Navigating the Maze of Article Discovery Presentation Transcript

Jun 24, 2020Transcriptions

Kael Driscoll:

Kael Driscoll and my paper is called Navigating the Maze of Article Discovery. I’m from the University of Western Australia. I’m a discovery librarian in collection and access services. My portfolio includes Primo VE, which we call One Search, Alma Resource Sharing, Alma Course Reserves and Leganto, which we call Unit Readings.

Kael Driscoll:

My presentation today is a story in six chapters. The idea: how could we improve article discovery for our users? The maze: navigating the complexity of access issues that appear every day, and which I’m sure you’re familiar with. The quest: from thought to reality with the article discovery project. The map: finding our way through the maze of the project to improve article discovery. The keys: what tools did we find to help unlock access for users? And the champions: once the quest was over, who were the heroes of our tale? I hope you enjoy the story.

Kael Driscoll:

Before we start a little bit of context about UWA. The university of Western Australia is in Perth and it’s situated on the land of the whadjuk noongar peoples. It was founded in 1911. It has 25,000 students, 1600 academic staff, 2,200 professional staff, and is a member of the Group of Eight universities in Australia. The UWA library is six libraries working together as one unified staff. We have 14,700 visits per day on average, 3.4 million Primo searches a year, 1.3 million e-resources are available and we have over 1 million print volumes, many of which are situated in offsite storage as our libraries form a very popular physical hub for our students and staff to research and study together.

Kael Driscoll:

So now we start the story with part one, the idea. It was an idea about improving article discovery. It really started in December, 2018 as a one minute pitch. What is one thing you think collection and access services should do? We all offered ideas at the time. Mine was one sentence long and it was promote Kopernio as a browser extension to our staff and students as a way to improve access to full text articles. That’s really all I said, I had one slide and one sentence. But what did I really mean by that? When I thought about it in a little bit more depth, I realized that it was about finding articles which can be like being lost in a maze, full of loops and dead ends and confusion and you can feel like you’re in a labyrinth of information. But at the same time, there seemed to be new tools appearing that were linking users with full text articles in new ways. And I wanted us to be agile enough to explore these technologies.

Kael Driscoll:

So that brings us to the maze. I needed to demonstrate that there was confusion in certain times when people were looking for full text articles. And I think it’s a place we’ve all been at some stage in our library careers. And I hope that this demonstration just gives you a small taste of some of the issues that we can face.

Kael Driscoll:

Okay, let’s say in this scenario that I’m a researcher who’s been talking to a colleague about a certain article that I really need to read from European Psychiatry, 2005 by Nicola Rouche. They knew the title, they also knew that there was a DOI and possibly that it was available through PubMed, but they weren’t sure. Being a good library user, I decided to start with our discovery layout and went to One Search and searched for the article. The great news is that it came up first in my search. So I knew it was peer reviewed, open access and had all the right details. So I followed the link to available online. I can see here that it’s available in full text from Elsevier from 2000 onwards so it should cover what I need.

Kael Driscoll:

Unfortunately, what I find is that the page I was looking for has not been found, which is a dead end and a bit unfortunate as a starting point. That’s okay. I won’t give up. I know that from the One Search record, I am able to get the DOI and so I decided I would go to Google and do a search for the DOI with PubMed. Now it did give me a PubMed result and that sounds like I’m on the right track. So when I went to the PubMed page, I could see that it was again the correct article, I couldn’t actually see any full text link there. So I had to look around and realize, okay, well there’s a link to the DOI so maybe that will take me to the full text.

Kael Driscoll:

Now again, I’m starting to get a little bit frustrated because it seems like I should have access to this article and I can’t work out why. Okay, this is looking good. Cambridge, access is provided by the University of Western Australia. It’s the right journal, it’s the right article. Everything’s looking great. Again, I still can’t see an easy link to the full text. I’ll try the DOI and see what happens. And now I seem to be going in circles.

Kael Driscoll:

Okay. So again, my frustration is starting to rise, but I’m not going to give up yet. I know that Google Scholar might be a great place to go. So I go back to Google Scholar and I try the article title to see what happens. Again, straight away, I can see it. Can’t see any full text link here, but I can see two over here. So I know that Science Direct, hm, had been there and it does say that it’s HTML, I’d rather get PDF. So I’ll try find it at UWA, that should link me back and we’ll see, maybe I’ll get the full text this time. So as it’s loading, I start thinking about, what other options might I have? Oh, well this is looking good. Elsevier, Science Direct, does look familiar. I’ve been here before and now I’m back where I started. Okay.

