论文标题
有效的网络导航，其中有部分信息
Efficient network navigation with partial information
论文作者
论文摘要
我们提出了一个信息理论框架，以捕获网络导航模型的过渡和信息成本。基于最小描述长度原则和马尔可夫决策过程，我们证明只有部分信息才能实现有效的全球导航。此外，我们在某些条件下得出了一种最佳解决方案的可伸缩算法。所提出的算法可以解释为网络上的动态过程，使其成为分析和理解现实世界网络的导航策略的有用工具。
We propose a information theoretical framework to capture transition and information costs of network navigation models. Based on the minimum description length principle and the Markov decision process, we demonstrate that efficient global navigation can be achieved with only partial information. Additionally, we derived a scalable algorithm for optimal solutions under certain conditions. The proposed algorithm can be interpreted as a dynamical process on network, making it a useful tool for analysing and understanding navigation strategies on real world networks.
