A Multi-stage Framework for Complex Task Decomposition in Knowledge-intensive Crowdsourcing

Abstract

Crowdsourcing is being used by more and more industries, which proves that this model is effective. In the crowdsourcing model, knowledge-intensive crowdsourcing (KI-C) has received widespread attention due to today's knowledge economy. The tasks submitted by consumers in KI-C are complicated and individualized, and it is difficult for one service provider to complete an entire task independently. Therefore, in order to effectively match suitable service providers for the complex tasks submitted by consumers, this paper proposes a multi-stage framework for complex task decomposition in KI-C. The framework contains the decomposition principles, searching similar tasks by word2vec and task packages obtained by genetic algorithm. The practicability and validity of this multi-stage framework are tested by a case study.