Tuesday 31 March 2015

Rachel Lammey - CrossRef Text & Data Mining Services: one year in

The background to this session:-

"The field of Text and Data Mining (TDM) is growing in importance with an increasing number of researchers interested in mining scholarly content. CrossRef Text and Data Mining Services launched in May 2014 and focuses on providing one common way to retrieve the full text of articles for the purposes of TDM for interested parties. This session will provide an introduction to and update on this service, and a short demonstration of it in action". 

"This is an introductory level talk" said Rachel, but indeed, this slide is great for people new to text & data mining.






This slide includes links to Announcing the PLOS Text Mining Collection &

Text mining: what do publishers have against this hi-tech research tool? by prolific science reporter/blogger Alok Jha.


 


Most in the room were aware of CrossRef and their services so as such, Rachel skipped through the opening section of her slides which can be found here.











An explanation was provided as to the importance of DOI's. 

CrossRef has been set up to do things that publishers don't do on an equal platform, they currently have. 27 - 28 staff.

May 2014 was the launch of CrossRef's TDM services.

There are a lot of technical components involved in the TDM landscape/industry/publishers.





CrossRef built a cross-publisher API for TDM'ing.

The live demo did not work as planned (the joys of doing a live-demo) But we captured some of this section on camera, as you do..



Negotiations / Permissions 




API Token



Publishers can upload their own T&C's to this service, they then get an API key.

A short video clip from some of the "Demo Version"


Rachel Lammey - CrossRef Text & Data Mining Services - DEMO from Graham Steel on Vimeo.

 

Rachel concluded her talk by talking about the benefits of TDM'ing. "Over 14 million articles with full-text links add license information deposited"

Importantly, it is also worth noting that this API is Open Source and free to play around with on GitHub, as are the likes of ContentMine.




 



Breakout session: Extending access to e-journals for NHS partners



Alan Fricker
Liaison Manager, Kings College London

Kings College London (KCL) is part of an Academic Health Sciences Centre - big institutions aimed at bringing together research, education and health care – and has been the library provider for a number of hospitals and NHS trusts in London.

A quick timeline

In the past, NHS staff had excellent access to paper journals.  Libraries were dotted around, so all they had to do was walk down and grab.

In the 90s, the switch to e-journals started, and they had excellent on-site access to journals across various formats.  All they had to do was use the sneakernet (Wikipedia), walk down with a floppy disc and grab.

Coming nearer to the present day, KCL, as a member of the UK Research Reserve,were able to get a good level of electronic access to content and were able to dispense with their paper content.

But to get access to these e journals NHS staff still had to go down to a library and log in – counter to the change in user expectations.  This led to frustration!

About NHS procurement

A NHS library is very different to one in the HE environment.  A big NHS library typically has 5 members of staff.

In KCL’s case, there have been lots of local efforts made to provide access for patrons, including:

  •  KA24 (Knowledge Access 24) service for health and social care staff in London and the south-east of England in the 2000s
  • National Care Content, an England-wide procurement program

And the London Health Libraries met their goal of being e-only by January 2014.

NHS working together with HE for procurement

There have been many attempts at getting these two to work together.  

Users First  (2003)  report  by  John  Thornhill  identified ways this could happen: 

  • develop joint HE/NHS licensing
  • longer contracts – monitored
  •  joint working at all levels
  •  explore common authentication
  • local and national initiatives

In 2008, Imperial, KCL, QMUL, UCL and St  Georges formed the London  Medical  Schools Procurement Group, focussing on purchasing for the medical schools and extending access to the affiliated NHS trusts. This was great for the affiliated trusts, but bad if your NHS trust was not affiliated.

2011-2012 saw the start of the AHSC pilot.  

Cambridge, UCL, Imperial, KCL, and Manchester University worked with a number of publishers including Elsevier, Springer, Nature and Thomson Reuters, to extend the licences to the NHS trusts.

A number of issues were highlighted, including:
  • Problems with varying licences
  • Low NHS usage relative to the HEI (0.5%)
  • The proposed business model was for no charge for  NHS  trusts  unless  the usage  in  current  year  exceeded 10% 

The NHS staff experience?

There are lots of different access options for the NHS at KCL:
  • You can walk into the library to register for an NHS account
  • Self-registration for NHS OpenAthens, where you can access national, regional and local content
  • Use university affiliate status, and log on and access your university’s research via Shibboleth
But this is tricky to explain and navigate and so it is basically an access nightmare for NHS staff.

And they weren’t very happy:
  • The GMC survey highlighted the problem
  • User surveys echoed dissatisfaction
  • Customer service staff received negative feedback regularly
A survey of doctors identified what they most wanted, which showed that there was a problem of people not realising what they already had access to.  People had a different impression of what they actually had. 

More outreach was necessary so they decided to try and extend the NESLi deals.

What KCL did

The Finch Report recommended licence extensions for the NHS, so this was the context for this approach.

