Software Crowdsourcing Allocation Algorithm Based on Publisher Heterogeneous Service Expectations

Aiming at the problem that reliable crowdsourcing workers in existing research are unstable in heterogeneous tasks, a software crowdsourcing assignment algorithm based on publishers’ heterogeneous service expectations (PHSE) is proposed. First, the efficiency and reputation scores of workers are calculated based on the explicit feedback of their historical behaviors. Then, candidate crowdsourcing workers are selected and their skills are scored based on the complexity and skill requirements of the tasks to be assigned. Finally, the optimal matching of crowdsourcing workers and tasks is achieved based on service expectations. Experimental results show that compared with the artificial bee colony algorithm based on skill matching, the ant colony algorithm based on efficiency indicators, the worker selection algorithm based on reputation indicators, the task assignment algorithm based on user reliability, and the task assignment algorithm based on grouping and reorganization, the task assignment success rate and task completion quality of the proposed algorithm are improved by an average of 1.98% and 32.27% respectively, which can effectively guide task assignment.