Agent Coalition Formation of Chinese Question Answering System Based on Improved Ant Colony Algorithm

In order to enhance the performance of Chinese question answering (QA) system, this paper presented a model for Chinese question answering system based on multi-agent. Optimal coalition was formatted by using improved ant colony algorithm. This paper described the problem, introduced solving steps based on ant colony algorithm (ACA), and improved ACA considering the characteristics of Chinese QA. The capability of agents and users' satisfaction were best shown in the process of optimization. The experiment demonstrated that the improved algorithm could avoid falling into local optimums, accelerate the convergence rate and improve the ability of searching an optimum solution.