Statistical physics analysis of collective human behaviors observed in the internet information

从统计物理的角度对博客和推特的海量数据进行了分析。首先，我们关注了恒定频率使用的日常词，观察了它们在出现次数波动中的统计特性。接下来，我们将重点放在关注消长和消亡的功能形式的“旺字词”和“新闻字词”上。我们发现，这种增长和衰退要么用指数函数描述，要么用幂函数描述。

From the viewpoint of statistical physics we analyzed the huge data of blogs and twitters. Firstly, we paid attention to "daily words" which are used in constant rates, and observed their statistical properties in the number fluctuation of appearance. Next, we focused on "booming words" and "news words" paying attention to the functional forms of growth and decay. We found that such growth and decay are described either by an exponential function or a power law.