Building scalable virtual communities - Infrastructure requirements and computational costs

Abstract

The concept of a "community" is often an essential feature of many existing scientific collaborations. Collaboration networks generally involve bringing together participants who wish to achieve some common outcome. Scientists often work in informal collaborations to solve complex problems that require multiple types of skills. Increasingly, scientific collaborations are becoming interdisciplinary - requiring participants who posses different skills to come together. Such communities may be generally composed of participants with complimentary or similar skills - ho may decide to collaborate to more efficiently solve a single large problem. If such a community wishes to utilise computational resources to undertake their work, it is useful to identify metrics that may be used to characterise their collaboration. Such metrics are useful to identify particular types of communities, or more importantly, particular features of communities that are likely to lead to successful collaborations as the number of participants (or the resources they are sharing) increases. © Springer-Verlag Berlin Heidelberg 2005.