Christina's LIS Rant
Thursday, July 20, 2006
Notes on Collaboration Expedition 52: Wikis
Collaborative Expedition Workshop #52, Open Collaboration: Networking Wiki Information Technology for Information Sharing and Knowledge Management
Held Tuesday, July 18, 2006 at NSF
I attended this intergovernmental workshop on wikis. This was a very large workshop with over 100 confirmed attendees. There were attendees all over the governmnet and from various vendors. Many of the government agencies already have significant experience with wikis.
Morning Keynote
The morning keynote was given by Dr. Calvin Andrus, Chief Technology Officer for the CIA Center for Mission Innovation. Dr Andrus has been a proponent of social software such as wikis and blogs to improve the end product at CIA. His 2005 article on complexity theory has been widely cited in the blogosphere (Andrus (2005) The Wiki and the Blog: Toward a Complex Adaptive Intelligence Community. Studies in Intelligence 49 (3)). His main point is:
The intent is to capture some of the tacit knowledge in the blogs and link it to the source data and reports which are stored in the repository. The wiki would be a place to take notes, float trial balloons, work on draft documents, and publish some complete documents. Since they started this effort about a year ago, they’ve created 1500 internal blogs of which several dozen are still active. The internal wiki has 12,000 pages after a year. It’s on JWIS – a top secret network. Intelink now also has blogs and Intellipedia wiki. Andrus feels that some of the work needs to be done at the SIPRNET level to get enough feedback to have real growth.
The major factor in getting this to work is trust. Managers have to trust analysts to publish before editing. An agency rule against collaborating with the larger intelligence community had to be rewritten to allow contributions to Intellipedia.
Panel Discussion
The next session was a panel discussion of wikis. Most interesting in this group were Clarence Johnson of the Department of Commerce and Olga Brazhnik of NIH. Johnson uses a wiki to share information with an international team of government technology ministries from Switzerland, Korea and elsewhere. Brazhnik is working on an NIH-wide wiki – there are already many wikis internally and externally at NIH, but she’s looking at how to do a larger effort while taking into account both the role of NIH as a federal agency responsible for health policy, and the NIH role as an innovator in bioinformatics and biomedicine in general. A DOD contractor spoke of layers of security for the wiki they built, and how new pages were tagged and disambiguated using preferred terms from the taxonomy.
Afternoon Keynote
Niall Sinclair spoke about “stealth km” – basically helping people find, keep, share, re-use information without taking it away from them and storing it in a central repository. Avoid the top-down mandate – km isn’t just another enterprise project to manage. (Much of what he said we’ve known for years.
Featured Presentations
(nb: Shankey and Rehberg went way deep into detail... I understood it at the time, but took no readable notes!)
MindTouch, Aaron Fulkerson
Held Tuesday, July 18, 2006 at NSF
I attended this intergovernmental workshop on wikis. This was a very large workshop with over 100 confirmed attendees. There were attendees all over the governmnet and from various vendors. Many of the government agencies already have significant experience with wikis.
Morning Keynote
The morning keynote was given by Dr. Calvin Andrus, Chief Technology Officer for the CIA Center for Mission Innovation. Dr Andrus has been a proponent of social software such as wikis and blogs to improve the end product at CIA. His 2005 article on complexity theory has been widely cited in the blogosphere (Andrus (2005) The Wiki and the Blog: Toward a Complex Adaptive Intelligence Community. Studies in Intelligence 49 (3)). His main point is:
“. . . that from intelligence officers who are allowed to share information and act upon it within a simple tradecraft regime will emerge an intelligence community that continuously and dynamically reinvents itself in response to the needs of the national security environment.” (emphasis his)In other words, intelligence officers need to collaborate to get the intelligence cycle short enough to fight modern enemies and move quickly in ways not predictable in the policy cycle. The enabling technologies are wikis and blogs linking to a repository of intelligence and findable through search and tagging. From his presentation, this “technology stack”:
The intent is to capture some of the tacit knowledge in the blogs and link it to the source data and reports which are stored in the repository. The wiki would be a place to take notes, float trial balloons, work on draft documents, and publish some complete documents. Since they started this effort about a year ago, they’ve created 1500 internal blogs of which several dozen are still active. The internal wiki has 12,000 pages after a year. It’s on JWIS – a top secret network. Intelink now also has blogs and Intellipedia wiki. Andrus feels that some of the work needs to be done at the SIPRNET level to get enough feedback to have real growth.
The major factor in getting this to work is trust. Managers have to trust analysts to publish before editing. An agency rule against collaborating with the larger intelligence community had to be rewritten to allow contributions to Intellipedia.
Panel Discussion
The next session was a panel discussion of wikis. Most interesting in this group were Clarence Johnson of the Department of Commerce and Olga Brazhnik of NIH. Johnson uses a wiki to share information with an international team of government technology ministries from Switzerland, Korea and elsewhere. Brazhnik is working on an NIH-wide wiki – there are already many wikis internally and externally at NIH, but she’s looking at how to do a larger effort while taking into account both the role of NIH as a federal agency responsible for health policy, and the NIH role as an innovator in bioinformatics and biomedicine in general. A DOD contractor spoke of layers of security for the wiki they built, and how new pages were tagged and disambiguated using preferred terms from the taxonomy.
Afternoon Keynote
Niall Sinclair spoke about “stealth km” – basically helping people find, keep, share, re-use information without taking it away from them and storing it in a central repository. Avoid the top-down mandate – km isn’t just another enterprise project to manage. (Much of what he said we’ve known for years.
Featured Presentations
(nb: Shankey and Rehberg went way deep into detail... I understood it at the time, but took no readable notes!)
MindTouch, Aaron Fulkerson
- Wiki software branched from Media Wiki circa 2004
- Runs on .net
- Will be open source (LPL and GPL), they will make money by selling a device (like Google appliance) (see opengarden.net, I think?, after next week’s release)
- Unlike MediaWiki, data is XML structured data, and user interface is a graphical text editor (not unique wiki language)
- Uploaded files are automatically versioned
- Pages can be hierarchically arranged
- The only “semantic” thing it does is microformats
- “discourse forms around or in the context of a topic”
- Basically looking for manual maps to an ontology to enable semantic searching
- ?
- Automatic extraction of semantic relationships
- Using reading styles to map the corpus to an ontology (but not one to one, can be multiple reading styles for a single ontology depending on need).
- ?
Comments:
Post a Comment
