Design Assignment
Kathy Lee
SI 884: eCommunities
Winter 2005

Enabling meaningful participation on del.icio.us

Introduction
Increasingly ingenious forms of organizing work, play, and social interactions are in emergenece online. In particular, a class of systems broadly categorized as user-generated content systems (UGCS) have received much attention not only for their unconventional (novel) approaches to organizing such activities, but also their usefulness and sustainability. A few characteristics of these systems are that they are open (anyone can contribute) and that all content is amassed bottom-up from the voluntary contributions of its users.

Research is by no way mature in this area; there is no taxonomy of UGCS or similar systems. They have gone by the names of open community contribution systems (as a class of collaboratories), micro-contribution systems, participatory organization, architectures of participation, or commons-based peer production systems (though, this latter category is a bit broader). Wikis and open source software development systems are prototypical examples.

Yet another emergent class of UGCS add an interesting twist in that they rely (to an extent) on design principles for single-user applications, while incorporating aggregate social effects. For example, applications like del.icio.us∞, flickr∞, or last.fm∞ help manage (traditionally) personal information—bookmarks, photos, music, respectively—while also allowing for sharing with the community at large.

This design proposal looks at the social dilemma present in one specific instance of such systems: del.icio.us∞. The design recommendations suggest ways to mitigate this dilemma by heeding some of social psychology’s most basic principles.

Dilemma: collaterally social
As mentioned above, systems like del.icio.us incorporate aspects of both single-user applications and groupware. Because these systems appeal at a purely individual level of motivation, the social dilemma traditionally associated with online public goods is slightly re-characterized.

As a point of clarification, it may be said that users of traditional UGCS also derive individual utility from knowing that they enrich knowledge on the web via contributions to wikipedia∞ or further scientific causes by contributing to NASA’s clickworkers∞. Still, the primary goal and drive of these systems is some sort of collective action; individual contribution defies the rational choice model while del.icio.us depends on it for its success.

Thus, one point of re-characterization is that the social dilemma in del.icio.us involves less a fear of under-provision of the public good, but more an over-provision that may impose on the community's common pool resource: attentional bandwidth (Kollock & Smith, 1996). This point, however, involves an important assumption about the users—that they perceive of the community's bandwidth as a common pool resource. This leads to another important re-characterization about the mental models of del.icio.us users. Herders that let their cattle deplete the commons or people who vandalize wikipedia are at least aware that it is a commons. But, can we assume that about del.icio.us users? Do they use del.icio.us primarily to share bookmarks or to access only their bookmarks from different machines?

At least for Joshua, the conceiver and designer, the aggregate social effects are secondary in priority. “[M]y goals are, in order: provide a good way for people to remember things for themselves (bookmarks); provide a good way to find other people's interesting things (discovery); and finally, see if the whole system can come together (social)” (Joshua Schachter on delicious-discuss, “Features | Tag Names∞”).

Informal user disclosures corroborate that use is purely individually driven at least among some. “Some people use del.icio.us as a social service, but I think they're in the minority. Most people I talk to (myself included) use del.icio.us as a way to organize and sync bookmarks between multiple machines” (Slashdot thread “Open Source Social Bookmarking Service∞”). This, of course, is by no means a representative sample. But, in designing primarily for the individual, is there potential value that is lost? Or is there a chance that Joshua’s individual and social goals are at odds?

I would argue (and perhaps the user quoted above would agree) that rewarding del.icio.us sessions consist of serendipitous discoveries that are new and relevant/interesting to me. All modes of discovery rely on the system's social aspects: by keeping track of the popular links or tags; drilling down to people who have bookmarked same things as I have; or simply monitoring my inbox. Indeed, the frequency of requests for various filtering mechanisms on the mailing list attests to this; users want serendipitous discoveries to be easier and more customized towards their interests. This is similar to what Reed (1999) terms group-forming networks. “A GFN has functionality that directly enables and supports affiliations (such as interest groups, clubs, meetings, communities) among subsets of its customers.”

Goal: encouraging social use
Thus, the goal of the proceeding design recommendations is to encourage social use on del.icio.us by facilitating the emergence of group-forming networks. Social use leads to meaningful interaction, where meaningful can be defined along two dimensions: not only do users derive social benefit from being in networks with similar others, but they also start becoming more cognizant/respectful of the commons and the common pool resource. We now turn to social psychology to inform the design proposal in two ways: 1) to give us ideas on how to encourage group-forming, and 2) to help explain their postive effects.

The presence of similar others
In order for users to form groups, they first need to become aware of the presence of others. That one is able to perceive the presence of others is a central assumption of social psychology.


Actual, physical presence modifies human behavior in very direct ways: I am less likely to litter when I see other people around. Imagined presence serves as a self-monitoring mechanism that is prompted by the imagined reactions of other people (imagining socially awkward situations). The implied presence of others refers to “the ways that social artifacts (human-made objects) in the environment imply the interests and presence of other people” (ibid).

Fiske goes on to add that “we are social creatures even when we are alone” (ibid), but the presence that is so fundamental of human sociality cannot be safely assumed in computer-mediated communication media. Designers must explicitly provide mechanisms that support social translucence (Erickson et al, 1999). Thus, one goal for del.icio.us is to incorporate more implied presence.

