Multi-objective Task Recommendation Method Based On Different Task Characteristics For Crowdsourcing

With the development of Internet technology,crowdsourcing,as an emerging distributed problem-solving model based on human-machine integration,has attracted wide attention in the industry and academia in recent years.On the crowdsourcing platform,there are problems that the task cannot be mind and the task does not match the participating workers,which resulting in inefficient crowdsourcing tasks and the waste of time and economic.Therefore,this paper studies the task recommendation method based on different task characteristics for crowdsourcing to improves the efficiency of crowdsourcing platform task solving.A multi-objective crowdsourcing new task recommendation method based on the hungarian algorithm is proposed to solve the problem that the new task participates can't match tasks.Firstly,the similarity between the tasks is calculated based on the task title and the requirement description text,and the similarity is used to obtain the completed tasks in the worker transaction record similar to the new task;then,selecting evaluation indicators to evaluate the quality,time and cost of the similar completed tasks.Efficiency is evaluated and used as an estimate of workers' efficiency in completing new tasks.Finally,a multi-objective optimization model for crowdsourcing new task recommendation problems is established with the goal of maximizing quality,time and cost efficiency,and with the hungarian algorithm the model is solved and get a new task recommendation.A multi-objective crowdsourcing emergency task recommendation method based on the evolutionary algorithm is proposed to solve the problem that too few workers participating in emergency tasks.First,based on the number of participants in a specified time,the number of participants is pre-judged whether the task belongs to an emergency task;then,the worker transaction record data and the number of task submission plans are collected to evaluate worker activity and task urgency,and based on worker activity and task urgency calculate the emergency efficiency of workers participating in emergency tasks;next,aiming at maximizing emergency efficiency,quality efficiency,time efficiency and cost efficiency to establish a multi-objective optimization model for crowdsourcing emergency mission recommendation problems,and use evolutionary algorithm solve the model to get emergency mission recommendation.Finally,this method is applied to the task recommendation of the Epwk crowdsourcing platform.By comparing with the actual operational data of the crowdsourcing platform,it shows that the method can achieve reasonable recommendation of new tasks and urgent tasks.The research results of this paper further enrich the crowdsourcing task recommendation theory and improve the operational efficiency of the crowdsourcing platform.It has certain practical significance for both theory and applied research.