How the Web Sees the Makice Family

On the surface, there appears to be some value in compressing all of the blogs, articles, and social media postings into a compact visualization. The high-level descriptive words about a person are easy to understand. In looking at our three most active online family members, one striking observation is that 9-year-old Carter has 20 category descriptors compared to just 11 for Amy. The outcome doesn’t completely make sense: I’m a designer, but politics and management both eclipse that category in my visualization.

Does it mean anything? Potentially, it could, if you revisit the process over time.

It would be interesting to see how Carter’s description changes as he continues to publish his own material, rather than relying on his parents to share his life online. Ambiguous design like this is a collaborative process between the viewer and the system to co-create meaning, often by comparing what one sees now with what was experienced previously. To mine value in this way, though, there needs to be some longitudinal consistency in how the data is processed.

That consistency isn’t there. I ran the same search for “Kevin Makice” three times in succession. With each run, Personas created a different interpretation of the same data:

Blink, and the interpretation changes

That’s OK, because this is one of the quirks the creators want to reveal.

Personas is a critique of data mining. While acknowledging the real value Google and Netflix bring to people through statistical analysis of large data sets, there is also a dark side that includes TSA watch lists (to name one). The Personas site explains this insight:

Data mining is “technologically neutral” in the sense that its power is derived from what people do with it. The creators of an algorithm choose how to model the world, deciding (somewhat arbitrarily) what inputs and outputs to use. You as the potential “victim” of data mining cannot control any of these factors, especially given the usual lack of transparency of the process.

The purpose of this project is to allow people to peek into one such black box, while still preventing access to controls to shape its engine.

Inside the Black Box
The analytical process of Personas starts with a Yahoo search for public data using “characterizing” queries (which are different from a simple ego search) and limiting the results to no more than 30 items. Some filtering is done on the results to remove hate speech and focus on English language, and words are also stemmed to remove suffixes and simplify the data sets.

Using a technique called Latent Dirichlet Allocation, the results are categorized with an unsupervised algorithm—which means the computer doesn’t know if its work is “correct.” In this case, a method of clustering called topic modeling was used to guess which categories best describe a given document.

Personas was created as an art installation by Aaron Zinman (with help from Alex Dragulescu, Yannick Assogba and Judith Donath) as part of an interactive exhibit—Metropath(ologies)—by the MIT’s Sociable Media Group. Zinman is a PhD student whose past work included “Is Britney Spears Spam” (PDF), an attempt to classify users by the humanness of their communication behavior and social structure.

The widget is one of a dozen or so Internet systems developed to support their mission of transparent government. The Sunlight Foundation relies on Internet technologies to get the truth out into the open. They support, develop and deploy new online systems to make information about Congress and the federal government more accessible to the American people, fostering more openness and accountability in government. These projects include:

• Capitol Words—For every day that Congress is in session, Capitol Words sums up the day with one word (the most frequently used word from the Congressional Record).
• Punch Clock Map—Transparent politicians volunteer their itineraries for public review through a Google Map
• PoliQuiz—an interactive political trivia game.
• PublicMarkup.org—gives you the opportunity to review and comment on proposed bills before they are even introduced in Congress.
• Where Are They Now?—Community effort to track whether and where former Congressional staffers are employed as lobbyists.
• Party Time!—Tracks where and when politicians are partying.

Other Sunlight projects are inspired by such things as the Foreign Agent Registration Act, new legislation, personal finance records, earmarking, and citizen journalism.

Launched in 2006, the Sunlight Foundation works hard to get citizens involved with the process of governing through awareness initiatives like OpenCongress. This site collects a wealth of information from sources like GovTrack.us, Google News, Technorati, and OpenSecrets.org, turning them into blog reports on what is really happening in Washington, D.C. “We think everyone should be an insider.”

Past initiatives included the successful Let Our Congress Tweet campaign earlier this year, which responded to rules changes that would keep elected officials from using Twitter. Sunlight also sponsored a visualization contest. The winner was Unfluence, a network querying tool that showed the context of political donations, connecting candidates to corporations.