However, everyone's and anyone's presence is unlikely to be of value. Here, we can leverage social psychology’s age-old principle of similarity towards better supporting group-forming networks on del.icio.us. “An individual’s attraction to another is no doubt the most frequent cause of voluntary attempts to initiate interaction with that other” (Berscheid & Reis, 1998, p. 205).

When users are better able to sense the implied presence of similar others, they can also more easily imagine the others' presence. Ostrom (1990) as cited by Kollock & Smith (1996) observed this very mechanism of imagined presence at work in real life communities that were able to successfully manage common pool resources. “A system for monitoring member’s behavior exists; this monitoring is undertaken by the community members themselves.” If I know that I am part of a small del.icio.us group, I am more likely to post meaningfully (e.g., tag and annotate meaningfully, avoid del.icio.us-ing sites like slashdot/google/hotmail, etc) in light of the imagined reactions of my subscribers.

Combining presence and similarity thus, we have something akin to a social recommender system that serves to moderate the behavior of del.icio.us users, making for more meaningful interaction.

Re-designing del.icio.us' posting interface
How can design bring out the implied presence of similar others? The inbox—where users receive their subscriptions—partially serves this function. As seen in the user testimony above, however, not all users take advantage of it. On the other hand, the bookmark posting interface∞ is accessed universally by users with individual and social tendencies alike, and is a better place to target. There are currently two interfaces for posting bookmarks, but neither contains much implied presence. The 'experimental' interface presents tag recommendations (what you should be tagging this particular URL) based on a comparison between the user's tags and commonly used tags from the community at large (for the specific URL). While these recommendations implicate the presence of other users who have gone and tagged before, there may be more direct ways of implying presence.

I suggest that del.icio.us’ posting interface generate a list of other similar users who have bookmarked what the user is about to bookmark. The similarity algorithm could be based on number of common bookmarks, commonly used tags, or some combination of both. For users who are not yet part of a group (who do not have inbox subscriptions), this list could be a random sample. In addition to a random sample, users who already belong to a network may also be interested in seeing a more contextualized list, i.e., ‘who, in my inbox, has bookmarked what I’m about to bookmark?’ or ‘who in my second degree network has bookmarked this?’

These additions not only highlight the implied presence of similar others, but also serves to moderate less meaningful posting behavior. If I see that most of my subscribers have already bookmarked what I am about to bookmark, their imagined reactions may cause me to reconsider re-posting the bookmark, for want of not flooding their inboxes.

One metric by which to measure the effectiveness of this intervention is the growth in number of inbox subscriptions to other users, i.e., did users who were previously not subscribed to anyone add any inbox subscriptions? Another way to represent and analyze this data would be to use social network analysis to gauge the presence and growth of group-forming networks.

Conclusion
The underlying assumption of this recommendation is that people will contribute more meaningfully to a collective (while curbing purely individual motivations) when they are able to sense the presence of similar others. Not only do they benefit from the bookmarking efforts of others with similar interests, the imagined presence of these others also serves to keep individual behavior in check. While the implementation above is specific to the del.icio.us user interface, these principles can hopefully be contextualized for any user-driven content system where meaningful contribution is of the objective.

References
Berscheid, E., & Reis, H. T. (1998). Attraction and close relationships. In D. T. Gilbert & S. T. Fiske & et al. (Eds.), The handbook of social psychology, Vol 2 (4th ed., pp. 193-281). New York, NY, US: McGraw-Hill.

Butler, B.S. (2001). Membership Size, Communication Activity, and Sustainability: A Resource-Based Model of Online Social Structures. Information Systems Research. 12(4): 346-365.

Erickson, T., Smith, D. N., Kellogg, W. A., Laff, M., Richards, J. T., and Bradner, E. (1999). Socially translucent systems: Social proxies, persistent conversation, and the design of "Babble." Proc. CHI 1999. 72-79.

Fiske, S. T. (2004). Social beings: A core motives approach to social psychology. New York: Wiley.

Kollock, Peter., & Smith, Marc. Managing the Virtual Commons. In Computer-Mediated Communication: Linguistic, Social, and Cross-Cultural Perspectives, edited by Susan Herring. Amsterdam: John Benjamins. 1996. pp. 109-128.

Reed, D. (1999). That Sneaky Exponential: Beyond Metcalfe's Law to the Power of Community Building

Comments

LaurieBuis: Nice paper! I thought it would have also been interesting if you had addressed the opaqueness of del.icio.us and made some suggestions for how to improve that. I know that several of us have been discussing this semester that del.icio.us is a cool site, but there seems to be a barrier to access in that it isn't intuitive to use.

PaulRenick: This is excellent! I don't want to detract in any way from your focus on not polluting the commons. But I think you might also consider the public goods aspect of producing valued tags. If you produce tags only taking into account the benefit they create for you, then you'll produce some, but possibly not as many as you would if you also took into account the benefit they provided for others. You might again try to draw on identification with a group of similar others to inspire a design where people would somehow find out how useful their tags had been to others "like them".