Since we started using the Confluence enterprise wiki platform at my current job, I’ve come to a better appreciation for the use of fields in wikis, particularly as I’ve worked more with wiki templates.  With hundreds of projects deployed on multiple tiers, it’s very important to have the information about the servers, infrastructure tools, and applications organized in a standard way, so we can find and share information more easily (especially when getting a 3 am page and troubleshooting a system).   Templates and field structures can certainly help with that.  For example, one could have a drop-down list of approved operating system versions to pick from on a server information page template.

I still think fields should be used judiciously (if at all) when trying to encourage wiki use.  When just trying to add some data, it can be very frustating to encounter a field mechanism that seems poorly designed, such as incomplete dropdowns or incorrect-seeming unalterable items (though sometimes these may be due to fundamental disagreements about the intended use of pages/information).   Some people will not have the time or patience to try to contact wiki admininistrators or other users to work out a fix to the field or a common understanding of the page; they may just give up and not make the contribution they were planning. I was happy to notice recently that LibraryThing had fixed the dropdown that had so annoyed me before.  There was a long stretch of time before that where I just wasn’t as enthusiastic about them.

Where there are wikis with multiple users and audiences, sometimes compromises will have to be made, or one path or another chosen in structuring input mechanisms.  Sometimes this just doesn’t happen fast enough to keep up with changes going on in the world.  Where possible, I recommend having some kind of an open notes section in addition to fields, so that there’s at least a hope of keeping the contributions flowing and finding the (up-to-date) info one needs.  Also, anyone with ownership of the page (or at least some feeling of responsibility towards its contents) should probably set up a notification so they’ll know if the contents change, and perhaps be able to address problems/disagreements in good time.

• “The Wachovia presentation this morning appears to have opened some people’s eyes on how things could happen.”  Having been familiar with Wachovia since they were a small NC bank, I wondered what they were up to.
  • Steve Newberger wrote more about it:  [Fields'] analogy: Web 2.0 is this generation’s equivalent of his generation’s company picnic and bowling leagues. Additional critical rationale: mitigate the impact of the maturing, retiring workforce, i.e., the attrition of knowledge assets.  They’re using SharePoint (WSS/MOSS), with extensions for instant messaging (I miss using IM for work) and video blogging (we talked about getting that going at my previous employer — 2-minute service call how-to videos, viewable over mobile devices, anyone? — but didn’t get the resources to get it rolling in time).
  • More important than the technology is Wachovia’s attitude.  From a preview of the talk,  Wachovia’s collaborative environment is designed to attract younger Generation Y employees who expect access to Web 2.0 tools at work. [...] “They’re coming to us with high enthusiasm but encountering arcane tools and bureaucracies,” he said, adding that many young workers’ engagement levels “fall off the table” after about a year on the job. “They are leaving Fortune 100 companies,” he said.  As I said at my previous employer and during my Knowledge Gardening presentation at Penguicon 2007, the ability to use modern, fun, low-entry-barrier collaborative and knowledge sharing tools is a recruiting advantage.
• The CIA presenters recommended starting small, such as with an acronym wiki page (I’ve had success with that approach!), and pointed to an OSS (CIA precursor)  manual from 1944 about how to sabotage organizations and production efforts: Insist on doing everything through “channels”, refer all matters to committees, haggle over precise wordings of communications,  advocate “caution”, and question whether a decision lies within the jurisdiction of the group.  The conflict such behavior would have with the architecture of participation is obvious (Note:  I do think precise wording and formal processes can be useful and important, but relative risks should be considered, and getting bogged down is a risk, too).
• “People get ticked off at twittering and live-blogging during an Enterprise 2.0 conference, just like they do at other events.”   They get ticked off?  I hope no one’s gotten mad at me for tapping on my laptop at presentations.  I guess I can see the negative potential (folks so busy twittering, blogging, taking pictures, etc. that they’re not engaging with the rest of the conferees), but I’d always had more of an attitude of “their loss” when it seems that they’re not fully there.  I suppose if a substantial proportion were doing so, that might take away a bit from the energy of the event (?).

As with ”knowledge gardening”, a web search reveals that others have used the phrase “sustainable development” (in the context I mean in this entry) before me.   In fact, Kevin Tate wrote what looks like a pretty good book about about it; here’s an excerpt.

• Cool Visualizations:  my Spanish-speaking brother may be especially interested in this tag display/analysis
• Cool analogy:  energy required for state changes between data/information/knowledge, like ice/water/steam

And old content from me:  the audio from my Knowledge Gardening presentation at Penguicon 2007 is online at the Internet Archive (mp3 and Ogg Vorbis):  Sarah Elkins on Knowledge Gardening

Thanks to whoever put it up there, and the 6 folks who found it and downloaded it already!

