Strategies to Control Quality in Crowdmining

Abstract

Crowd sourcing is emerging as an effective method for data gathering and analysis, which introduces a new paradigm of data mining process. Crowd workers perform small chunks of larger tasks in return for a reward, which can be monetary, material, psychological and so on. One of the biggest challenges of crowd mining is controlling the quality of the answers collected from the crowd. The diversity of workers’ expertise, questions of wide-ranging degrees of difficulty and the lack of enough after-work evaluation techniques are the reasons behind it. To address the problem, we propose some strategies for improving the quality and diversity of the explanations produced by the crowd workers. We have focused on controlling quality from three different perspectives i.e. before working, during working and after working in crowd mining.