The core objective of SOCIAM is to establish the research, methods, tools, networks and collaborations to allow us to understand social machines, in order that they can be designed and deployed by the full range of potential beneficiaries.
The research is complex, as the 'components' of the social machine are both human and technological; the incentives for participation vary widely from personal gain to reciprocity to social responsibility to altruism, while problem identification and solution design are both radically decentralised.
This central goal is decomposed in the programme via six themes, each of which has its own narrower set of objectives.
- Social computation. This has the overall objective of building the computational infrastructure to support advanced social computation. This requires the engineering goal of generating the relevant tools and methods, and the social goal of relating human interaction to computational interaction in social architectures.
- Data curation. The objective of this theme is to design the data infrastructure to support social computation. This entails the development of a computational infrastructure, and the propagation of understanding of the infrastructure through the set of potential social machine users, as well as the creation of incentives to use it.
- Privacy, trust and accountability. The objective here is to maximise the utility of social machines without compromising the privacy of data subjects, the intellectual property of data developers and curators, or the accountability of data users. This will require the development of an infrastructure to support sensitivity to issues of privacy, accountability and security without imposing a large computational overhead on social machines. This theme also has the social objective of fostering and preserving trust in social machines.
- Interaction. The objective of this theme is to design interaction models to support users' defining, requesting and coordinating computation in social machines; where necessary innovating design and evaluation methods for social machine interaction. This requires meeting the engineering goals of developing instances of interfaces/workflows from the interaction models for evaluation; and deploying new evaluation methods across social machines of sufficient variety to validate methods for predictive reliability. In social terms, the objective requires supporting lightweight, "natural" interaction with social machines that will enable new kinds of ways to engage with information for innovation, creativity and discovery.
- Social machine implementation. The objective here is to use the insights from our implementations and our observation of their performance and behaviour to enrich the theoretical understanding of social machines. We can decompose this as an engineering objective to build social machines that implement and embody the insights from our theoretical work, and to feed back insights about how the work in other themes needs to be modified or developed, and a social objective to build classes of social machine that users engage with, develop, extend and from which they can derive value.
- Social machine observatory. This has the objective of classifying and modelling existing social machines in order to support design of new social machines and prediction of behaviour, especially for data-intensive problems. This subtends two further objectives: first, data collection from case studies, analysis of data, simulation, and provision of an instrumented environment to support extensions of existing machines and development of new machines; and second, creation of models to capture the sociotechnical assembly of social machines including the human context and the relationship with existing processes within the case study domains.