Report on VALA 2010 for ITSIG

Warren Curran

 

Sponsored by LIANZA ITSIG

 

As part of receiving the generous sponsorship to VALA 2010 from the LIANZA ITSIG I have to report back on what happened and what I learned.

 

Amanda Brown’s 2006 report has an excellent summary of VALA that I can’t improve on, so I won’t try. http://wiki.lianza.org.nz/index.php/ITSIG/VALA2006Report

 

The conference website is http://www.vala.org.au/conf2010.htm

 

Thoughts while flying back to New Zealand

• The conference may not have looked that exciting on paper, but it ended up being very interesting and a lot was learned and some things confirmed.
• The Cloud, Linked and Communication are the things.
• Doing things the way the patrons search / use instead of how we think they should
• Discovery layers in the cloud.
• Hosted library systems in the cloud (OCLC WorldCat Local, ExLibris ERM)
• There was something about counting and using use data.
• At ANZREG the head of AARLIN, which will be winding up at the end of the year, said what they found out in the 10 years of work was that Federated searching was a broken technology. Forget it and go straight to a discovery layer if you can afford it?

   

• One Search to rule them all and in the metadata bind them. In the land of Discovery, where the information objects lie

 

The conference began with this excellent video of a Digital Native http://www.youtube.com/VALATV#p/a/u/3/7_zzPBbXjWs

 

The first plenary was Karen Calhoun’s (OCLC) warning us that the web is causing disintermediation and the disruption of libraries. That 89% of people begin research with a search engine.

 

“We need to be responsive to change”. This has been the case for libraries for ages!

 

Of course coming from OCLC her solution is cooperation on a huge scale to:

• Embed collections in the web
• Enable discovery and delivery of a wider range of Information objects
• Understand and engage with local communities
• Raise a culture of continuous improvement

 

As well as sharing efforts in collection management and metadata.

 

University of Auckland’s Kate Edgar Information Commons was interesting in that is effectively a multi storied version of what we want to do at MIT.

 

Jane Burke (Serial Solutions) continued the theme of Disintermediation. Article and chapter are the information objects of choice for researchers/students. Libraries, their websites, catalogues and databases aren’t being used because too intimidating, confusing…

 

Discovery is the way to go. Starting with a compelling and simple interface which searches everything – catalogue, databases, fulltext, the library’s website. Federated searching is the bare minimum. Ultimately we need web scale discovery i.e. Hosted, preharvested, pre co-ordinated, with contributions directly from publishers and with collection and ingest capabilities - all at Google speeds.

 

Some companies / organisations are moving towards this with their Discovery Layer products (including - surprise - Serials Solutions). These hosted solutions locally index the metadata as many of an organisation’s resources (to solve the delay issues with Federated searching) and begin with a simple search interface. Limiting options and facets are made available at the results stage.

• KISS - don’t gum it up.
• Give up bibliographic instruction (it’s given at the wrong time and the wrong place).
• We have to sell to patrons at the point of need.

 

There was a slightly boring presentation on eBook usage, partly due to the man’s voice. Curtin has tried a number of different eBooks vendors with differing approaches to delivery and payment. I was particularly interested in the effect of EBL’s service, as we started using it near the end of last year. One buys individual titles in EBL rather than leasing access to all titles (eBrary’s model). The interesting bit is you can get your patrons to purchase the titles when they agree to the offer to loan it to them, after about 5 minutes reading.

 

What surprised me is the relatively few titles that were bought in one year at Curtin with open slather (a patron “loans” a book for the first time and it’s bought), only 180 or so titles.

 

His conclusion – needs more research!

 

I am not that interested in citation analysis and other bibliometrics, but in Australia, and probably in NZ with PBRF, researchers need to show that people are reading and using their research to get more funding. So researchers need to know how to show this and consequently so do Librarians who will probably have to teach the researchers. They are also helping researchers by checking the researcher’s lists and checking them twice. At least that’s what Curtin is doing.

 

Another research report was on trying to find out how useful eresources are to researchers. 3 ways of measuring this was described – 1) Implicit (download counts etc.), 2) Explicit value (interviews, bibliometrics) and 3) Derived (ROI’s etc.). This research was looking at the “Critical incident of last reading”. i.e. what was the last eArticle you read?, why?, was it useful? For researchers the older articles were valued more, were more likely used for research and came from the library. Probably the ejournal back files the libraries subscribe to.

 

Science type disciplines got the best rate of return from eresources (i.e. for every dollar spent on eresources / the library, the institute got so much in return). Humanities had the lowest ROI – almost 1 to 1. Not surprising really, humanities are more book oriented when it comes to research. The point of all this is to have evidence to disseminate on how the library contributes to the institute’s income. The next phase of the research is to try and come up with ROI measures for teaching and learning, which is more relevant to us at MIT.

 

The Semantic Web – metadataing everything on the web. Tom Tague from Thompson Reuters talked about the impact of them providing free access to their product Open Calais, which goes through the text of documents and generates lots of metadata. It breaks content up and runs it through 2,000+ rules and places terms in to categories -= Named Entities, Facts, Events etc.

 

You only have to agree to let them keep the metadata it produces, so they can harvest it for their own nefarious devices. Actually they seem quite benign. In fact they haven’t worked out what to do with all this data yet. A lot of people are using the service to create tag clouds. One submitter was processing Doctor Who scripts every 90 seconds, and crashing Open Calais, but they don’t know what it was crashing on.

 

An interesting example on how the metadata is being used was investigative reporters pooling their research and the Open Calais metadata and using the metadata to identify relationships e.g. if the Cities contracts are being awarded to relatives / associates of officials and elected representatives.

 

Makes me think of Batman with his super computer linked to everything and connecting disparate clues together to find out where the Joker will strike next.

