repository-archived ones (though we do subscribe to databases which include such tools).

 

A growth in gold OA would be more helpful in overcoming barriers to discoverability for society as a whole, as it would strengthen the comprehensiveness of indexes like the Directory of Open Access Journals (DOAJ), which allows for searching of articles as well as journal titles. Yet as Caren Milloy noted in her presentation on Jisc's recent review of transitional agreements in the UK (slides of her and co-presenter Chris Banks’ talks are available via Slideshare), the pace of transition to gold OA among journal titles has been sluggish and at the current rate it would take the Big Five academic publishers 70 years to fully embrace gold OA. In addition, publishers currently lack concrete strategies to achieve such a goal.

 

The overall impression from the panel was that the OA movement is at an impasse and that new models are needed to maintain momentum and ensure that OA publishing is sustainable. Many of the speakers talked with an impressive frankness about their views on the extent to which publishers’ current practices were part of the problem – something I had also noted in the research integrity plenary. A cynic might think that there might be pressure to “play nice” at a conference that brought together academic librarians and academic publishers, but that clearly was not the case. As someone who is all for openness and honesty, I was there for it – though I did wish I had brought some popcorn.

 

Following this, it was time for another workshop in which Judith Carr and Rachel Bury of Edge Hill University led us through an exploration of tools to support research workflows. After an introduction that discussed the research project that inspired the session, participants got started by identifying the questions researchers would want answers to regarding a research tool before they would consider using it.

 

Our table engaged in a lively debate identifying plenty of salient points, including whether stored data would be portable. Particularly important for paid-for, proprietary products, especially if a researcher needed to move from one institution to another – whether it facilitated collaboration, and its interoperability with other tools the researcher uses. My own favourite (not one I came up with) was “does it solve a problem” – almost certainly the most important question to ask of any new product, as it is likely to define whether someone is willing to invest time and effort in learning to use it. As someone who considers themselves a techno-agnostic (or to put it another way a technological pragmatist – I am happy to adopt a new technology if I feel it is actually useful), I could really identify with this question. It certainly strikes me as something to always bearing in mind when promoting any product.

 

The groups were subsequently shuffled and we were tasked with applying some of the criteria to a range of different tools. Each table were given details of tools for a different stage of the research cycle – ours focused on discovery. After some questions about what we were being asked to do, we decided to pair off to examine the products on our list. Some of the sites included under the “discovery” heading could be seen as not being discovery tools as such or as having a more roundabout relationship to the discovery process. For example, Scholarcy, which can summarise text, would make it easier to get the gist of a publication, but you would already have to have found the publication you want to summarise before using it. To their credit the facilitators were happy for us to express such views: there were not any wrong answers.

 

Our newly rearranged group picks up the threads of the task: "So, what exactly is it that we're meant to do...?" (Simon Williams)

Many of the other sites we looked at were AI-enhanced academic search engines such as Perplexity and Consensus, which can process questions posed in natural language and return quality sources in response (and in Perplexity’s case, synthesise a prose answer using an LLM in the manner of ChatGPT). Tools like this have a clear appeal to those folk outside the library who are not search practitioners by profession. They do not require the user to learn a particular search syntax or set of commands (though of course prompt engineering is as much of an art as designing a good search strategy), whilst also being more discerning than conventional general search engines in terms of material retrieved. On the other hand, their algorithms are opaque and the grounds on which they have selected particular publications are not obvious. I think there is definitely a place to promote tools like this to those looking for research and evidence, such as our users. But, as ever, it is important to present a rounded picture, highlighting the pros and cons, unpacking their workings or at least explaining how they are difficult to unpack, sharing how the user can exert more control on the results, and encouraging patrons to use them as part of a range of diverse information seeking tactics to obtain a balanced set of results on a topic.

 

Not all of the sites listed were free of charge and the fact that we were being asked to evaluate them without being able to use them first led to some interesting discussion about the practicalities of familiarising yourself with a resource before recommending it. The facilitators rightly acknowledged that many of us are busy and may not have time to test out a site personally, so it may be reasonable to signpost a tool to others in the hope that it may be useful to a patron and that they can experiment with it and find out for themselves. I can understand the logic, but I would be reluctant to advocate a resource without having at least a little firsthand knowledge. I am in no doubt that researchers are also extremely busy, and if they do go out on a limb to try something new, I am sure they would prefer there is a good chance it will actually help them. Not cultivating knowledge about a tool that you are advocating for does seem potentially counter-productive, undermining claims of expertise and potentially damaging customer relationships and possibly leading to negative word of mouth.