Bill Moyers describes the Sunlight Foundation

Big Treetop, for example, recently held a demo party for their beta testers. This Tuesday, SproutBox will be working a room full of “buzz builders” to kickstart their project. And after working on their business for two years, isocket is starting to put their advertising platform into action.

isocket is a local startup creating a custom advertising platform

Part of the motivation for our recent face lift ("BlogSchmog goes to the spa") was to prepare the way to be an active beta tester for isocket, a development platform for the advertising industry. We carved out a couple places for advertising badges in the new theme design: on the home page, in the footer, and in the sidebars next to every article. Interested advertisers can click on those placeholders and be transported to the isocket site with instructions on how to purchase the ad.

This week, we sold our first 7-day homepage ad to PlanetHTTP, an Internet hosting service.

A place where opportunity grows
isocket is the first commission-free advertising platform—they don’t charge for this transaction. Their business model is dependent on cultivating a transparent environment to support interactions between the buyers and sellers of ads. isocket will eventually offer up their data to interested developers to create the widgets, tools, and services to keep the transactions flowing.

Buyers can plan and manage “360°” ad campaigns, allowing the same message to be placed all around their target audience. Unlike with larger ad businesses, each buyer controls exactly where every dollar goes by searching isocket’s marketplace for the best deals and best audience match. Sellers earn more revenue because no commission fee is charged, but they also have control over which ads they accept. isocket facilitates direct sales by exposing smaller and niche ad venues to more advertisers and by collecting payment.

Down the road, sellers will be able to decide how the price is determined. Closed bids, possible now, allow the seller to set the price. He asks, “Who wants this space?” and a buyer responds, “I do.” Open bids allow the market to determine the value. The seller declares, “I’ve got this space,” and buyers inquire, “Can I have it?” When bids are collected, the seller retains the right to accept her preferred deal.

The simple click-to-buy process took just minutes to set up. To create my available ads, I submitted a few simple forms on the isocket page to define two BlogSchmog ad sockets. Each socket is assigned to a specific ad configuration, determining color and size. Adding two lines of script to the blog theme’s design generates the ad placeholder. When PlanetHTTP was interested enough to place an offer, I was notified to return to the isocket web site to review and ultimately approve the new ad for the listed price. Moments after doing so, the new banner showed up on our blog.

Advertise anywhere
Eventually, isocket will allow for other ad channels, too, beyond web site banners. Widgets, applications, email, video, digital billboards, and traditional forms of advertising are all expected to be supported. From the long list of options in their web forms, here are just a few ways our lives can be offered up for ad revenue:

• Websites—Beyond what is described above, web site campaigns can be engineered in creative ways to allow ads to change based on date or category content. Each adcode can be programmed to dynamically select from a list of created sockets, rendering the most appropriate one based on the content being displayed.
• Video—Product placement for the YouTube age. Imagine Carter and Archie fighting as Jedi with a banner for The Game Preserve hanging on the wall behind them.
• Social Networks—I could sponsor every 20th tweet with a link to some local business. Or maybe a restaurant can buy me lunch if I tweet about eating there.
• RSS Feeds—Subscribers to BlogSchmog could get a small link to a business or organization added to the bottom of every post.
• Podcast—There would be a much greater incentive to create a local area or domain podcast if I could find a sponsor ahead of time.
• Large Signage—Our roof could be crafted into an ad visible to passing planes and satellites … including those taking pictures for Google Earth.
• Athletic Sponsorship—The next time Amy runs a marathon, perhaps Smith’s Sport ‘n Shoe would pay for childcare while she trains in exchange for wearing their logo on her running shirt.
• Conferences—It would be much easier to find sponsors for another local Startup Weekend with such an easy mechanism to define and recruit financial help.
• Parties—Archie’s fifth birthday in October could be sponsored by Target and bring him all the Bionicles his heart desires. There might even be some YouTube footage of the cake-n-candle ceremony with opportunities for bonus product placement.
• Clothing—Every Wednesday, I’ll wear a t-shirt with a company or product logo and information as I walk around town, interacting with people.
• Body Art— How desperate are we for money to get us through the final years of grad school? Enough to consider tattooing the Bloomington Bagel Company logo on my hand. (I’m guessing the price would be pretty steep on that one and include payment for a post-graduation procedure to get it removed.)

