A multi-dimensional approach to crowd-consensus modeling and evaluation

In this paper, we propose a multi-dimensional approach to support modeling and consensus management in collective crowdsourcing applications/problems.We define the notion of crowd-consensus, and, for each dimension of analysis, we set pre-defined dimensional levels capturing the different variabilities characterizing the crowd-consensus along that dimension in different applications/problems. The design of a crowdsourcing task for a given target problem requires to characterize the task with respect to each dimension following a pattern-based approach.