Procrastination is a behavioural aspect of a task executer who delays the task submission in spatial crowdsourcing. It degrades the quality, introduces errors, and affects the verification quality of the submitted tasks. A bipartite matching approach was proposed to mitigate procrastination by distributing tasks into slots, but not in a balanced manner (based on cost and number of tasks). There are two other approaches that address procrastination by using a balanced distribution of tasks into slots. In this paper, we have proposed a Consistent Hashing Based approaches with improved balanced distribution than the existing algorithms. Our proposed approaches achieve a better balanced distribution of tasks. Additionally, a currently executing task can be retrieved for real-time monitoring in \(\mathcal {O}(\log K)\) time by our proposed approach. We have shown that our proposed methods outperform the existing algorithm using extensive simulation.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Prevention of Procrastination in Spatial Crowdsourcing: A Consistent Hashing Based Approach

  • Naren Debnath,
  • Sajal Mukhopadhyay,
  • Fatos Xhafa

摘要

Procrastination is a behavioural aspect of a task executer who delays the task submission in spatial crowdsourcing. It degrades the quality, introduces errors, and affects the verification quality of the submitted tasks. A bipartite matching approach was proposed to mitigate procrastination by distributing tasks into slots, but not in a balanced manner (based on cost and number of tasks). There are two other approaches that address procrastination by using a balanced distribution of tasks into slots. In this paper, we have proposed a Consistent Hashing Based approaches with improved balanced distribution than the existing algorithms. Our proposed approaches achieve a better balanced distribution of tasks. Additionally, a currently executing task can be retrieved for real-time monitoring in \(\mathcal {O}(\log K)\) time by our proposed approach. We have shown that our proposed methods outperform the existing algorithm using extensive simulation.