With isocket, advertising transactions aren’t limited to banner ads on web sites. Sockets can be created for almost any channel that a seller can imagine and a buyer finds appealing enough to purchase.

isocket is currently in private beta, launching their initial banner ad service this summer. The number of ads served has grown by 30-40% each week, possibly altering their funding strategy. The team may decide to wait on major investment capital and continue to expand services on their own.

A map of digital identity. (source: Fred Cavazza)

Anonymity is one of the pillars of online communication. The ability to become someone else, mask some part of yourself, or lurk in the shadows increases paths to participation. The dark side of anonymity is irresponsibility, and we have already witnessed other social networks tackle Jon Swift problems by censoring their own communities to strengthen legitimacy. Even as we collectively accumulate personal profile pages that express our real identities, however, there are initiatives emphasizing anonymous disclosures. The Experience Project is designed around anonymity, asking members of the two-year-old community to connect through their experiences rather than extending existing social circles. In the end, though, this approach is about lowering barriers for people who could not otherwise participate in discussions. The impact of what is shared is dependent on the relationships we form with the identities we assign to ourselves and others.

The consequence of self-disclosure centers around the resilience of information.

Twitter, a microblogging service that exploded to a million members in about a year, uses the SMS constraint of 140 characters to lower the barriers to entry for potential authors. It is much easier to conceive of sharing a simple sentence or two than several paragraphs. The custom nature of the personal information stream (everyone can decide whose content they want to follow) implies a sense of control. However, the reality of Twitter is that the content is public. Even with private streams—where a member can require a mutual handshake before someone else can see their posts—the act of sharing content with anyone exponentially increases the likelihood that information will reach a public audience. The age of the intelligent web is here, and innocuous posts made in a semi-protected context one day can give rise to unexpected revelations in the future.

This has implications on future career paths, as comments in a Web Worker Daily article last September attest. Tim O’Reilly expects a Web 2.0 backlash and a return to private data. Perhaps. At the start of the year, Duncan Riley published a poll asking, should some things remain private in the age of lifestreaming. The nature of that flawed question led to a predictable response—less than 10% of respondents said “No”—and false evidence that we disclose too much about ourselves. A more relevant line of questioning would be what kinds of information should be private, for ourselves and from others.

Being naked isn’t only about what you publish online. Dartmouth’s Alice Mathias wrote in a New York Times op-ed piece last fall that the only privacy setting in Facebook that mattered to her friends was the one revealing if they are logged in. She speculated that the ability to search anonymously is significantly more important than remaining anonymous when other people search: “If our ability to privately search is ever jeopardized, Facebook will turn into a ghost town.”

Rob May had another take on the professional cost of social networking. Using the life of economist and political scientist Joseph Schumpeter as an illustration, May observed:

The web makes it easier than ever to connect with new people, but the flip side is that it also keeps us connected to people from earlier times… people who may not understand or support our goals. Is it possible that in some instances, social networks hold us back? In earlier times students could go away to school and carve their own path, but now with old friends judging every move we make, are we likely to be less unique, less aggressive, and perhaps not live up to our creative potential?

Molly Holzschlag prefers the term authenticity to transparency, criticizing the later as implying a reality that may not be present. There is some degree of self-censorship that goes into any interaction with the world, be it this blog, my Twitter account, or a conversation I have on campus. Throughout it all, I aim for authenticity. My own hypothesis is that greater transparency leads to more empathic censorship, not decisions to omit based on fear or power. Being open requires more awareness of our networks, which leads to a higher consciousness about communication.

Keeping track of locals depends on accurate self-disclosure.

My interest in Twitter revolves largely around microblogging as a conduit for strengthening local community. I routinely seek out Bloomington-area authors, adding them to a master list in the Hoosier Twitosphere and following many of them. Sometimes, it is reciprocated and our social circles join. We inspire each other through this ambient connection, planting the seeds of more meaningful relationships capable of addressing business, emotional and intellectual needs. None of this is possible without a transparent culture willing to share something as small but significant as their location.

