- Reduce redundant effort by using shared services and use savings to fund new services
- Make your data freely available and reusable - can add value
- Sharing content can help engage users
Owen Stephens liveblogged the day and has given a much more in depth and accurate write up of the days than I could manage. So rather than repeating what's already been written below are the main points I took from sessions I attended.
Opening keynote - Dan Greenstein (University of California)Offered a number of fairly radical suggestions for tackling shrinking budgets and growing demand for digital content and collections:
- In a digital world why are we managing print? Especially why are we managing general collections at the expense of special collections? Should focus on what makes a library unique.
- Suggested solutions include having central repositories for general print with Print on Demand and downloads to handheld devices and providing secure access to digital copies
- Digital collections should be supported from core resource budget e.g. Funding Open Access
- National shared services to reduce costs and savings can then be put into digital content/ next generation services e.g. consortial licences for resources which allow each institution to tailor collections to their needs; National instututional repository - more interesting to see by subject / department than grouped by institution
- New models for delivery - micorpayments e.g. deepdyve.com 99 cents to rent an article - can libraries build on this rather than letting commercial providers do it?
- Reduction in physical collections/ increase in shared services would lead to smaller physical space needed - could move to providing support in departments instead - new type of academic librarian delivering innovative services.
If you love your content set it free - Mike Ellis (Eduserv)
- Traditional value in content was gained from locking it away and charging to use it.
- Now copying/ piracy of contetn is almost inevitable so need to find value elsewhere
- Still valuable but has shifted e.g. Paulo Coelho found all his works were on P2P sites. So he set up own site Pirate Coehlo with links to the downloads. Sales of his books went up. Similarly Lady Gaga says she's not worried about file sharing as money is in touring rather than music sales.
- Find value in what can't be copied e.g. support rather than software
- If putting data on web then may as well make it reusable as with screenscraping/ yahoo pipes it can be used even if you don't
- Making availble can work for you - 75% of twitter traffic comes through their API - growth is due to external developers.
National Archives & Flickr - Jo Pugh
- National Archives record 'business of government' using Flickr as a shop window
- Put images they already had on Flickr Commons
- User generated content (comments/ tags) used to enhance their catalogue
- No judgement on how users interact with content (e,g, comments vs geotagging) - all mean they're engaging with content
- TNA labs (hadn't been launched at conference) to allow to showcase developments. Could we do something similar?
Geospatial data - James Reid
- Rapid uptake informally and formally and driven by availability of technology on phones etc
- Various projects - Unlock uses georeferencing; Inspire - improve sharing of geospatial data
- Augmented reality predicted to become mainstream technology in next 5 years
Linked Data - Tom Heath (Talis)
- Linked data has generalised structure to enable connections between data.
- Need unique identifier for data e.g. URI
- Using linked data can reduce cost for creating services as don't have to start from scratch in connecting data sets each time you try something new
- Networks add value to the things that are connected (compared with railway linking 2 cities)
Digital New Zealand - Andy Neale (Digital New Zealand)
- National vision for archives - rather than asking people to sign up to whole vision started with collecting data for one project on Armistice Day in NZ (also included projects like search widget, remixing content, search API)
- Whilst collecting data for this project asked if contributors could send everything else at same time
- Low barriers to contribution - would accept any data and enhance themselves
- Used Scrum methodology in development - producing something every 2 weeks so easy to gather support as could see what the project was doing
- Next step was supporting digitisation process - Make it digital with rather lovely website
- Branding/ design really important - products look really good so can feel proud of them and inspire commitment
Open Science at Genome Scale - Liz Lyon (UKOLN)
- Genome sequencing producing vast amounts of data - issue of how to store - looking at cloud technologies
- Researchers can be reluctant to share data - need to think about how to incentivise
Galaxy Zoo - Chris Lintott
- Project to classify galaxies - was being done by one person before realising that was impossible. So built a website so that amateur astrolgers could classify galaxies - citizen science
- Picked up by media so huge interest and within 24 hours getting 70,000 classifications an hour
- Built community around project
- Surveyed users and top motivation for taking part was wanting to contribute to research
- Zooniverse now using this idea to apply to other projects
Getting your attention - David Kay (Sero Consulting)
- Attention data = data about what people are looking at / are interested in e.g. usage data
- How can use? Would national scale be useful? Could it be used to help collection management?
- Projects: TILE project; MESUR (e-journals)
- Survey as part of MOSAIC project - 90% of students want to know what other students reading.
- Paul Walk suggests that in library usage data undergrads will be in high usage data whilst postgrads will be in the longtail so automated recommendations would be useful for undergrads but creating networks around usage would be more useful for academics (although would they want to take part?)
- Model of bit torrent - shows how many downloaded and comments about quality - helps you decide which file to download. Also shows quality of contribution is important - if know who's reviewing can judge if its valuable.
Panel discussion
- How to fund - how to make sustainable? Savings elswhere e.g. redundant services , prioritising spend. Doing things along with business as usual (e.g Digital New Zealand Search API)
- Metadata - how to ensure good enough in e.g. repositories? Depends on application and what you want to do with it. Mark a record as not having been checked - ask users to report errors. More specialised metadata costs more - but need to weigh up whether will actually be useful. Putting metadata 'out there' gives motivation to go back and check it. TROVE project using crowd sourcing to improve records.
No comments:
Post a Comment