Not that most of them are quick to identify them as knowledge management issues, but many of the recent organization-community conflicts I’m aware of have a lot to do with knowledge management, as affected by ownership.

• EVE online (game): How is sensitive information secured/accessed? How is information about controversies managed? Who has the ultimate say on what topics are allowed on the community boards?
• Wikipedia: What information is appropriate for Wikipedia? Is it a good venue for knowledge development (semantics: is organizing references knowledge development, or only when new analyses are allowed is that knowledge work (as opposed to information management))? What happens when the neutrality of editors/judges is called into question? Are links to sites critical of Wikipedia going to be allowed?
• LibraryThing: What information should be on a work page or an author page? What if there’s disagreement about the legitimacy/accuracy of including/excluding certain types of information (see previous LibraryThing entry)? Does “Common Knowledge” imply truth, or truthiness? Which should it imply? What if one workflow is better for adding books from the user point of view, but another is better at generating accurate book data?
• One Laptop Per Child: Is “go explore” a reasonable response to a new Give One Get One XO laptop owner’s question about the laptop? What if the US user base gets way ahead of OLPC in generating information about it, but, impatient with the official wiki, puts most of that information over on discussion forums on OLPC News (not associated with OLPC; in fact, sometimes highly critical, though often enthusiastic about the technology and mission) instead?
• For any of these, who owns the information users contribute, and how portable is it?

An obvious answer from the organization’s point of view is “we sponsor the site/project; someone has to decide, and we’re the ones who’ll make the final decision.” A vulnerability of this approach is when they depend on a large user community for contributions for their reason for being (Wikipedia) or for profit (LibraryThing). If their community grows disenchanted, they’re in trouble (especially if/when the Next Big Thing comes along). If the organization depends more directly on ads for survival (e.g., LiveJournal) than its users, it’s not such an immediate issue.

I think most flourishing communities have growing pains from time to time. Some resolve the conflict and move on, some paper it over and move on, and some splinter or go out with a whimper. I’m not sure what the sociology is of which direction a community is likely to go depending on whether it’s a mission-oriented community (e.g., an open source project) or a community of proximity (geographic, or common interests without a strong mission, e.g., people who like to chat about movies). I’d be interested in any studies about how the financial angle (e.g., non- versus for-profits) plays in, too.

Read more on the photo management software presentation:

Ted is presenting a relational DB management system he wrote which uploads photos in photo sets, groups in albums, allows derived photos, stores metadata, tags group photos arbitrarily , and allows fast searching.  Albums can  be exported to websites/media.  I don’t take that many pictures, partly because of  not being one to get around to doing much photo management.  I’m aware of Flickr etc. but do like the idea of photo management software I’d run on my own server.  Needs Apache, PHP, PostgreSQL etc.  But Ted’s Easy Foto Box package can be installed easily on a virtual (or real) machine.

http://www.tux.org/~tbr/easyfotobox/

Some of their search examples suggested a tagging solution to me. However, there’s no saying that random taggers will be interested in adding the sort of metadata that leads to commerce or facilitates the kind of use an organization may wish to make of its data. I can see why an organization might wish to encourage certain kinds of contributions through structured fields and other input mechanisms. This approach comes with its own risks, of course. Whatever they set up may become outdated quickly if their system/mechanism is not very flexible. What’s more, the more that inputting seems like “work”, the less enthused people will be about adding metadata. This may be just fine if what’s wanted is really just the experts/authorities adding metadata (not the hoi polloi), and there are enough of them, and they’re motivated, of course. I think mashups between social software and structured information resources have a lot of potential, depending on the implementation and the critical mass. Where things get tricky is when folks have different expectations about the interactions, as in my recent post about LibraryThing. Tim, the founder of LT, wrote a very thoughtful comment, and I’ve been meaning to reply to it at length; this ACM article has helped me sort out my thoughts a bit, preparatory to that (next entry, I hope).

* By the way, I was cheered to notice that someone at ACM appears to have seen/taken seriously the feedback about their beta site. Not just my feedback, I imagine, but anyway, I’m glad their post-beta implementation went in the direction I’d hoped.

Recently I noticed that Pieter Spronk, the author of my favorite PalmOS game, Space Trader, is the Director of Studies of Knowledge Technology at Universiteit Maastricht. He also teaches Logic and Games & AI and guides students in related studies. This intersection of interests makes a lot of sense to me; more work is being done these days in what games have to teach us about learning, and implementing these lessons depend in some degree to the capabilities of the knowledge technology infrastructure involved. Spronk has also written about various issues with knowledge-related technology (e.g., his short analysis of the trouble with USB sticks).