 

Provenance is an issue for Thompson Reuters. Indentifying the “truth” of a document submitted to Calais. In the future they will be providing a scale of truthfulness to submitters of the documents. How Librarians may use it?

• Automatically tag content in a useful way.
• Detect connectors, provide context and deliver insights.
• Chunk content so it can be mashed up in new ways.
• “Near enough is good enough”.
• Creatively mine collections.
• Integrate relevant social media and real time web resources in to your collection.
• Use to build intuitive, web 3.0 interfaces that take the burden off librarians and patrons.

 

Tom confessed he had never heard of OCLC nor WorldCat until this conference! But he was going to have a chat with them.

 

Marshall Breeding gave us an overview of the ILS market, what direction they seem to be moving in and what they aren’t doing they should be doing.

 

Technologies are in transition: xml / web services, beyond Web 2.0, Cloud platforms delivering to the full spectrum of devices (device neutral).

 

ILS are either evolutionary or revolutionary in their development and then there’s Open Source. It looks like those who wanted to go Open Source have already done so. Not many libraries interested now.

 

Vendors are slowly opening their systems up, via API’s, but not as far as Marshall thinks they should. “Back room” processes should be included.

 

There is movement towards Discovery Interfaces and web-scale discovery services i.e. harvest the metadata from ALL your resources into on big index and search rather than send our queries to them (Federated searching). Also provided as a service so can reduce duplication. Webscale I think means including ones website and other websites in the index. Library websites are still silos of information. We still require patrons to serially search resources. There should be only one search box which searches everything, content and services, including what’s in the library website. But that’s not simple to do. It has to be precise, predictable (search results) and nuanced.

 

Movement towards this:

• Discovery layer (ExLibris, Serials Solutions, EBSCO etc.)
• Social Discovery (tagging, user supplied ratings, BX recommender service from ExLibris)
• Deep indexing (not just metadata)

 

Federated searching was an interim solution (AARLIN says it’s a broken technology). With Discovery Interfaces one still federate searches articles. Web scale search is locally indexing as much as possible. Now it’s being put in the Cloud. Providing the platform and storage as a service. The next front is providing all of this to mobile devices / being device neutral.

 

Money is always an issue (library systems aren’t a real money spinner), but it does force innovation.

 

Next 5 years?

• Most libraries are using evolved systems.
• Increasing ranks of next-gen LMS.
• Library resource discovery metadata (indexing everything)
• Mobile
• Transition from local to Cloud computing (50% confidence)

 

Roy Tennant (OCLC) on Libraries at the network level – API’s, Linked data and cloud computing.

 

Structured data is the next revolution, via API’s and Linked Data. Cloud computing is the new thing (as a service).

• Service over the Internet
• Low(er) barriers to entry – just need browser and credit card
• Pay as you go
• No need for local server capacity
• Save staff

But

• Do not have complete control
• Reliance on network connectivity and speed (issue for NZ)

 

I am personally interested in CMS as a solution for developing our website, So I attended a concurrent session on the installation of one and on Thursday 2 boot camps (practical sessions) looking at 2 open source products, Joomla and Drupal. They still didn’t answer my questions. Mainly that I was hoping for information on setting up a database driven website. Note to self – don’t get developers of something to teach it, get implementers of it to teach.

 

Thursday morning’s plenary had to be re-jigged as the booked speaker cancelled at the last minute, due to the snow in the USA. With power outages etc. he couldn’t teleconference either. A panel was organised, but I didn’t attend as I was still trying to get Joomla working.

 

Later Stephanie Orlic from the Museum Lab, Musée du Louvre, tried to show us some interactive installations they paid a Japanese company to produce, but her strong accent, fast talking and the technology breaking down made it hard to understand.

 

The interactive installations were really neat, but the museum had the money to pay for someone else to produce them and if my experience at the Pompey exhibition at Te Papa was anything got go buy, ¼ of them won’t be working and there would be crowds waiting to use them.

 

Quite a few people were Twittering and posting to blogs during sessions (I made a few posts to Paul Sutherland’s Kiwis at VALA blog) – this Back Channel is becoming an integral part of conferences these days.

 

The Final Plenary was Mckenzie Wark who showed us how he wrote a book on Gamer theory with audience feedback. He wrote a draft which he made available online and provided an online feedback mechanism, down to the paragraph level.

 

A paragraph is displayed on the left and people’s comments are on the right. Feedback was used to improve the book which was then published physically. The online version was free, but people still paid good money for the print copy (pre-sales were greater than usual for this type of book). The second edition included précis of the comments made online.

 

Asides:

• “If you give stuff away you can still sell stuff, but lawyers don’t like it and seek those forwarding the stuff”.
• In another book he found royalties for images (stills from movies), are just way too expensive, so he got someone to create his own versions of the images in graphic form (scenes from movies in this case).

The conference ended with a short round up and thank yous and then drinks and nibbles.

 

General comments There was lots of walking. My dogs were sore and tired after the second day (although it could have been the new corrective insoles I was wearing). Distance wise it was probably about the same to get there as the previous venue, but psychologically walking past all the exhibition halls (about 3 rugby field widths) to get to the convention centre made it seem longer. The food was excellent, which is important, but you had to get there quick before the good stuff was all gone. More places for people with laptops and even though we had to pay $30 for Wifi, it was very useful (I had my bosses laptop to use because of the boot camps).

 

On paper the conference looked a little boring, but was in fact very useful, enlightening and educational. It also confirmed some of my opinions, changed a few and showed the boss that we should change our approach with searching. I thank ITSIG for awarding me the sponsorship and encourage everyone to apply the next time it is offered.

 

Warren Curran - 25 February 2010