艾滋病目标人群复杂网络多属性的表达与统计分析

Multi-attributes of HIV/AIDS target population complex network and the statistical analysis

把属性作为一个独立的领域,而不是依赖于实体及其所在的网络进行研究,具有一定的探索性。以"对象、实体、属性、网络关系"作为基本范畴,构建多属性复合网络的研究范式。用集合工具对属性进行表达,引入属性的正负性,并基于传播路径叠加和属性层叠加,结果表明,它们具有较好的统计价值,对艾滋病的传播具有很好的解释功能。利用数学的方法重点挖掘了属性在网络传播过程中的三大发展规律:一是探索了属性的理想增长模式;二是属性类型的数量分布客观上服从正态分布规律,揭示了拥有同一属性类型人群的规模大小状态;三是属性在实体上的分布则服从于二项分布规律,揭示了同一属性在人群中扩散的规模大小和对应的概率水平。可以预见,未来把属性研究领域与现有的网络研究领域相结合,将是复杂网络研究的新方向。

We take the attribute study as an independent research domain instead of depending on its entity and network like before. Taking ＂obieet, entity, attribute, and network＂ as the basic research categories, we can con- struct the research paradigm of the multi-attribute complex network. We use aggregate as the attributes research tool, giving attribute the positive or negative nature, which explains the network of HIV/AIDS more effectively. We use mathematics tools, and find the three major laws of attributes： the ideal growth model of the attributes; the number of attribute types to obey normal distribution law, revealing the size of groups with the same attribute types; and the distribution of attributes on the entity in accordance with binomial distribution, revealing the entity size of one attribute and its probability of occurrence. It can be predicted that the attribute study will be a new direc- tion of complex network research.