Research On Methods Of Recommending Crowdsourcing Tasks Incorporating With Interests And Capabilities Of Workers

In this era of rapid development,the existence and development of the Internet is closely related to the development and progress of enterprises.With the development of Internet technology,crowdsourcing mode has attracted wide attention of scholars and become a hot topic of many scholars.In the crowdsourcing mode,how to accurately recommend tasks to workers who are more likely to complete the task has always been a concern of scholars.The precise recommendation of the task directly affects the completion efficiency and quality of the release task.If the task cannot be accurately recommended to workers,it will greatly increase the difficulty of the worker's choice of tasks,waste a lot of time and effort on the receiving side,and reduce the efficiency and quality of the task completion.Therefore,for crowdsourcing task recommendation issues,how to accurately recommend tasks to workers who are more likely to complete the task is a top priority.Crowdsourcing is an important mode of developing swarm intelligence.By publishing and recommending tasks to workers on a crowdsourcing platform,it can improve task completion efficiency and reduce the cost of task completion.However,the lack of information and the limitation of recommendation methods make it difficult to recommend existing tasks to workers accurately.In order to improve the accuracy of task recommendation,this paper proposes a crowdsourcing task recommendation method based on interests and abilities of workers.Firstly,the task recommendation problem is modeled as a constrained single-objective optimization problem with with unknown time to task completion.Then,based on similar tasks and similar workers on a crowdsourcing platform,the time for workers to complete the task to be recommended is estimated.Finally,the model is solved by genetic algorithm,and a set of workers to complete the task is obtained.Further research has found that in the process of finding similar tasks,only considering the task attributes is often not accurate enough.On the crowdsourcing platform,some of the contractors have a bad reputation.They will deliberately extend the commission time,be picky about the workers' submissions,or maliciously terminate the task,causing many workers to spend a lot of time and energy to participate in the task,and ultimately in vain.The crowdsourcing platform recommends tasks to workers,and for those tasks issued by less-reputable contractors,the likelihood of workers choosing to complete the recommended tasks is reduced.Therefore,when looking for similar tasks,the credibility of the task contractor should be taken into account.At the same time,the factors affecting the willingness of workers to participate mainly include the interest of the workers and the trust of the workers in the contractor.It is obviously not comprehensive enough to only consider the interest of the contractor.Therefore,this paper improves the original method and proposes a crowdsourcing task recommendation method that incorporates workers' trust in the contractor and the creditworthiness of the contractor.In the process of finding similar tasks,the method considers the task attributes and the creditor's reputation comprehensively.In the process of finding similar workers,the workers' willingness and ability to participate are considered.The proposed method is applied to the typical domestic crowdsourcing platform-task China,and compared with other methods.The experimental results show that the proposed method can accurately recommend tasks.