Space Trader is a single-player game (turn-based), with pirates, police, special missions, and, of course, trading. In this game, the only trading involves material goods (from water to robots). There is no trading of information or knowledge, though one can buy a newspaper or pass on warnings to others. On the other hand, I remember that CJ Cherryh’s Tripoint features Merchanter ships, and one of them makes a bundle by (spoiler below the cut) a particular information run (being the first ship to show up with a lot of new news and entertainment downloads at a remote location).

Cherryh has written directly and indirectly about the implications of technology (effects on society) in many of her novels. She made the point at a WorldCon years ago that as we have more and more information overload to deal with, more and more of us are going to rely on our friends to filter our information for us. This has benefits and drawbacks. I’m using this approach increasingly, and I’m seeing the pros (as discussed in the Penguicon blogging panel, I don’t have time to read all the news and blogs that I’d like, but I have a number of friends who blog the most interesting science news, for instance) and cons (e.g., risk of groupthink) myself.

Penguicon overall attendance: over 800
My presentation’s length: 1 hour
Q&A Length: 30 minutes (after my hour-long talk)
# of other events scheduled at the same time: 13?, including the wedding of one of the founders of Penguicon
# of attendees at start of talk: about 20, including professional librarians, knowledge managers, and various open source project participants.
# of people who left before Q&A: Only 1!
# of people who asked questions: about 10, including people who addressed each others’ questions in their questions (i.e., not just a Q&A but a group discussion, which I was happy to foster)

Feedback: One of the folks I gave a dry run to before the con said it was the shortest hour presentation he’d ever heard (yay!). Matt Arnold (head of programming for Penguicon 5.0) said “her presentation and ensuing discussion were stimulating.” Constructive criticism was that there were a couple of times where I was so deep into the material that I didn’t make it clear enough to the audience what the connection was between one of my gardening metaphors and knowledge work. I hope the Q&A helped address that, but that’s something I’d want to rework if I give the talk again anywhere.

I had said I was going to take the slides (which are not standalone; I didn’t want a wall of words behind me distracting the audience from what I was actually saying) and construct blog entries or a wiki section from them. However, one person who’s been published in Linux Journal suggested I turn it into an article for them (perhaps with more of a how-to focus, or more like reviews of different tools). I’m not sure if “we purchase first rights” would preclude conference presentations, blog entries, or what.

Penguicon prep: I was very glad that the time I spent pulling slides together, doing dry runs (thank you co-workers!), and adding last-minute Penguicon images paid off with a well-received presentation. I had also spent time emailing with my co-panelists (who was covering what), copying don’t-miss events into my Palm Pilot (with over 10 events going on simultaneously much of the time, it would have been easy to overlook items of interest), and acquiring and reading Nifty Guest John Scalzi’s The Android’s Dream (I’d already read stuff by most of the Guests of Honor). Fun read, with more to it than reviews had led me to expect.

Travel and transit: People reacted with concern when I mentioned flying with Northwest, but it worked out fine for me. I liked the Detroit Metro airport (DTW), especially its shiny red monorails. Driving to/around the con was fine except for construction slowdowns and rush hour Friday night.

Convention organization: Check-in was quick and painless. Con Ops was right on the main hall and appeared available and in control every time I passed by. In most cases programming had guessed how to match room size and audience size pretty well, though I was sad that I got to the RepRap talk (open source personal manufacturing device) too late to be able to really see in — that one was packed and flowing out the door. The program book was above average. It came with a map of the main floor, pointed to the Consuite (which provided sandwich materials and fresh veggies, not just chips) and reminded readers that eating real food and getting sleep would help maintain stable moods (another blogger noted what a drama-free con this was), and had a cool essay in the back on Knowledge Ecology by Programming Chair (Wrangler?) Matt Arnold that tied in neatly with my presentation.

Pictures and more entries forthcoming.

“Metadata” seems like a good word to describe my use of “game” in that context. I don’t like the thought of turning off audience members with careless buzzword usage, but I’m not sure how else to capture what I’m getting at. “Extended context data” doesn’t flow so well for me. Also, a number of the criticisms of the term don’t seem to recognize how metadata can be cheap and fun (tagging/feedback), or how emergent metadata (e.g., from the general public using unstructured tagging) can lead to interesting informatics possibilities.

I’m going to stick with metadata … for now.

I agree. I think more companies are waking up to the importance of helping their people deal with all the information overload (for their productivity) and capture/generate knowledge (IP/competitive advantage). Here are another couple of interesting articles I came across regarding companies, knowledge work, connections, and infrastructure:
1) Open Wide
2) Some Thoughts on “What is a Knowledge Worker?”

I’m not in their league, but I’ll also be giving a talk:

Knowledge Gardening, Saturday, 4-5:30 pm
How innovators can use open source social software to nurture knowledge and collaboration within an organization. Wikis, blogs, tagging and RSS feeds will be covered, along with reasons why you might choose to set up behind your organization’s firewall (at least initially). No special tech knowledge is needed to attend.

