37th Annual MCN Conference, “Museum Information, Museum Efficiency: Doing More with Less!”, Portland, OR Nov. 11 – 14, 2009.
This conference by MCN came to my attention through a posting on the LITA list serve. The museum perspective on data and information management was interesting and I found the conference useful and plan to follow up on several presentations. Participants were from a full spectrum of museums, from art to anthropology, from the Smithsonian to the Shelburne Museum in Vermont.
Below are some notes on the workshops and presentations I attended. Some of the presentations are online and linked to from the program schedules for each day:Some are also on SlideShare tagged with “mcn2009″ or in the MCN 2009 event section of SlideShare.
Cloud Computing | Keynote | Case Studies I | Museum Data Exchange | More with Less Roundtable | Case Studies II | Strategery | Case Studies III | Semantic Web II
Cloud Computing Workshop (presentation)
by Moad, Stein, and Davidow.
The workshop started out with a helpful introduction to terminology, concerns, and implementations of cloud computing. Gartner’s Hype Cycle for 2009 showed the timeline for cloud computing expectations vs. time to mainstream adoption. Gartner says it is the #1 strategic technology area for 2010. They differentiated between cloud applications (Zoho, Gmail, Google Docs) and utility computing (“elastic IT-related capabilities, renting CPU and/or file space, ie., Rackspace, Windows Azure, Amazon Web Services). In 2008 21% of companies were piloting software as a service. Gartner says cloud is the top trend for 2010.
Top concerns of those looking into cloud as an option: total cost, security, not able to find the applications they need.
- Pros: fast deployment; lower cost/no capital expense; reduced IT maintenance; elastic and unlimited scalability; reliability in service may not be 99.999 %, but as good or better than their own services.
- Cons: information security; long term offline storage (where is it?); potential vendor lock-in (ie., Google apps); some bandwidth bottlenecks at your connection end; latency; lack of control of downtime
Presenters provided a quick summary of the items to consider in making choices about cloud computing: what kind of security requirements fit your data; how granular is your information; where are the performance bottlenecks; and what is your IT staff like.
Jungle Disk was mentioned as one example of a way to backup office files to Rackspace or Amazon S3. Upload of large files can be relatively slow, depending on your connection to the larger pipes, but you can do scheduled backups. It retrieves files gracefully with a drag and drop interface.
Amazon Web services (AWS) has simple storage (S3), “Elastic Compute Cloud” (EC2). Transfers between S3 and EC2 are relatively cheap; the transfer to and from AWS can add a lot to the cost of using AWS. With S3 you pay for what you use; with EC2, you pay for a block, regardless of how much or little of it you use.
As an alternative to the cloud and AWS, the Indianapolis Museum of Art plans to continue to use cloud for web apps, but not for storage. They have 16 TB stored onsite, 16TB offsite. They benchmarked growth rate (~14 TB in 4 years) and looked at storage area network (SAN) vs. S3 and concluded that co-location was more cost effective than S3 by approx. $70 K, but there are other costs to colocation, such as server maintenance and admin.
There are a couple of Firefox addons that make it easier to work with AWS: Elastic Fox and S3fox, and the EC2 console in AWS.
Moving Drupal to the Cloud: A step-by-step guide and reference document for hosting a Drupal web site on Amazon Web Services: http://www.imamuseum.org/cc-mcn09.pdf
Fedora for a Digital Asset Management
This was an exploration of using Fedora on AWS for preservation work. Spent ~$35 K with a vendor to do Ruby on Rails for Fedora, on AWS. This saved them a lot of money vs. a hosted ISP (went from $1200/mo down to $900/mo now). They use subversion for version control of code, and the vendor handed off everything once the instance was set up. They required documentation on a wiki, and an “Amazon Machine Instance” (AMI) that they could then check out of Subversion and run on AWS. The reviewed some technical “gotchas” that are worth looking at in the slides if you want to pursue something like this.
Presenters also talked about the ArtBabble site (neat implementation of AWS for video hosting), and Rightscale.com.
Keynote address by Karen Donoghue.
Most of Donoghue’s talk focused on examples of interface design she has worked on, such as for handheld devices, incorporating a browser into a cell phone. She also talked about content distribution platforms, such as RSS but with more interaction capabilities. She likes to use translucency (content layers display partially through those above them), via CSS. Affordances are important in design, where the representation conveys the action (ie., button graphic -> push/click). Other concepts she covered: presence [indicates whether or not a person is connected]; personalized didactics [museum visitor uses a hand-held device to read a barcode at an exhibit item which translates the text into their language].
