Reputation-based task allocation for improving realibility in crowdsourcing

Crowdsourcing has been increasing popular nowadays. The act of outsourcing job or task to the group of unknown individuals without any commitment like usual benefit given to the employee has gained acceptance to the not only corporate world but also to a world leader. Website that can be access anywhere and anytime is one of the attraction or key factor for them to use this new wave of employment. The crowdsourcing system such as Amazon Mechanical Turk (MTurk) which has millions of registered user base offer requester to post a task that can be solved by a group or an individual of workers with a minimum fee. In return, the reward will be given to the successful worker determine by the requester or evaluator. In a borderless world, chances of malicious or adversaries attempt is high. Moreover, the requester does not have any profile of workers who work for them. Any slight of bad result will jeopardize the outcome of their company’s operation and the reliability of crowdsourcing system. In realizing that, we are proposing to introduce a reputation management to be embedded into crowdsourcing before appointing tasks to the worker. The workers will be scrutinized their trustworthiness before assign the task to them. Hence, our objective is to design the trustiness scheme for crowd worker and develop a reputation-based resource allocation using trust factor The experiment has been conducted and the result shows that the reliable worker always guarantee that all task can be executed within task duration time given.

Preview Text FSKTM 2017 11 - IR.pdf Download (1MB) | Preview

Actions (login required)

View Item