基于超网络的社会化标注行为

Exploring Behavior of Social Tagging by Hyper Network

构建了以标签为节点、以标注活动为超边的社会化标注行为超网络,定义了超度、超度分布、余平均超度和超度条件概率等刻画社会化标注超网络性质的指标,结合Delicious数据进行了实证分析,实证结果表明:(1)用户会不断地使用新标签来进行资源标注;(2)利用新标签标注资源的同时通常会与大众化标签共同进行标注;(3)用户一般会同时使用多个个性化标签和一个或少量大众化标签对资源进行标注。

The hyper-network for social tagging is built,in which tags are taken as nodes and tagging actions that are taken as hyper-edges.Some measures are defined to depict features of social tagging hyper-network,such as hyper-degree and its distribution,excess average hyper-degree and conditional probability distribution of hyper-degree.In terms of these measures,an empirical study is done based on data from Delicious.The empirical results show:(1)users will constantly add new tags to annotate resources;(2)the added new tags generally are combined with popular tags to annotate resources;(3)individual tags and popular tags are usually put together to annotate resources,however the number of individual tags is normally more than one and the number of common tags is generally only one or in small amount.