It looks like I’ll also co-panel Blogging Your Life Without Losing Your Job, Or Your Mind with Cathy Raymond (finalizing schedule). And, as I’ve done at a number of science fiction cons in the past, I’ll run (demo) some games. I’ve presented at a company conference (poster session), and I’ve sung solo in front of a crowd, but I’ve never done a presentation or panel like this before. I’m a little nervous, but also pretty jazzed.

However, it turns out that Jon Udell talked about “collaborative knowledge gardening” over two years ago. Which is great; I can point folks to his excellent column as an intro to the concept, and to him as another person excited about the possibilities of social software for collaboration and knowledge work.

Possibly further back (2000?), there’s the article Knowledge Gardening Through Music. I very much like the “Cognitive functions of gardening” section.

See also Knowledge Ecology.

Where I would like to be this afternoon: “Putting the Fun in Functional: Applying Game Mechanics to Social Software” at the O’Reilly Emerging Technology Conference. I may write an entry later about what we can draw on from games to engage people in learning.

When I started doing configuration management, and more specifically having to maintain software configurations (and builds and recreatability), I saw how vitally important it was to be able to know six months down the road just what I’d done to create a build.  This seemed to involve two major components:

• saving the information of what I’d done in the first place
• being able to find that information later

Since then, if we’d had yearbooks at different jobs, you’d probably have seen “Most Likely to Document” by my name and photo each year.  The “being able to find it” bit was somewhat ad hoc and somewhat dependent on the technologies in place at the different jobs, though.  The locations ranged from document folders, floppies, Lotus Notes databases, internal web pages, etc.  Formats ranged from massive “books” to FAQs to simple text files to hyptertexting all over.  I tried to write for other readers besides just myself, but the degree of time I spent on that bit would depend to a degree upon whatever evidence I’d gotten that others there would actually look at what I’d put down.

What bothered me most, but I wasn’t able to articulate at the time, was frustration with trying to keep process and configuration information current — particularly when other people would make changes (as the needs of the projects changed), but not update the documentation.  Occasionally I found out what had changed, and everyone would wait for me to enter the changes (even though I was not the only person with edit capabilities).  Looking back, I think part of the problem was technology that didn’t really facilitate collaborative writing, and another part was a lack of emphasis on a culture of openness and knowledge sharing.  One co-worker “joked” about not writing stuff down as “job security.”  That’s a powerful current to swim against.

• Mining Blog Stories using Community-based and Temporal Clustering
• On the Structural Properties of Massive Telecom Call Graphs (clique v. communities; indivuduals may not know they’re in a “community”; reach is either very small (fewer than 6) or very large (1 million +); indegree/outdegree; even totally anonymized data is of some use, e.g., can tell us how likely word-of-mouth is to work in a given “area”)
• Tracking Dragon-Hunters with Language Models (analyzing scrapes of WoW/other MMORPGs)
• Efficient, Automatic Web Resource Harvesting (OAI-PMH; website alternative to building sitemaps; incremental harvesting of XML metadata; good for intranets; Apache module mod_oai; runs on Linux)
• Pushing Task Relevant Web Links down to the Desktop

One cool thing this conference did was give an hour for poster presenters to present a couple of slides each on what they were going to talk about.  With something like 50 posters, it was very helpful to watch and listen to their soundbites, so I could mark up the poster list (I’d printed it out before attending – it wasn’t distributed in the onsite materials), highlighting some discoveries, and crossing off others which had sounded good (for me) from the titles alone. 

Stats:  there were about 100 attendees, 85% men / 15 % women (by my eye).  Speakers:  Many academics, followed by a fair number of practitioner companies (Microsoft Research Asia, Yahoo, Fuji Xerox).  A few consumers (NIH, me) were just there to soak up what we could.  Perhaps a simple majority of the speakers were from the US, but China wasn’t far behind, then Germany-India-Japan-Taiwan-Italy, and some folks from Canada and Greece.

Weaknesses of presentations:  All too many papers at CIKM were given by substitutes (not the authors) who were unable to respond substantively during Q&As.  Or perhaps I just picked badly; there were usually several paper sessions going on (multi-track).  Too many (1 is too many!) pointless animated slides.  Occasional sociotwaddle (painful when folks wander outside their expertise that way), and in one instance, conclusions drawn from tiny samples and ad hoc methodologies.

A close second was “Geek Spaces” (organization and display of one’s books, art, and collectibles), chiefly due to the enthusiasm of an attendee for the game room project I mentioned (leading to much drafting on paper and brainstorming of possibilities). 

Alas, the Regency Dance was cancelled.  Fortunately, the waltz and tango classes were great.  What does dance have to do with Knowledge Management?  I’ll give you a hint:  buzz, bzzz, buzz…