Analyzing writing styles of bloggers with different opinions

Abstract

Understanding customers is crucial to companies’ decision-making. With the advent of Web 2.0, more and more people choose to express their feelings and articulate their attitudes through online social communities such as blogs and web forums. These new sources of information offer the potential to obtain large quantities of customer feedback using automated analysis techniques. In this paper, we study how people with different opinions toward a commercial product write differently in their blogs. We define and extract four types of stylometric features – namely lexical, syntactic, structural, and sentimental features – to represent a blogger's writing style. Based on multivariate analyses on a data set of iPod-related blogs, we found various writing style patterns of bloggers. Our analysis shows that a blogger’s writing style is marginally related to his or her opinion toward a product.