Macroscopic and microscopic statistical properties observed in blog entries

TL;DR

This paper analyzes the statistical properties of blog entries, revealing that small word frequencies follow a Poisson process and confirming the universality of Zipf's law across different blogger behaviors.

Contribution

It introduces a normalization method to accurately evaluate word frequencies and identifies two distinct blogger behaviors, confirming Zipf's law's universality.

Findings

Small word frequencies follow a Poisson process.

Two types of blogger behaviors are identified.

Zipf's law is universally applicable.

Abstract

We observe the statistical properties of blogs that are expected to reflect social human interaction. Firstly, we introduce a basic normalization preprocess that enables us to evaluate the genuine word frequency in blogs that are independent of external factors such as spam blogs, server-breakdowns, increase in the population of bloggers, and periodic weekly behaviors. After this process, we can confirm that small frequency words clearly follow an independent Poisson process as theoretically expected. Secondly, we focus on each blogger's basic behaviors. It is found that there are two kinds of behaviors of bloggers. Further, Zipf's law on word frequency is confirmed to be universally independent of individual activity types.