Crowdsourcing e-markets have been widely used to complete various complex tasks with the help of seamlessly integrating the ubiquitous intelligent systems and artificial intelligence. Most of traditional crowdsourcing e-markets focus primarily on individual incentives to motivate workers to participate. However, workers generally have heterogeneous values, which brings forth an undesirable outcome that workers with larger values monopolize a fixed total payment. This monopolization discourages the workers with smaller values from participating, thereby severely reducing the quality of individual-oriented crowdsourcing e-markets. Thus, this article proposes a team formation (TF) games-based mechanism in which workers are incentivized to form teams and obtain payments according to their contributions. Moreover, the existence of Nash equilibrium for a given TF game is rigorously analyzed. Finally, the experimental results in terms of both simulated and realistic datasets demonstrate that, as compared with state-of-the-art methods, our mechanism can achieve higher crowdsourcing e-markets quality.
Reputation detection in traditional distributed systems (e.g., electronic commerce systems and peer‐to‐peer systems) relies heavily on the factor of interaction reputation, which can be derived from direct interactions between agents through bidirectional relationships. However, in the current information diffusion in social network systems (SNS) (e.g., Twitter), the characteristic of the unidirectional relationship between agents and the decay property of diffusion will result in lacking direct interactions; therefore, interaction reputations will be difficult to be obtained by agents in a distributed manner. To solve this problem, a novel distributed reputation detection model following the pattern “from path to individual” (FPTI) is proposed, which can provide a new reputation factor as an alternative to interaction reputation in such environments. The main idea is that the positive (or negative) observation of an information diffusion process increases (or decreases) the belief of the corresponding diffusion path, which further increases (or decreases) the reputation of each involved agent. Thus, the reputation of a target agent can be assessed by the superimposition of reputations of multiple paths on which this agent has participated in past information diffusion processes. Furthermore, being aware of agent’s limited capacity for reputation detection in SNS, we then propose theenhanced FPTI model(eFPTI), which simplifies the detection source to reduce detection costs and achieve the approximate performance as FPTI. Theoretical analyses and experimental evaluations validate the efficiency and effectiveness of our models and also show several properties of the models, for example, the robustness for dynamic environments.
Principal–assistant agent teams are often employed to solve tasks in multiagent collaboration systems. Assistant agents attached to the principal agents are more flexible for task execution and can assist them to complete tasks with complex constraints. However, how to employ principal–assistant agent teams to execute time-critical tasks considering the dependency between agents and the constraints among tasks is still a challenge so far. In this paper, we investigate the principal–assistant collaboration problem with deadlines, which is to allocate tasks to suitable principal–assistant teams and construct routes satisfying the temporal constraints. Two cases are considered in this paper, including single principal–assistant teams and multiple principal–assistant teams. The former is formally formulated in an arc-based integer linear programming model. We develop a hybrid combination algorithm for adapting larger scales, the idea of which is to find an optimal combination of partial routes generated by heuristic methods. The latter is defined in a path-based integer linear programming model, and a branch-and-price-based (BP-based) algorithm is proposed that introduces the number of assistant-accessible tasks surrounding a task to guide the route construction. Experimental results validate that the hybrid combination algorithm and the BP-based algorithm are superior to the benchmarks in terms of the number of served tasks and the running time.