Kael Driscoll:

I’m not going to report a broken link from this point, I’m going to go back to Google Scholar and out of desperation, I’ll try this link to Science Direct. And what I find is, much to my surprise, it’s not really HTML, although there is some here, but I can see that there’s a link to download the PDF. Couldn’t be that simple, could it? And here I have the article I wanted. There it is in PDF from Elsevier. So now I start to wonder, why did one search take me to Elsevier in the first place, but I wasn’t able to find this article? And why wasn’t Elsevier able to actually deliver it once I was there? So I start to lose confidence in the library’s ability to give me the full text that I need.

Kael Driscoll:

So we’ve come out of the maze. We’ve got what we needed, but I still feel confused about why it didn’t work in the way I expected. And that was really what I wanted to explore. So it became a quest. I had to move from the idea to the project. So the project timeline was set out like this. In December, 2018 was when I came up with the initial thought and pitched it in the end of the year presentation session as a one minute, one sentence idea. It really needed to be explored in more detail. In January, 2019, the project was formally endorsed by Executive to be part of the library operational plan with the Project Team and Project Board, so I had to start writing what the project was actually going to be. The aim was that by December, 2019, we wanted to present recommendations to the Project Board with recommended technologies or tools that we could use. And we planned to implement them from January to June in 2020.

Kael Driscoll:

So we really needed to go back to work out, what did our users want when it came to article discovery? And we were able to look back on all the surveys and feedback that we’d received over the years. And so that was our starting point. We knew that our users wanted less clicks. We had to find a way to reduce the number of steps required to access full text articles in one search. Less barriers, users wanted to decrease the incidence of broken links, dead ends and loops through different publisher sites as we saw in the maze demonstration. Increased presence, they wanted the library to be integrated into their preferred discovery spaces, to provide clear links to subscribe resources. And confidence, users wanted to be assured that they will find what they need in a wide variety of online situations in the simplest way possible.

Kael Driscoll:

So we took those main points and it kind of suggested two main paths for us to explore. We had direct linking technology where full text in Primo would be available with less clicks. And we had browser extensions, which were tools for the user at the point of need integrated into their workflows. So to get through the project, which became a maze itself, we needed a map. So really the article discovery project had four main parts and these were the products of the project.

Kael Driscoll:

First of all, we needed to scan the field to find all the potential tools and services in the rapidly changing environment and investigate the viability. We needed to categorize and select so we could arrange everything by the type and the functionality, integrations and the possible costs, and then had to decide which products to pursue. We needed to explore and test these products by setting up product trials, test out integrations and do some user experience testing with library staff and library users. And then hopefully, choose and implement. So we wanted to recommend the best options, make a business case for the budget, if it was required, and liaise with the vendors about implementation.

Kael Driscoll:

So first off was the environmental scan, which actually took about six months because it was a very changeable area. We ended up with two main options for direct linking, five options for browser extensions, and also a new set of tools that we hadn’t really considered, which were hybrid tools that kind of had parts of direct linking and parts of browser extensions combined in terms of being a bridge between your subscribed resources and citations. There was three tools there that we also took into consideration. It was a very rapidly changing field and over that period, we found that it was moving in all different directions and we had to stay very focused and very agile and keep changing our framework slightly to make sure we were covering everything that was available.

Kael Driscoll:

So we had new products appear with new vendors and mergers between previous products and vendors. We had new features appearing within products. So there was different options, different functions and changed ways. And in some of the cases, there was actually reduced functionality for some of the original options that we chose to look at. And there was also unknown quantities. There were products that were basically vaporware and had a lack of documentation. People seem to be using them, but we couldn’t find out anything about them. So we had to make sure we were looking for things that were viable.

Kael Driscoll:

So it really came down to finding the keys, which were the outcomes of our exploration and allowed us to unlock what we wanted to achieve with the project. We had two major recommendations. We had the Third Iron products, Browzine and LibKey, and we also had Google Scholar and the browser extension, the Google Scholar Button. And we found that these two definitely came out ahead once we’d done a lot of the work. However, we also found that over that time, the Third Iron product LibKey Nomad really developed into a very strong browser extension and so we added that to our list of recommendations.