The value of local connections cannot be underestimated. The impact of transparency is routinely exhibited on Homeless Man Speaks, a blog about a man experiencing homelessness in Toronto, as transcribed by a neighbor:

So, of course, the guy gets out of the car, of course he’s pissed at me, and I’m telling you, he’s six-foot-something and he’s got biceps that could crack walnuts. So he looks at down me and says: ‘I know you. You’re homelessmanspeaks.com.’

Small steps. The journey begins by gaining awareness of those around you.

What information do you keep private? What should others keep private?

Battletweets is a group Twitter game, currently in development.

According to the FAQ, Battletweets is a team-based game for the Twitter community. There are plans to host different weekly activities, such as web scavenger hunts, pic & post, trivia, wikijam, and others. Initially, the beta release will feature two established teams the rest of the world can follow along. Winning teams earn badges, with the hint of other more substantive prizes possible as the game develops.

I have two very diverse opinions about the kind of use for Twitter that Battletweets proposes. Since it is a good news / bad news kind of thing, I’ll start with the bad:

Battletweets will produce a lot of noise.
For the same reason I think Tweet 140 is a good idea with bad consequences, my big concern with using Twitter as a platform for a game. The value of Twitter for myself and many others is primarily its ability to help members feel connected to others despite gulfs of both time and space. Each person can shape their own information stream by following only the other members they want to know about. If the noise becomes too great, they can stop following, maintaining control over their stream. When the noise becomes mixed—which includes both frequent directed conversation (@replies) and artificial content (like making a goal to use all of the available 140 characters, a twoosh or perfect tweet)—that separation goes away. People I want to or even must follow now have a split identity that cannot be separated. One is the individual whose connection is important, and the other produces high-content noise that has no relevance.

Battletweets will engage the community.
It isn’t that Twitter games are inherently bad. Minnesotan Aric McKeown has been using Twitter for a local version of Where’s Waldo, sitting in local coffee shops and tweeting clues to his whereabouts ("Hide ‘n Tweet"). The Least Dangerous Game is something that would work well here in Bloomington as a means to mix online and offline, and to create awareness about local businesses and organizations. Battletweets may do some virtual-physical mixing, too, but even if it doesn’t, the games are task-based group activities open to the world. That means, people can engage on several levels, befitting their interest and available resources, and connect to people simply by paying attention to what they contribute. If Battletweets takes great care in their games to emphasize the phatic nature of Twitter—to base the games around revelations of individuals about their day-to-day identity—these games could enhance the content of the personal information stream.

What I find most compelling about Battletweets at the moment is their development process. When I first visited the site, there was a logo (with a needless reflection of the image) and two buttons, one linking to their twitter account and another with a contact email address. On my second trip, they added a FAQ and a currently-dormant link to a blog. They also have a little progress bar to show where they are in their process, a nice little use of tech culture that is very effective, provided it continues to move. With my recent fascination with startup companies, I decided to pay attention mostly for the promise of transparency.

After we passed New York City in the number of participants registered to attend, the Herald-Times published a front-page story on our BSW event at City Hall next weekend. Thanks to the generosity of Richie, I have my name attached to both a Work-In-Progress and a Workshop for CHI, avoiding the shutout for my trip to Italy in April. Overall, estimates are 33 students and faculty will participate in the CHI 2008 program.

The sad news for my design team came in the form of mixed reviews on our own submission, one of just 40 international submissions by students willing to take on the wicked problem of designing for people experiencing homelessness. More than 1/4 of those submissions came from Indiana, which is one of the reasons we were able to put five teams into the second round.

Community Weekend: Facilitating Social and Economic Opportunity

Abstract:
For people experiencing severe poverty in the United States, homelessness can be an involuntary symptom of a complex social ill. The cost of being homeless is the vicious cycle of deteriorating social support and increasing barriers to economic opportunity. A proposed local intervention—Community Weekend (CW)—presents people experiencing homelessness with a viable exit strategy based on human connection and in situ exposure to practical skills. Participation in the shared task of developing a new local business decreases stigma and establishes authentic ties between the homeless and the homed. A web-based visualization of each participant’s evolving social network is generated effortlessly with sociometric badges. The CW connection manager provides a persistent means of contact even for those with transitory lifestyles.

