Maximizing user type diversity for task assignment in crowdsourcing

• Published in: Journal of combinatorial optimization Vol. 40; no. 4; pp. 1092 - 1120
• Main Authors: Wang, Ana, Ren, Meirui, Ma, Hailong, Zhang, Lichen, Li, Peng, Guo, Longjiang
• Format: Journal Article
• Language: English
• Published: New York Springer US 01.11.2020; Springer Nature B.V
• Subjects: Combinatorics; Convex and Discrete Geometry; Crowdsourcing; Greedy algorithms; Mathematical Modeling and Industrial Mathematics; Mathematics; Mathematics and Statistics; Operations research; Operations Research/Decision Theory; Optimization; Theory of Computation; Crowdsourcing; Crowdsensing; User type diversity; Task assignment
• ISSN: 1382-6905; 1573-2886
• DOI: 10.1007/s10878-020-00645-6

Crowdsourcing employs numerous users to perform certain tasks, in which task assignment is a challenging issue. Existing researches on task assignment mainly consider spatial–temporal diversity and capacity diversity, but not focus on the type diversity of users, which may lead to low quality of tasks. This paper formalizes a novel task assignment problem in crowdsourcing, where a task needs the cooperation of various types of users, and the quality of a task is highly related to the various types of the recruited users. Therefore, the goal of the problem is to maximize the user type diversity subject to limited task budget. This paper uses three heuristic algorithms to try to resolve this problem, so as to maximize user type diversity. Through extensive evaluation, the proposed algorithm Unit Reward-based Greedy Algorithm by Type obviously improves the user type diversity under different user type distributions.