Case Studies Session I
Using Open Source Software in an Era of Tight Budgets
Robert Schimelpfenig, Maria Schenk. WSU used WAMP and PHP MyAdmin to work with metadata for ContentDM that are compatible with Dublin Core. Case study summary:
Schimelpfenig and Schenk, Using Open Source Software in an Era of Tight Budgets: A Case Study of WSU Vancouver Library’s Digital Archives Project (PDF)
Crazy Quilts to Patchwork Technology
Narda McKeen LaClair. One of the neat ideas that Narda talked about was the use of Flickr to highlight donor-funded projects throughout the life of the project. Example: adopt an animal for preservation. This information is also linked to from their Facebook page.
Using Server and Storage Virtualization to Build an Economical and Scalable Infrastructure
Sarah Winmill. Sarah reviewed their project to set up 60TB of storage, using VMware. They followed business continuity planning from the start to avoid service breaks. For secondary storage off site, they located servers at another organization that is also using Hitachi, and provided space for that group’s off-site storage at their facilty. This enabled both organizations to reduce costs of their secondary storage while maintaining a good baseline of support. A case study is on the vendor’s web site. Here’s another summary of the project.
The Flickr Commons Experience: The View from the Oregon State University Archives
Tiah Edmunson-Morton. OSU is the only university archives that are currently on Flickr Commons. They joined in Feb, 2009. Commons is more formal than the normal Flickr account, but only costs $25. Photos are scanned into ContentDM, then added to Flickr Commons with batch uploads of the images. Unfortunately, it requires manual upload of the metatdata into Flickr, but they hope to move to auto-updating. Updates of ~50 images twice/month keep the manual aspects manageable. They also use the indicommons blog for Flickr.
Freedom to Experiment! The Luce Foundation Center as Testing Grounds for Innovation
Georgina Goodlander. The take-home lesson from the Luce Foundation presentation was to encourage and try out innovative low-investment ideas, (without a lot of marketing support), and then move forward with the results.
Fill the gap: The Center has a storage area of approximately 3000 objects that are also on display. If an object will be on loan (or out for repair, etc.) for more than 12 months, a replacement is chosen, and online viewers can submit their suggestions for what should be moved into the open spot via their Flickr page
Ghosts of a Chance was an interactive text-based scavenger hunt where the participant(s) use a cell phone with texting. It became a 2009 Webby Awards Honoree. It enabled interaction with the museum’s resources, and could be configured for individuals or groups of more than 10 people. It can still be accessed as a downloadable module.
The Field Museum Gone Google – case study summary
Drew Ruginis. Drew summarized their move from Microsoft Outlook Exchange to Google Apps for email and calendaring. Some of the issues mentioned were incomplete integration across apps (ie. Reader, Picasa); lack of password synchronization with Active Directory, and lack of granularity in the control panel. Positives were cost savings, and fewer outages.
Museum Data Exchange: Making hay with harvestable data http://www.slideshare.net/RLGPrograms/museum-data-exchange
Goodwin, Museum Data Exchange: Introduction (PDF)
Rubinstein, Museum Data Exchange: COBOAT (PDF)
Oberoi, Museum Data Exchange: Learning How to Share (PDF)
Waibel, Museum Data Exchange: Making Hay with Harvestable Data (PDF)
Full report on this Mellon-funded project was released February, 2010.
This presentation summarized a project by nine museums to create tools using COBOAT (http://www.oclc.org/research/activities/coboat/default.htm) to extract CDWA Lite XML records out of Collections Management Systems, publish records via Open Archives Initiative Protocol for Metadata (OAI-PMH), then analyze data, and examine issues such as standards compliance and interoperability. Implementation was generally easy, the politics were more difficult. Collaborative development required the developers to host and use the same technology, which they were not familiar with. The data aggregation collected more than 850,000 recoreds, and is a source for more research opportunities to look at consistency of implementation of standards, etc. across museums. Other tools and resources mentioned that are worth a look: Omeka, The Cultural Objects Name Authority (CONA). Also see the blog, hangingtogether.org
Doing More With Less: A community-software-based technology roundtable.
CollectionSpace. With sustainable funding, CollectionSpace could be an opensource alternative to commercial products for museums, possibly with natural history collections too. Maturity date is 2010. There are opportunities for partnering. for webinars, the community design workshop results (http://www.collectionspace.org/sites/default/files/Community%20Design%20Workshop%20Report.pdf), how to join the project, etc.