While the overall scores were lower than we would have liked (and needed, to advance), we have a nice rebuttal for the critiques. Our mPath entry that won it all in 2005, I recall, had equally mediocre reviews in the contest’s second year. I include oiur reviews below to (a) solicit more feedback, and (b) help give future student teams some insight into the review process for the competition.

Review 1
Design Process Quality: 2 (Poor)
Design Solution Concept: 3 (Average)
Feasability of Design Solution: 2 (Poor)
Clarity of Writing & Presentation: 3 (Average)
Overall Rating: 3 (Borderline: Overall I would not argue for accepting this paper.)

Review Comments:
The authors review appropriate literature on economic and social determinants of homelessness. Additional factors are also discussed. Benefits of job training padded with other holistic support and training are discussed, with focus on the example of the Startup Weekend initiative where participants from different backgrounds spend a weekend together to develop start up businesses and nurture social connections. The authors propose a derivative solution: Community Weekend, where members of a community spend a weekend with homeless persons on community projects. Smart badges are proposed to track conversations over the course of the weekend, and visualize these connections over time.

The authors propose a sound approach to helping the homeless on multiple fronts: skills-wise through holistic training, and socially through prolonged contact with homed individuals. The authors’ proposal was apparently informed by blogs, however the vague mention of online artifact analysis or user involvement (bottom of p3) is not clear. The number of users and/or websites are not specified and the analysis methods are not described. The proposal of badges and communication visualization do not seem useful for the homeless because it is unclear how they would benefit considering the common lack of computer access. Also, it would be helpful to see a prototype of the visualization. The authors need to explore barriers to this proposal, such as finding willing homed and homeless participants, and connecting them with realistic, integrative tasks and goals.

Of the three reviews, this is the most constructive and informative. The ratings of “2″ are harsh, but understandable given the emphasis on showing process. In a short-form six-page paper, however, we made a decision to emphasize the rationale for the argument at the sacrifice of some of the things mentioned (screen shots, detailed discussion of web sites reviews, etc).

There were two primary reasons for this decision, one practical and one philosophical. The practical one is that our three-member team only formed in mid-December, without an HSC-approved study we could use as a foundation. Relevant inquiries were conducted, and we did manage to push through two studies before the paper was submitted (only benefiting from one). We met only once in the same room, as one member was moving as the other was becoming available, and worked over the holiday season. A steep hill to climb, man-hours-wise. Our process was designerly, but decidedly informal.

Philosophically, our design required buy-in to two key notions. First, homelessness is a byproduct of economic disadvantage, not simply an absence of home. Second, developing positive social networks outside of the peer group is vital to infusing resources into that community. Spending two pages explaining which web sites we analyzed and how we iterated web screens wasn’t going to get those points into the foundation.

Eight pages and two more weeks, we would have been golden.

Review 2
Design Process Quality: 5 (Very good)
Design Solution Concept: 4 (Good)
Feasability of Design Solution: 4 (Good)
Clarity of Writing & Presentation: 5 (Very good)
Overall Rating: 5 (Definite accept: I would argue strongly for accepting this paper.)

Review Comments:
This is an impressive submission from start to finish. Secondary literature is used sensitively and effectively. Both social care literature, existing designs such as StartUp weekend, and emerging technologies such as MIT’s smart badges are used as a source of ideas and inspiration. This has been shaped, deepened and grounded by collaboration with local domain experts (Acknowledgements). Although the design is at an initial conceptual stage here, there is a rich enough feature set under consideration for iteration towards worthwhile outcomes to be possible. Although I am not absolutely confident on feasiblity, I do feel that the proposed design can be scaled back and still be effective if some design options turn out to be too costly or challenging. Note that technical feasibility isn’t an issue, it’s the effectiveness of a design synthesis with all the envisaged features that is an issue at this point. This is a consequence of an ambitious design, but I do not feel that it is overly ambitious. The breadth of generated options (49) is very impressive, and gives the group many opportunities to steer through the design space to an effective outcome.

