arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1701.06153physics.soc-phcs.SIq-bio.PE
收藏

规则网络中基于声誉的合作演化

The Evolution of Reputation-Based Cooperation in Regular Networks

Tatsuya Sasaki, Hitoshi Yamamoto, Isamu Okada, Satoshi Uchida

摘要:

尽管声誉技术最近取得了进展,但声誉系统如何影响社交网络中的人类合作尚不清楚。虽然它是已知的,在合作的演变中的两个主要机制是空间选择和基于声誉的互惠,这两种机制之间的相互作用的理论研究仍然几乎是未知的。在这里,我们提出了一个新的基于个人的声誉和网络之间的互惠合作的演变模型。我们比较分析了四个主要的道德评价规则-回避,形象评分,严厉的判断,和简单的立场-和基础上的合作者,叛逃者,和歧视者在定期网络的给予游戏模型。歧视者依赖于适当的道德评估规则。通过使用基于个人的模型,我们表明,这四个评估规则是不同的特点,在合作如何演变,这取决于效益成本比,网络节点的程度,以及观察和错误的条件。研究结果表明,最宽容的规则-简单立场-是四个评估规则中最强大的,在促进定期网络中的合作。

英文摘要:

Despite recent advances in reputation technologies, it is not clear how reputation systems can affect human cooperation in social networks. Although it is known that two of the major mechanisms in the evolution of cooperation are spatial selection and reputation-based reciprocity, theoretical study of the interplay between both mechanisms remains almost uncharted. Here, we present a new individual-based model for the evolution of reciprocal cooperation between reputation and networks. We comparatively analyze four of the leading moral assessment rules---shunning, image scoring, stern judging, and simple standing---and base the model on the giving game in regular networks for Cooperators, Defectors, and Discriminators. Discriminators rely on a proper moral assessment rule. By using individual-based models, we show that the four assessment rules are differently characterized in terms of how cooperation evolves, depending on the benefit-to-cost ratio, the network-node degree, and the observation and error conditions. Our findings show that the most tolerant rule---simple standing---is the most robust among the four assessment rules in promoting cooperation in regular networks.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