 

Speaking of word of mouth, one of the search resources listed was a paid-for site which I think would be useful to us in the Welsh Government. One of the other delegates in my group had access to it via their institution and raved about it, which was useful feedback to take back home with me.

 

Following another excellent lunch, we had another plenary which I had been particularly looking forward to – the intriguingly titled “There is No List” (slides available via Slideshare). The panel taking part in this discussion reflected on ways of teaching others how to avoid predatory publishers and whether lists of journals or publishers have a place in this. In my view, lists have their place, but I am interested in methods that go beyond simple heuristics that effectively delegate the task of exercising critical faculties to an external authority and that instead help to develop the individual’s ability to evaluate a journal by themselves. Katherine Stephan of Liverpool John Moores University and the Think. Check. Submit. initiative to help researchers choose appropriate outlets for their work outlined some of the reasons to take this approach: it empowers people to make choices for themselves, whilst the compiling of lists opens questions of power around them. Who compiles them? What process do they use? These questions were among the issues that engendered criticism of Jeffrey Beall's blacklist of predatory publishers. Biases towards the developed world and the English language have also been noted.

 

The predatory publishing panel answer questions - Katherine Stephan taking her turn (Simon Williams)

Another view was put forward by Cenyu Shen of DOAJ, who noted that a blacklist could never be comprehensive. New predatory publishers were constantly being established, but it was argued that the Directory was a wise choice for a whitelist due to the rigour of its evaluation process and the transparency of its inclusion criteria, which constitute a gold standard for OA journals. I have some sympathy with this argument: if journals are going to be listed, the carrot of being recognised for high standards seems to be a better incentive than the stick of being blacklisted: an approach based on critical skills seems to be a better way of steering people away from dubious publications, for the reasons outlined previously. DOAJ also has the advantage of a diverse multinational and multilingual set of editors, which should help address issues of bias.

 

There were also parallels with the discussion on research integrity at the first plenary with factors such as embarrassment and cultures of blame, which make it challenging to ensure researchers, who had fallen victim to predatory publishers, could be supported and openness and transparency being key to facilitating this. Also, the discussion of the pressures on early career researchers to publish and the need for research assessment processes to encourage good behaviour.

 

After this it was time for another breakout session on a topic touched upon in one of the breakouts I had attended on Monday. In “Demystifying AI: (Data-centric) Uses and Limits for Library Collections”, Siobhan Haimé of the Open Library of Humanities explained what AI is, what types of AI there are and the ways in which generative AI has changed the game, her own experimentation in applying AI to collections management tasks as well as important considerations and tips on undertaking such work successfully. I think this was one of the best received sessions I attended at the conference. It explained the fundamentals of AI in a non-technical way, and also outlined the nuts-and-bolts of how it could be applied in a library context whilst it also stressed the importance of ensuring the right solution was being chosen for the problem being tackled. As a techno-agnostic I was pleased that Siobhan had a healthy sense of realism about AI, pointing out that a viable business model for the AI sector is yet to be established, and that there are environmental and copyright challenges that also need to be resolved. Her sense of humility in answering questions about her work and other possible library applications which were further away from her own library experience was also admirable.

 

By the end of the breakout session, it was 4pm. Those who had previous experience of the conference had warned me that it could be an intense experience for a first timer, and I was starting to feel pretty tired at this point. Although there was another breakout I had wanted to attend, I decided to retire to my hotel room and give the evening’s gala reception a miss – which was just as well, as I ended up falling asleep at about 8pm.

 

Don't know much about this lot, but they do look a bit like they were booked for the cabaret at a holiday camp and took a wrong turn. I will say that I could have done with an ounce of their energy though: several turquoise suits were seen partying well into the night at the UKSG disco (Simon Williams)

And on that sleepy note the second day was brought to a close. If my musings have not sent you to sleep yet, be sure to keep your eyes peeled for the third and final post in this series, where I will cover: the cutting edge of integrating generative AI with searching for academic information; what happens when you use gamification and storytelling to teach the history of DOAJ; how a small publisher and their working partners are collecting and enabling use of author identity metadata; and why you need to lead yourself in order to lead others, as well as some closing thoughts.