I’m very impressed by this front end to a design process. The team should definitely advance to the next stage. Figure 4 and its caption provides reassuring evidence of how far the group have thought ahead.

PS There’s typo on p.6 col 1, form should be from, first line of first new paragraph.

Clearly, this reviewer got the rationale. This was the kind of reaction we were hoping for when we structured the paper, including the use of margins to address the process and technical gaps that our more extensive literature review would not permit. This reviewer accepted our argument, was OK with the high level of concept, and gave us the benefit of doubt that we would be able to iterate sufficiently by April. Check, check and check.

Review 3
Design Process Quality: 3 (Average)
Design Solution Concept: 4 (Good)
Feasability of Design Solution: 3 (Average)
Clarity of Writing & Presentation: 4 (Good)
Overall Rating: 3 (Borderline: Overall I would not argue for accepting this paper.)

Review Comments:
The proponents have not done extensive enough research into the “pathways” to homelessness and, as a result, cast their proposal too narrowly. Economic development opportunities are clearly important, but they are not sufficient to address the fundamental questions of shelter, food and income (to name just three). The central role of housing (both as a factor in causing homelessness and a solution to homelessness) is mostly absent from the discussion, and therefore sets in place a fundamental flaw. I’m not clear how the community weekend model would work without the fundamental stability of housing. That said, the proponents do a credible job of matching their proposal to the stated goals. However, due to the fundamental flaw that I have outlined, I would not recommend that this paper be accepted.

This review was the sink in our sunk.

Although technically scored higher than the first reviewer, this critique is based largely on a literal definition of homelessness as one who is without home. We explored the meaning of the term deeply, drawing different affinity diagrams and relational charts to figure out what it meant to experience homelessness. In the end, all roads led to economics, not physical shelter. To design for “shelter, food and income” offers a band-aid to a much more complex dynamic. Those problems involve a short-term necessity; we were attempting to address homelessness as a symptom of a greater social ill.

It’s the fish vs. fishing. Clearly, even addressing the credible weaknesses of the first reviewer would not have been enough for this critic. In order to pass muster, we would have needed to revert to a more tangible and limiting definition of the problem that we rejected after a lengthy exploration and deliberation of the options.

What do you think?

My expectations were a bit higher this year than last, when I submitted mainly to get in the habit of creating short-form documents about research interests and meeting deadlines. Since the consortium is looking for students who “have a clear idea and an area, and have made some progress, but who are not so far along that they can no longer make changes,” the reality is that a Ph.D. student who hasn’t even cleared the qualifying exams (that comes a year from January) is a long shot. Still, the theme of this year’s CHI is “Art. Science. Balance.” … which seemed like a good fit for my interests.

In the interest of transparency, here is the long abstract—”Complexity in Interaction Design“— and the comments from the two reviewers. The first reviewer wrote:

The idea in this DC-proposal is that we can analyze online communities in terms of “complex systems” and design them bottom-up rather than top-down. That is, the different components (e.g. the activities of the users) can be seen as independent units that together form the system, actively constructing themselves and thereby the overall (complex) system. This is different from a top-down perspective where the system “enforces” certain behaviors through its design.

What worries me about this proposal is that we do not get presented with a lot of solutions yet, only a set of high-level concepts like flow, fuzzy determinism, etc. Since the author has been doing his PhD-studies for quite a while, I would have expected something more? The concepts presented look really interesting but I am not certain how they translate into design? I know for sure that designing for flow is a hard task. And the other concepts do not look any simpler…

Very little has been published yet.

And from the second:

This is an interesting topic, but I wonder about the join between the theory (complexity) and the application target- e-communities. There is a considerable work on e-communities in CHI (see Preece Maryland, Kraut, CMU), so improving design of community systems needs to take this into account. Also I think the theory angle on complexity needs to be
sharpened, some of the influences you have chosen have little to say about complexity (e.g. Flow), and if you are taking a contextual approach better to look at McCarthy and Wright and Dourish’s books. How will you validate your complexity framework or demonstrate its utility ?

I draw a few conclusions from this, most notably about the selection criteria.

