Improvisation often requires performers to anticipate each other's upcoming actions to coordinate their performance. Traditionally, improvisation does not involve explicit representation of upcoming actions, relying instead on implicit cues and overt gestures. Providing a concrete representation of these actions may impact the nature of improvisation. To explore how performers use such a representation, we introduce the "temporal window", a shared workspace for collaborators to indicate upcoming actions shortly before those actions are carried out. We describe versions for music and storytelling, and present results of an observational study of use of these systems by 11 pairs of participants. Participants used the temporal window in various ways, including using knowledge about upcoming actions when performing live, and adjusting upcoming actions based on knowledge of both the live performance and upcoming actions. These observations can inform design to support improvisation as well as other systems that support anticipating upcoming actions.

Reuse of the works of others has become common practice on the Internet and has formed the basis for collaboration in some online communities. However, some works are reused much more frequently than others. In this article we build a quantitative model that explains which factors are most salient in determining the likelihood that an author's work will be reused. Controlling for common factors, such as the work's popularity, we show that the probability of reuse depends on (a) the degree of derivativity of the work in question, (b) the specific ways in which it derives meaning from other works (intertextuality), (c) the audience's preferential attachment to authors of high fecundity, and (d) the author's social embededness in networks of reuse. We use trace data from an online community that was built for the purpose of demonstrating the ability of open sharing and reuse to spur collaboration and innovation in music. Although our model is designed for broad applicability, we explain that the size and direction of the effects reported in this paper may vary, when reuse is performed with other media or for different purposes.

Categorization of online videos is often treated as a tag suggestion task; tags can be generated by individuals or by machine classification. In this paper, we suggest categorization can be determined socially, based on people's interactions around media content without recourse to metadata that are intrinsic to the media object itself. This work bridges the gap between the human perception of genre and automatic categorization of genre in classifying online videos. We present findings from two internet surveys and from follow-up interviews where we address how people determine genre classification for videos and how social framing of video content can alter the perception and categorization of that content. From these findings, we train a Naive Bayes classifier to predict genre categories. The trained classifier achieved 82% accuracy using only social action data, without the use of content or media-specific metadata. We conclude with implications on how we categorize and organize media online as well as what our findings mean for designing and building future tools and interaction experiences.

This paper describes an ongoing research project exploring the business practices of mobile entrepreneurs, specifically mobile food vendors, and the potential use of location aware and mobile phone-based application to support their information needs. Mobile food vendors are a ubiquitous phenomenon in the developing world and can be seen hawking their wares in carts, bicycles, or motorcycles. In this paper we report findings from 28 interviews around business practices, challenges, personal histories, and economic wellbeing among mobile food vendors and 10 of their customers from five major Indonesian cities. Based on these findings, we point to characteristics and features of a mobile phone-based application that will enable these vendors to advertise their current location, accept orders from customers, enable customers' recommendation of vendors, and inform fellow vendors of various special events.

Open contribution (OCS) are online applications that encourage users to contribute and share content in a public and open manner. While these systems lower the barriers to participating, what is less clear is why users are motivated to contribute time and effort in these online environments with relative strangers. My dissertation proposes that one way to explain high levels of participation on OCS is to use the lens of social performances. This lens suggests that individual participation on social software websites involves elements of both individual and collective performative behavior. The social performance framework suggests that the participatory behavior is part of a larger sensemaking exercise that rationalizes and aligns individual contributions to the collective effort. The view here is that OCS and its users are part of a socio-technical ecology and are mutually dependent on each other. Understanding participation as a form of social performance can enable us to better design systems that encourage participation, collaboration and sharing.

This paper describes a community-based effort to preserve organizational knowledge and to orientate newcomers to a graduate school. It presents a very brief review of recent research on wiki use in corporate and organizational environments and initial data from two wiki implementation iterations within our academic community. We contrast use of a TWiki with that of a WetPaint wiki. Our data suggest that with low barriers to participation and a great deal of patience, wikis can be useful stores for community information and knowledge sharing.