Kael Driscoll:

So now we’re kind of at the end of the quest, we’re looking at the champions, which are the tools we’re now using and promoting from everything we looked at. So LibKey, which is a suite of integrated tools, including Browzine, LibKey Discovery, LibKey Link and LibKey Nomad. So it’s all based on the Third Iron browsing technology. We were really interested in LibKey Discovery to start off with, but the other parts developed throughout the project and became integral to what we wanted to achieve.

Kael Driscoll:

So to look at the LibKey functionality in more depth, the reason it works so well for us is that there was DOI matching. So rather than using link resolver, metadata strings, and open URL, LibKey uses the DOI, or in some cases, the PubMed ID, as the match point to increase accuracy and improve reliability. And this has definitely been the case. Because we upload our library holdings too, it’s going to be cross-checked and LibKey can decide on the best source for delivering the full text directly. So this means that it will prioritize or preference publisher sites over aggregators, for example, because the metadata allows it to be delivered directly in a more reliable fashion. And there is the full integration in that LibKey Discovery creates direct links to PDFs in Primo. So that’s the direct linking. There’s LibKey Link, which integrates with native databases to show PDFs. So that’s bridge into the researchers space. And LibKey Nomad provides the point of need access to PDFs through the browser extension. So it really does create a holistic approach to article discovery.

Kael Driscoll:

It’s been very popular and seamless with our users since we implemented in February, 2020. So from March to July, we had nearly 218,000 full text downloads via our LibKey direct link, which has an average of nearly 1500 articles per day. And during our busiest main teaching period from March to May, the average was over 1800 full text downloads from LibKey links. The other amazing thing is that we’ve had virtually no complaints or problems with it. It’s fitted in seamlessly and our users. We didn’t do any education about the LibKey Links, users just see the link and they use it.

Kael Driscoll:

Now, Google Scholar has also proven itself to be very popular and we knew it was a very popular space with researchers and we knew it had strong search power. When we explored it a bit more, we found that the Google Scholar Button has a browser extension that was very good in that you could highlight, say, a DOI or a title within a webpage and the Google Scholar Button will allow you to search for it straight away. It can be linked to UWA resources by users choosing UWA as their preferred library. So there is access to subscribers direct from Google Scholars. It also has the ability to find articles that others can’t, and I don’t know how to explain it, but sometimes there are just certain hard to find things that Google Scholar is able to tell me I have access to through UWA, but I can’t find through any other means. And because of the LibKey integration, it does provide direct linking now to our finders service, rather than returning to One Search. If it can, it will actually show the PDF straightaway using the LibKey technology.

Kael Driscoll:

So in conclusion, this was our path through the maze. Less clicks, so LibKey discovery links directly to PDFs from One Search, from the brief results page and it also now integrates with Leganto, which is fantastic for students using unit readings. Less barriers, the broken link reports have reduced and users don’t need to sign in as often when navigating resources from off campus, which has been fantastic. Increased presence, LibKey Link has replaced Find It in databases and LibKey Nomad connects users to own content. So like PubMed, LibKey Link has now expanded to all of the major databases that we use and so there’s more access to direct content. And confidence, we’re really going to be looking for increased satisfaction with One Search and access to online services and resources in our next user survey. And that will be the real measure of whether this article discovery project has been a success with the users. We feel from the anecdotal evidence and just the seamless way that people are using it, that it’s already been a success, but we really want to look for increased satisfaction.

Kael Driscoll:

Just wanted to very quickly give credit to Roger C Schonfeld who’s dismantling the stumbling blocks video that I watched before I did the project, was a huge inspiration to me in terms of showing the blocks that can impede researchers. And also thanks to Aaron Tay for his musings about librarianship blog. During our environmental scan, we were using his input about the six broker access extensions that he was looking at doing his own environmental scan and sharing with everyone. And it was great to be able to compare our findings with Aaron’s. So thank you to Aaron Tay as well. Thanks to the Project Team, Erin Montagu, Katherine O’Brien, Kirstie Nicholson, and thanks also to Kendall Bartsch and the Third Iron team who have been a great support during the project. There’s my email address if you’d like to contact me. Thank you for listening and it’s time for questions.

François:

Thank you. Thank you, Kael. I hope you enjoyed the session. I saw some questions. You saw them also, Kael?

