Agent-based simulation approach to understanding the interaction between employee behavior and dynamic tasks

In this paper, an agent-based simulation approach is applied to explore how employee behavior interacts with dynamic tasks. An assessment model for matching between employees and tasks is presented, and two algorithms to allocate tasks to employees are designed: the minimal matching and the greedy matching algorithm. The algorithms are then translated into multi-agent simulation systems, which are programmed in Java based on Repast J 3.0. The simulation experiment results showed that minimal matching is better than greedy matching for rapid task allocation. The former can reduce interface communication cost and effectively promote the use of employee capability. Moreover, with the minimal matching algorithm, the following effects are evident: (1) the different percentage of generalists and specialists have a distinct effect on completion of tasks; (2) the different preference of manager has a rarer impact on the completion of tasks than on the increase of individual capability; (3) the higher rate of individual capability is positively correlated with collaborative learning rate; and (4) the higher rate of individual capability has a marginal, significant effect when the variance of task capability distribution increases and the expectation remains constant. The increase will be significant when the expectation of the task capability requirement increases. The increase of task number has positive impact on the average rate of increase of capability of employees.