ConservationSpace is for in-house art conservation. Standalone, as a module in CollectionSpace, or interoperable with other collection management systems. Expected 2011.
Museum Scholarly Infrastructure. In development. To improve ability of scholars to collaborate with each other in study of collections. Shared technology infrastructure: Sakai (collab support and social media); ConservationSpace; CollectionSpace. Extensible infrastructure to cut costs. British Museum as tech lead, Partners are National Gallery (Raphael), Courtauld, RKD & Martitshuis, etc. Maybe by summer of 2011.
Should partner well with CollectionSpace. Common authoring environment for museum content; banners, kiosks, wall tags; usability best practices; accessibility compliant to manage risks for museums. Small device support (smart phones). Many partners, advisers, contributors. Version 0.1 is available now. Being built in a modular fashion. Pull and push info (web presence, plus user comments, tagging). Hope to release 0.3 in January 2010, more user interface improvements. 0.5 by April. Focused on user experience and usability.
Long tail digitization project, for collections too small or too fragile or rare to leave the institution for digitization. Initial target is paper, but in principle anything 2D. One-button free
publication to Internet Archive. Optimized for scholarship, books, boxes, collections. Multilingual OCR. About 300-pages/hour. Target: anyone, 1-page instructions, error checking, etc. target cost is $1100 (except for laptop). Self-correcting (de-warping, de-skewing).
Case Studies II
Pathfinder, a New GUI for the Art Institute of Chicago.
They use 7 wall-mounted information kiosks to provide interaction and way-finding. Kiosks were made with Slate Roof Studio and AvantLogic, using a Troll Touch Overlay on 30-inch Apple Cinema displays. They incorporated “ligth boxes”, popup images, and “show path” and “show accessible path”. Nice!
Art Babble: Play Art Loud!
Indianapolis Museum of ARt. This site is a niche site intended to offer something better than the typical YouTube content and an option for institutions to collaborate in providing video about art, which can be hard to find elsewhere. They provide full text transcriptions of the content, and try to make instructional videos exciting and fun. The amount of time a user stays on the site reflects the typical length of most videos.
MOA-CAT: UBC Museum of Anthropology Collection Access Terminals.
They replaced data books in the visible storage galleries with touch-screen interactive information systems, and have all 37,000 objects in the system. Launch date was January 2010. Visitors can view where an object is from via links to Google Earth.
Presenting Musical Compositions as Works of Art.
Wellesley College. The woodworking for the construction of an overhead speaker system for the exhibit was what was most interesting to me. Unfortunately their web site has no information or pictures that I can find.
Strategery: The Realities of Strategic Planning
Edson, Iannacone, Honeysett. I’ll provide more in-depth notes on this session, though the slides for one part of this session are also online. This was one of the more interesting and entertaining presentations–and extremely applicable to museums and libraries. The focus was on digital strategy and “strategy-creation processes, successes and failures, and the relationship between technology platforms and organizational readiness.”
For the Getty (Nik Honeysett, no slides online).
They did a strategic review to redefine academic computing as a networked environment.
- Success: define it, give outcomes that indicate it.
- There’s no place like home (their web site): use platforms to bring people to the website.
- Create more than you see (iceberg): a solid infrastructure, process, and workflows are needed to implement mobile technology
- Courage: have the courage to start again when a project fails or falls apart.
- Creativity: to make the complicated simple
- Give people permission to fail
- Improve with use–the essence of Web 2.0 is that it works better as more people use it (ie., social media)
- Do more with less: do less with less, but do the stuff that matters more. Segal’s law: “A man with a watch knows what time it is. A man with two watches is never sure.”
They used a set of internal guidelines to inform the reorganization, developing and using a process and infrastructure that is flexible, scalable, and sustainable. A common thread is that people know who’s in charge, who to go to for help, and the process being employed.
