Crowd-based data-driven hypothesis generation from data and the organisation of participative scientific process

Abstract

In scientific process, hypothesis generation is one the most important steps where creativity is needed most. As the science becomes more open and data-driven, it becomes interesting to analyse whether a crowdsourcing approach might be beneficial in this step. First, we characterize the process as a design process. Then, based on a real-life case study, we analyse and highlight difficulties and challenges for crowd-based hypothesis generation. Last, we give a generic process model for organizing in similar challenges in other data-based scientific hypothesis generation contexts.