The rationale was:
  • The same people will be accessing the content as before – the current setup was an oversight, as the JISC deals doesn’t allow access for users who aren’t institutional, but KCL is their library support
  • They were not undercutting any NHS subs
  • Mobile access was very important.  The same content needed to be delivered, but in a way that is easily accessible to offsite staff.

There has been a national 1 year pilot that started in April 14, and this had led to agreements with six publishers (pilots and purchases) and 7,000 additional titles through OpenAthens.

Results so far

It is not all plain sailing, as there are a number of issues with link resolvers, authentication and challenges in promotion, but overall there are more NHS staff logging in via Open Athens each month and the GMC survey shows an increase in positive perception of access to e-journals.

Q: Why is NHS usage level low?

The NHS usage is not the same kind of intensity as in HE – there is a fall in the time staff have to read content, but not an increase in number of papers read.
Much of the content available is research, not “hands on” clinical content, so not all is relevant to them
The usage is low but very broad, e.g. one journal used 50 times, which would be very expensive if subscribed to in the traditional way.

Q: Barriers to access?

In Norway it has been proven that IP based access is best, but there are difficulties in NHS as they NHSN3, so you can’t tell the difference between IPs.

Catherine Allen – Touchpress – “Innovation in non-fiction content”

Excellent and informative presentation from Catherine, showing the possibilities of taking Ebooks and publishing to the next level.

Brilliant interactive learning tools with 3D images of chemical elements. 

The Wasteland – T.S. Eliot brought to life with Fiona Shaw reading the poem, interspersed with T S Eliot reading his poem, the reader and the text of the poem together and original manuscript with T S Eliots annotations.

Apps started in the 1990’s think back to the fabulous Encarta.

Apps as a whole generate more revenue than Hollywood and people spend more tiime with their apps than watching TV!

The future of apps = being integrated into everyday items such as fridges, Google glasses.

What do apps mean for scholarly communication?  We are at a point where there is so much scope.

Fantastic presentation showing the use of apps in the future.  The opportunities for science and medicine are incredible.  I am looking forward to it.

By Jo Milton, Medical Library Cambridge University Library, School of Clinical Medicine

Monday 30 March 2015

Breakout Session Group A: "Through a behavioural lens darkly: how ethnography can illuminate research into users"

(NOTE Mon 30/3, 3pm: this session will be presented again at 11am Tuesday 31st March - do attend if you can, it was fascinating! cheers, Paula Cuccurullo, EDINA)


Through a behavioural lens darkly: how ethnography can illuminate research into users
Presented by Bryony Ramsden (University of Huddersfield - Twitter @librarygirlknit) and Gareth Johnson (Nottingham Trent University - Twitter @llordllama)

Bryony's presentation: http://www.slideshare.net/UKSG/uksg-conference-2015-through-a-behavioural-lens-darkly-how-ethnography-can-illuminate-research-into-users-bryony-ramsden-university-of-huddersfield-and-gareth-j-johnson-nottingham-trent-university

Bryony Ramsden started by telling the attendees that both presenters will be talking about how they might use ethnographic research for their own libraries.  She told a ‘once upon a time’ story about the library being renovated at Huddersfield; she was appointed as a research assistant to see how the library changes had impacted their users.  She then got in touch with ethnographer Amy Whiteside (University of Minnesota) who had done similar work and helped Bryony learn what she could do with her data.  The research project was small-scale (with a short qualitative amount of data), studying their library, use, seat counts etc.  The library spans 5 floors, with 4 floors holding study areas - some more popular, and some soft furnishings not used.  But why?

Survey data was collected at the door of the library and in various sections - SO MANY NUMBERS - the people who were collecting data complained about this, but Bryony points out she had to type it in ;)  Some students were identified, and asked to fill in diaries but not many really did it (even with a gift given) - students weren’t engaged like US students often are.  There was also an opportunistic survey of students in the spaces they were using, asking why they were where they were.  It was noted that people don’t equate what they’re actually doing with what their stated purpose of use is - 'I'm studying' often meant 'I'm playing with my phone sending email'.

After her interest was piqued, Bryony is 4 years into her part-time PhD continuing this type of research into space issues - she is taking a qualitative approach, and getting away from the numbers.  She is looking at user behaviours in several different libraries across institutions, considering how students interact with each other, and with staff.

So, what is ethnography?  It's learning about cultures, but keeping in mind you’re not the same as the people you’re studying, plus you can never truly understand if you’re not part of the group.  You can use consideration of differences to inform your observations.  Also, you don’t want to change those you study.  She quoted Margaret Mead: ‘Never doubt that a small group of thoughtful, committed citizens can change the world'.

What goes on with cultures in environments we create for them?  Have we created the right environment?  How can we improve what we create?  No matter how much you use the library, you aren’t connected in the same way via a work purpose as via a study purpose.  You can also take a critical look at the relationship between environment and culture (considering approach to gender, race, disability etc).

