A Review on Crowdsourcing Models in Different Sectors

Abstract

This paper systematically reviews association rules mining using crowd data in different sectors. Various sectors like health care, education, and tourism require human thinking to be harnessed to answer the queries that can be difficult for computers to answer. Traditional techniques, like Delphi, require experts in the domain which can make the system expensive, and results can be time taking. Crowdsourcing has emerged as a popular approach out of open innovation which can be used for problem solving, ideation, and even in software development because it enables us to make use of experience and background knowledge of crowd using any platform or system or generated forms. It involves a group of experts and non-experts with certain knowledge toward the problem which they solve together, and the solution is shared further. In association rule mining using crowd data, we outsource work to the crowd and mine association rules from their answers using different methods of aggregation. Crowdsourcing systems are beneficial when information required is not in a systematic manner or it is complex to aggregate. These systems can be utilized for many commercial and non-commercial platforms like Amazon Mechanical Turk (AMT) or Kaggle and they can be structured in different forms like crowdfunding or micro-tasking. We first review the current work of mining association rules using crowd data in different sectors. Then we propose suggestions in other sectors and problems which can be solved in those sectors. Finally, we conclude the limitations of the above framework in different sectors that have been covered in this paper.