For the Smithsonian (Edson, Iannacone): some slides here: Smithsonian Web and New Media Strategy — Drivers, Process, and Execution
The assumption is that strategy has four components, in this order:
- Pain, Fear, or Opportunity
- Some kind of process
- Some kind of assertion
- Some kind of work
The Smithsonian used an internal blog and wiki as part of the process, to:
- Build a shared vocabulary
- Keep focus on mobile, UX, other subtle things
- Celebrate internal experts
- Practice skills to be used later on (“you get what you practice”)
They wanted a process that was public, transparent, and fast. Advantages (copied from a slide):
- Faster than traditional committee-driven process (Time is the enemy)
- Increase size of brain trust (Joy’s Law)
- Improve the odds for change
- Improve odds for execution (public promises not easily forgotten)
- Outside champions more likely to support “commons” goals than status-quo insiders
- Walking the Talk vis-à-vis crowdsourcing and innovation model
- “You get what you practice”
In the Smithsonian process, the process was the wiki; meeting agendas, notes, etc were all on the wiki. “The main intent of the workshops is to move relevant information to the wiki where it can be openly evaluated, sifted, weighed, and considered by all.” As participant comments in meetings were added to the wikii. This increased transparency, accountability, and speed. Action items and themes were highlighted after the workshops, to promote further synthesis and thought. Presenters also addressed the risks of using such a process. See the SI Web and New Media Strategy wiki.
The Change Model (slide 74)
(Borrowed from software and social entrepreneurship)
- Think big, start small, move fast
- Focus on doing things that matters (via Tim O’Reilly)
- Cultural institutions exist to do work in the culture
- Drive change through building A Sense of Urgency (John. P. Kotter)
A few points from Carmen Iannacone’s part of this session (slides not online):
What are hallmarks of disruptive success? innovation, flexibility, creativity, productivity, organizational discipline (better, smoother function), morale, achievement. The “tools” we have are: human capital, technology, goodwill, and luck.
He believes in aligning the infrastructure and operations staff with a maturity model (Gartner’s infrastructure and operations (I&O) maturity model: survival -> awareness -> -> business partnerships, see this example for more details; embracing a process framework, and MBWA – management by walking around. See also Good Projects Gone Bad for more on process maturity.
Other tips: follow the money in technology; assess your capital planning and investment control (CPIC) process; use zero-based budgeting; use earn value management (EVM) techniques if you can; use operational forensics to address and avoid issues with IT roll-outs and changes, and prioritize IT project list. Here are some of the things on their priority list:
- cloud computing, server virtualization, storage abstraction, mobile computing, ubiquitous wireless access, immersive user experiences, and sourcing & licensing.
Case Studies III
Digital Media Served up Using XML and the Google Maps API
Evelyn Lindberg, Laura Robinson. Based loosely on the concepts behind the Map of Knowledge project (http://www.nytimes.com/2009/03/16/science/16visuals.html?_r=1) they employed the Google Maps API to do a mashup with newspaper coverage.
Print quality downloads
by Sarah Winmill. They put the catalog online to facilitate download of print-quality images from a collection of ~150,000 images, using a proxy to download the files from a digital asset management (DAM) system. Images are cropped, which has been contentious. They’re now using a crowdsourcing approach to get feedback on the cropping. The total cost of this project was equivalent to a mini-exhibit; 6 month timeline, with 3 months of it in heavy codework.
Introducing imaging quality control practices into sustainable workflows.
The vendor, Image Science Associates http://www.imagescienceassociates.com/ , seems to have a good combination of hardware and software for providing a sustainable approach to ensuring that images are consistently high quality, whether from scans or photos.
Enhancing the usability of in-gallery media through data visualization
Paco Link. The topic was touch screen displays, something we don’t often deal with in the library but which I think could greatly improve way-finding and reference services for our users. They used cluster/contour maps of screen touches, generated from logs, to visualize how patrons were using the touch screen displays. The prerequisite is to include the logging capacity from the start in the design of the screen systems.
Hi Definition History
Tamara Georgick, Gabe Kean. Another presentation on kiosks, that looked at the development process for including from 1 to 3 panels, and employed a content management system for a lower-cost approach.
Semantic Web II
Koven Smith, Don Undeen. This project used Semantic mediawiki for quickly creating a system in which conservators could record their treatment notes, and add semantic relationships between terms. This was easier than the old system of written forms, and facilitated search, web output, etc. Since it’s built on the wiki framework and incorporates annotations, they didn’t need to spend a lot of time on setting up a platform. The focus is on terminology and nomenclature; annotations in the wiki create an rdf triple. “Queryable” data appears in a fact box at the bottom of the page, which can be clicked on to view inverse relationships. Other components of the system: CAMEO, a webservices piece written in C#, the CIDOC conceptual reference model, and the Halo graphical annotation extension.
Powered by ScribeFire.