Remote Collaboration Over Video Data: Towards Real-Time e-Social Science

Abstract

The design of distributed systems to support collaboration among groups of scientists raises new networking challenges that grid middleware developers are addressing. This field of development work, 'e-Science', is increasingly recognising the critical need of understanding the ordinary day-to-day work of doing research to inform design. We have investigated one particular area of collaborative social scientific work -the analysis of video data. Based on interviews and observational studies, we discuss current practices of social scientific work with digital video in three areas: Preparation for collaboration; Control of data and application; and Annotation configurations and techniques. For each, we describe how these requirements feature in our design of a distributed video analysis system as part of the MiMeG project: our security policy and distribution; the design of the control system; and providing freeform annotation over data. Finally, we review our design in light of initial use of the software between project partners; and discuss how we might transform the spatial configuration of the system to support annotation behaviour.

RelicPad: A Hands-On, Mobile Approach to Collaborative Exploration of Virtual Museum Artifacts

“…Annotations are a common feature of collaborative technologies, and increasing interactivity has seen annotations move from being solely about "managing data and metadata" to becoming "critical" resources in "supporting communicative practice" [19]. The Vannotea system used annotations as 'metadata stores' to enable "the collaborative indexing, browsing, annotation and discussion of [video] content between multiple groups at remote locations" [20], while the Kinected Conference sees annotations used to convey users' whereabouts in 3D, using video depth, audio cues and face-tracking algorithms to assign dynamic and interactive context tags to remote collaborators in a videoconference [21].…”

“…For collaborative use, this provides "a persistent record of interaction and collaboration" that can be easily referred back to [23], enabling users to revisit earlier interest points and "remind [themselves] of the process by which they reached previous interim conclusions" [19].…”

The e-Social Science research agenda

“…The MiMeG first-phase node (Bristol University and King's College, London, 2004London, -2007 developed tools to support distributed, synchronous collaborative video analysis (Fraser et al 2006Tutt et al 2007). Digital video has become an invaluable tool for social scientists to capture and analyse a wide range of social action and interactions.…”

Human-centered e-science: A group-theoretic perspective on cyberinfrastructure design

“…As one of the major proposed benefits of CI tools is collaboration within and across disciplines, data management in CI environments necessitates supporting several data models, re-analysis of data under differing assumptions, validation of many computational models, and the exploration of multiple measures of validity and analysis [31]. Furthermore, when sharing data across large distributed teams, there arise issues with how to develop and store standardized annotations (without sacrificing individual researcher ability) [1] and how to develop CI environments that support separate collaborative and individual perspectives on stored data [19].…”