Conducting observations is really important - people in the library environment are aware of what they’re doing but maybe not consciously, or may not feel that you want to know the full story, so that’s where observation comes in to see what they don't say.  Interviews aren’t the same in ethnography as you imagine when you hear the word 'interview'; they want to go deeper, and they’re semi-structured with few questions to dive deeper.  It’s really important to give voices to the people being interviewed, to bring their stories out.  When doing space research (how people use spaces - still small scale in libraries) as Bryony is, observation and interviewing are especially important.  Libraries who have done space research on a formal level often make large changes to layout, admin etc - they are usually looking at improving user experiences (though perhaps not always).  It's not just space issues though - for those looking at open access and the ref, her research and methodology can show what works for academics (and what is difficult) in helping get their research out.

There are always difficulties in ethnographic research; it’s always going to involve much time and staffing resource to carry out the methods, and in an ideal world you would hire an anthropologist to do this work.  This is also very context-specific based around the people you speak to and the observations you conduct - you can’t really generalise across the board.  At the end, it’s completely up to the people who look at your data what they do with it, so it can be frustrating when it’s out of your hands - what they do may not be what you wanted to happen.  That said, what seem like minor changes you make based on your research can add up to big changes and make major differences to the user experience in the long term.  If you can talk to people, give them the voice and share their voices, this makes people feel involved in the process and empowered

Gareth then picked up the narrative, looking at emerging open access academic paradigms for his PhD work.  The whole idea of dissemination is entwined with current academic practice, but rank and file are still reluctant to deal with ‘the open’.  Gareth is interested in the qualitative rather than quantitative, the perceived ‘mundane’ things which are actually shaping institutional environments.  For example, he started looking at and working with open access in 2006, when it was a niche idea; it’s now gone mainstream.  The UK has a rich history in the infrastructure that props up open access, as well as producing a great amount of the world class research output, so we should be leading in this field.

Ethnographers can find interesting narratives and trends, but they can’t say ‘things are so’ - it's only 'so' with the people you work with.  Ethnography is the framing around which he hangs everything else; he is ‘the research instrument’ through which the voices speak, but with context and further information added.  No one can be truly subjective in research, but ethnography admits that the researcher is coming at the subject from a certain direction.  He discussed the neoliberal context of open access in a marketed sector, and got into the theory and intellectual framework that frame his research, including that of Foucault and discourse (shaping behaviour).

In his research, there were qualitative semi-structured interviews which permitted genuine perceptions and serendipitous insights.  He has 220000 words of data to transcribe and analyse through Qualitative Comparative Analysis. [I will add a photo here of Gareth's slide of all emergent themes he found]  He focused on the theme of barriers - in this case, why aren’t academics adopting open access.  His narrative reflections include the following:
- Academic deficiencies in OA awareness/knowledge are perceived.
- Positive engagement is more typically found in the STEM sector.
- AHSS lag can be attributed to dissimilar dissemination practice.
In Gareth's next phase, he wants to do further research across groups and disciplines and develop a bigger picture.  Finally, he reflected on ethnographic methods: time demands are not trivial for gathering (or analysis), but there is great value in generating a rich narrative and dataset.

Audience questions:
- How many times did they have to change questions?
Gareth kept his questions the same (or kept them of a similar theme as he modified) but he will move on to new questions that came up at a later time (questions should first be tested on a small subset? he didn't have that much time though); Bryony had very specific questions - you do get led a bit by the data but again you can always come back to it later.
-How do you balance asking the questions with what you observed your clients doing in the library?
Bryony got her data to work with to prompt questions, so she goes into interviews on the basis of knowing these things - whether what people tell her goes along with that is another thing!
-Bryony, are the library still collecting all the seat count numbers at Huddersfield?
Not that she knows of.  She can’t say where she’s collecting now, as there are issues of institutional anonymity.
- Health studies are dominated by quantitative data - how do you feel about ’no stories without data’ / ‘no data without stories’?
Gareth has always been told that there must be qualitative data - it's interesting how they work together in different sectors.
- Gareth, are you going to propose ways to destroy the current networks of scholarly access? ;)
He was told that he should show the data and leave it to others to do the destruction! Though he’s seeing how he can work in advocacy - might have some of these things in his final work…
- What is Gareth’s timescale?
He’s in his 3rd year hoping to finish interviews over 3-4 months, writing up from September and available next year (via open access of course!)
- So Bryony, why don’t people like soft furnishing?
She suspected that they didn’t lend themselves to PC use but were often located near them, so it wasn’t condusive - if we’re designing environments with comfy seating, it should be more of a lounge environment and other comforts are more welcome than PCs.  She also thinks libraries are guilty of not playing around enough, changing little things and seeing what happens!
- What about anonymisation?
For both, some people really wanted to have their names included, but that option was there; Gareth was worried about what he does with data after it finishes (AHRC will want that but as he said it’s an issue in the academic culture!)

Sunday 29 March 2015

Hours to go until the 2015 conference starts.  Check back for posts regularly.

We'll tweet from @UKSG with #UKSG15 when a new post appears.