I’m in the second year of a Ph.D. program, only two years into contemplating this notion of complex design, among other interests. What the first reviewer is telling me is that, in order to be considered next year, I not only have to up my pedigree through publication but also develop the thoughts sufficiently to have reasoned and supported conclusions formed. In practical terms, my CHI 2009 entry won’t have a snowball’s chance, either, and I might be best waiting until CHI 2010 to have a good candidate. Unfortunately, there is a better chance that I could have my dissertation completed before that particular conference begins. The ideal time to get the feedback would be in the spring after my quals, and to get accepted I need to spend the next ten months focused on publishing some preliminary ideas that will make the complex design idea official.

The short form of the abstract is a bit unwieldy, too. The challenge of a big new theoretical idea is that there are a lot of concepts to not only introduce but synthesize, and there are many resources that could be cited to help this. In a 4-page CHI format abstract, though, there isn’t enough room to do so. Does that mean my idea for a possible dissertation is too broad? Maybe. It could also be that the constraints of a 4-page document bias the selection toward very narrowly-focused and traditional research.

The comments are helpful in reinforcing what I already know: this is going to be a difficult idea to explain. The depth of research in any one of these ideas (situated action, complexity, flow, criticality, online community) is sufficient to guarantee that even in a full paper I’d have to leave out relevant archival work. I’m also dealing with talking to two if not three distinct communities with this research: complex systems scientists, human-computer interaction folk, and designers. It is a tall order to demonstrate both interest and relevance to all groups when there is also some educating needed to bring the groups up to speed on the basic concepts and terminology. Perhaps the only way to do this is to publish the groundwork first, citing those papers as a means of cutting to the chase in the Doctoral Consortium submission.

In the short-term, I’m going to let this research rest for the holidays and pick it back up after the new semester starts. I would appreciate any thoughts you have on the short paper, submission process, or complex design as a possible dissertation.

A trailer for a documentary on Wikipedia.

Wikipedia is having a mid-life crisis of sorts. Robert Rohde reported that much of the constructive activity—article edits, new accounts and uploads—are in decline, as are some of the signs of controversy, such as blocks, protection and deletion. The most notable increase is in edits to revert to a previous state, a sign of spam or inanity.

There are two ways to interpret this, of course. Either Wikipedia is actually in decline, or that this is a natural stage of community evolution. I believe it is the latter.

First, there are several new tools that effectively re-paint Wikipedia in new norms. WikiScanner got a lot of press this summer by using IP lookups to take the anonymity out of some key edits, but that isn’t the only new Wikipedia tool for new discovery. WikiRage is a new website that keeps track of the top 100 article edits in a given period of time. This can show what stories are getting the most attention. Further, it is coded to identify articles with a history of vandalism, those under dispute, and current viewership. The wiki dashboard is PARC project to improve the attribution of an article by visually showing the editing activity detail of its authors.

The arrival of these tools onto the scene changes what people can know about the Wikipedia site, improving the accountability and making more transparent the meta context behind each article. When the environment changes, so do the agents as they adapt to their surroundings. The WP stats decline could be due to a combination of factors, ranging from new roles to the elimination of masking behaviors. I’m anxious for Rohde or someone else to do a follow-up in six months.

Second, there is evidence that the traditional roles are evolving. Anonymity began as an inherent right, and moved toward symbolizing deception and bad intent. According to a recent Dartmouth study, however, anonymous users contribute as Good Samaritans. Again, this may be a sign that people are adapting to the current environment, or it could be that the pool of anonymous users has been pruned of its worst element.

Evolution also implies death, at some point. Wikipedia has some competition. It is nothing new to have someone upload some files, write some community rules, and claim to be better. Sustainability is the big concern, but it is a knife that cuts both ways. The Wikipedia community is divided into two primary groups: editors and consumers. The former is loyal but very diverse. It is common to suffer from burnout or move on to other projects. If one of those projects is something like Citizendium, the content may start to materialize. The latter group is going to Wikipedia now because (a) they have heard of it, and (b) it comes up highly placed in Google searches. As search paths change, so too could startup encyclopedias find some traction and Eric Goldman’s prediction of failure by 2010 will come true.