Kael Driscoll:

Yes. Yeah. I’ve got a couple of questions here. Yeah. Thanks everyone for attending. It was very strange listening to myself. I’ll just say, we’ve had one from Tamar here, I hope I’m saying that correctly, “Our IT strongly objects to browser add-ons due to instability and security breaches.” That can be an issue. The way we were really focused on this with the browser extensions was about people who are using their own computers and own laptops they were bringing to university, but we were also able to talk to IT and they were able to make it part of our image if you like on all the library computers and they dealt with all the security. So I can’t talk to that too much, but they certainly weren’t concerned about those particular browser add-ons and we didn’t need to go through a lot of work to get them there. So maybe I can follow up with that by email.

Kael Driscoll:

“Is there an issue with content not covered by browsing?” That’s an interesting one. Probably would need some time to do it. It just means the content is still there, but it just doesn’t have the direct linking. So I probably would need to go into that in more detail to really answer that well.

François:

I think concerning browsing and LibKey, you need the DOI and it must be Crossref or provided by Crossref, I think.

Kael Driscoll:

Yeah, the DOI. Yes. Sorry, you’re right. Yeah, the DOI. I think we can certainly, Kendall Batsch and Third Iron are presenting as well so I would highly recommend checking out their presentation at this conference. And Kendall has been great help, just contact him and he will help you with everything you need to know.

Kael Driscoll:

Someone very kindly asked if I could do a quick demonstration, but I won’t be able to do that, but I may be able to record a short video and email it to anyone who’s interested. There’s also one here about usage statistics. That’s actually been really good. So BrowZine have produced lots of statistics for us. In terms of authentication, we’re using EZproxy, which is currently on the way out and we’re hoping to move to OpenAthens. Time-consuming for cross-checking practice? I’m not sure about that. I’ll have to come back to that one.

Kael Driscoll:

And Jacob says DOI and PMID, but it doesn’t solve the example you showed us in One Search. Yes, that was a little bit of a trick because I had to kind of use one that I knew wasn’t going to work, but there was no way for me to actually solve it. The way that one was actually solved was by contacting Third Iron and they were able to fix it for us. So that was well picked up by Jacob that my example was a little bit of a lie because it wasn’t solved, hence why I didn’t show it being solved at the end. But one of the things to say for it was it was very hard to find an example where it didn’t work since we’ve had LibKey. Whereas before we had LibKey, I had thousands of examples I could have used for you.

Kael Driscoll:

So that’s probably my time up, but thank you for all your questions and I’ll certainly get them in text and I can send out some more stuff as well. So yes, we did look at Lean Library. I’m not sure when Third Iron are presenting and yeah, Martin, I will do a demonstration. And if you guys contact me, well I’ll have your details, I’ll send it to everyone who’s asked a question as well. But thank you so much.

François:

Thank you. Thanks to you. Thanks for the questions. Anyway, as I said, I will send all questions to Cal so you can collaborate on them, no problem.

Kael Driscoll:

Yeah, definitely.

François:

Was it easy to convince your colleagues to promote LibKey Nomad especially? Because it’s an add-on, it’s not integrated it into Primo. So also in my institution, I’m not sure that I was able to convince many of my colleagues to enable the add-on.

Kael Driscoll:

Yes. So our promotion of Nomad has only really started in earnest in the last couple of months. So we’re really trying to see that now with mainly the COVID situation kind of changed our focus. We had to stop for a little while, but it will be really interesting to see over the rest of this semester and next year in particular, now that we’ve kind of recommended it on our websites and that’s part of the information literacy, but yeah, it’s being advertised at the moment on all our screens around the library and yeah, so far we’ve had nothing but positive feedback.

François:

Okay. If you can make a small demo later, a video, we will also make it accessible on the Eagle website with a link.

Kael Driscoll:

Yeah, for sure. I’ll just do another quick Zoom video as an MP4 so people can watch it. Yeah. Yeah. Great idea. If I’d had a half an hour, I would have done some more demo.

François:

Maybe next year.

Kael Driscoll:

But yeah, no, I really appreciate everyone attending and especially the questions, I’ll make sure I get in touch with you. And yeah, I’m very happy to make some more networking contacts with everyone. So thank you so much.

François:

Okay. Thanks to you, Kael, and thanks to everyone who attended, 110 people. It’s more than expected.

Kael Driscoll:

Yeah, amazing.

François:

Yeah. Thanks everyone. Bye-bye.

Kael Driscoll:

Thank you.

François:

Bye-bye. Thank you.