针对当前指数膨胀的网络资源使得搜索引擎查询出来的结果经常成千上万、良莠不齐的研究现状,通过充分借鉴社会学、经济学、心理学和分布式人工智能等多门学科的重要内容,提出了基于角色的多Agent协同式强化学习技术在这一领域的创新应用,并适时地为该技术引入了一种新颖的经验交流与共享学习机制。比较实验表明,该技术具有较高的搜索效率和运行准确率。
The paper concerns the present application situation that the network resources inflating frequently make the current search engine index tens of thousands of results,the good and evil intermingled.Referring social science and economics and psychology and distributed artificial intelligence,the paper proposes the innovation application of cooperative reinforcement learning technology based on the role,and introduces the novel mechanism of exchanging experience and sharing study.The compared experimental results show that the technology proposed in the paper has the high search efficiency and the running accuracy.