A Taxonomic Framework for Social Machines

Abstract

As the Web has developed into a global social platform there has been increasing interest in a particular class of systems known as 'social machines'. Social machines are typically presented as systems that combine some form of social participation with conventional forms of machine-based 'computation'. Beyond this rather general characterization, however, there is little consensus as to what the term 'social machine' actually means. Furthermore, little has been done to explore the core features of social machines and examine differences between them. This limits our understanding of the kinds of social machine that currently exist, and it also limits our ability to imagine the kinds of social machine that could emerge in the future. In this chapter, we introduce a taxonomy for the description and classification of social machines that could be used to frame the scholarly discourse around social machines and identify aspects of the social machine research effort that deserve further consideration. As part of this effort, we propose a definition of social machines that puts them in relation to the broader class of socio-technical systems, while distinguishing them from other kinds of technology-mediated social participation system; for example, human computation systems and collective intelligence systems. The taxonomic framework we present serves to extend our understanding of social machines. It includes a total of 33 dimensions and 106 associated characteristics. Together, these specify the space of all (theoretically possible) social machine types.

A Collaboration Model for Community-Based Software Development with Social Machines

“…Many crowdsourcing systems can be seen as social machines. Of the existing commercial platforms, of particular relevance here are Topcoder 6 and ODesk 7 , which use different mechanisms to organise diverse participants around software development. As crowdsourcing platforms are becoming widely used as research tools, a number of solutions appeared providing overlay abstractions offering more advanced workflow management and allowing users to perform more complex tasks/computations.…”

“…While automating the coordination and execution management, Jabberwocky is limited to the MapReduce-like class of problems. 6 http://www.topcoder.com/ 7 https://www.odesk.com/ AutoMan [24] integrates the functionality of crowdsourced multiple-choice question answering into Scala programming language. The authors focus on automated management of answering quality.…”

“…One example of such dynamic systems is the field of social machines-systems where computers carry out the bookkeeping so that humans can concentrate on the creative work [4]. This viewpoint can be used to model and produce a diverse class of systems, spanning task-oriented (Wikipedia) to generic (Twitter); scientific or humanitarian (GalaxyZoo, Ushahidi) to social (Instagram) [5,6]. In these systems, interactions between computational intelligence and human creativity are deeply woven into the system, making it difficult to draw a clear line between the human and digital parts, separate their analysis and manage coordination.…”

A Social Machine for Transdisciplinary Research

“…These include, among others, Wikipedia-a collaborative editing site for democratically producing an encyclopedia, Facebook-a social networking system for allowing people to interact with other users, GalaxyZoo-a crowdsourcing approach for distributing the massive task of classifying galaxies and stars, and OpenRov-a distributed sensor network composed of underwater drones and their human handlers for exploring the ocean (Smart, Simperl, & Shadbolt, 2014). A Social Machine for supporting transdisciplinarity may help improve outcomes in such critical areas as sustainability and healthcare research.…”

Human-Machine Networks: Towards a Typology and Profiling Framework

“…Another example, actor-network theory, argues that we explicitly need to take into account that any social system is an association of heterogeneous elements such as humans, norms, texts, devices, machines, and technology, thus granting equal weight to humans and non-human (machine) entities in the analysis of the social [8]. A third and newer perspective, the study of social machines, focuses on online systems that combine social participation with machine-based computation, connecting with Berners-Lee's original vision of the Web more as a social creation than a technical one [9]. Though insightful, these theories tend towards a narrow scope, too restrictive to provide a unified framework for understanding HMNs.…”