基于能力层次聚类和角色协同的众包任务分配  

Crowdsourcing task allocation based on capability hierarchical clustering and role collaboration

作　　者：黄黎[1] 赵璐 陈嘉豪 HUANG Li;ZHAO Lu;CHEN Jia-hao(School of Information Technology,Jiangsu Open University,Nanjing 210019,China;School of Computer Science and Technology,Nanjing University of Posts and Telecommunications,Nanjing 210023,China;School of Computer and Information Engineering(School of Artificial Intelligence),Nanjing University of Technology,Nanjing 211816,China)

机构地区：[1]江苏开放大学信息工程学院,江苏南京210019 [2]南京邮电大学计算机学院,江苏南京210023 [3]南京工业大学计算机与信息工程学院(人工智能学院),江苏南京211816

出　　处：《计算机工程与设计》2024年第12期3739-3748,共10页Computer Engineering and Design

基　　金：国家自然科学基金项目(61672022、U190411189);江苏省自然科学基金项目(BK20230351);江苏省高校自然科学基金项目(22KJB110012);江苏省教育科学规划课题基金项目(B/2023/01/50)。

摘　　要：针对现有众包任务分配研究中群组角色与任务之间多对多的协同匹配问题,设计一种基于能力层次聚类和角色协同的众包任务分配方法。考虑群组成员间的协作能力和交互成本度量,定义基于能力的角色层次聚类方法,量化基于能力层次的工人综合胜任力度量。引入群组角色协同理论,将群组角色与任务之间多对多协同匹配问题定义为执行者与任务之间的组合优化问题,定义基于社会网络分析方法的协同候选服务交互成本计算方法,提出一种基于群组角色协同的众包任务分配算法,解决群组绩效和交互成本的双目标约束优化问题,提高任务分配的效率和质量。实验验证该方法在众包任务分配满足服务行为兼容性和功能可靠性等方面具有较明显优势。Aiming at the problem of many to many collaborative matching between group roles and tasks in existing crowdsou-rcing task allocation research,a crowdsourcing task allocation method based on capability hierarchical clustering and role collaboration was designed.Considering the collaboration performance and interaction cost among group members,an interaction cost calculation method was proposed,and the comprehensive competency level of workers based on competency levels was qualified.The group role collaboration theory was introduced,and the many to many collaboration matching problem between group roles and tasks was defined as the combinatorial optimization problem between executors and tasks.The interaction cost calculation method of collaborative candidate services based on social network analysis method was defined,and an approach for assigning tasks in crowdsourcing based on the collaboration of group roles was introduced to solve the double objective constraint optimization problem of group performance and interaction cost,so as to improve the efficiency and quality of task allocation.Experimental verification shows that this method has significant advantages in meeting service behavior compatibility and functional reliability in crowdsourcing task allocation.

关 键 词：服务众包 任务分配 角色协同 群组绩效 层次聚类 协同匹配 交互成本 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]