The Internet promotes the development of the social media, these new media offer open platforms for participants to share product/service reviews with each other. This study applied the theory of conformity behavior to explain online community members’ information consumption behaviors by using text-mining techniques. NetBeans7.4 was used to conduct Chinese tokenization and data analysis. Next, factor analysis and correlation analysis were conducted to reduce the attribute size of products. Our findings demonstrate that more attributes a product/brand has more discussions found in an online community. The conformity phenomenon is seen in help to accumulate sufficient and complete eWOM to reach a sufficient quantity. Thus, brand is more likely to be mentioned. However, the few brand vendors with high product strength have the impact of conformity, in which, may result in a lower spread power with wrong marketing strategy. Therefore, we argue that the reputation bias generated by conformity will make a misleading purchase decision. Based on the conformity effect of eWOM, we establish the effectiveness of text mining technology applied to information search platform design and brand marketing strategy. Implications were proposed in the final section.
Huang, Echo; Yang, Ya-Hui; and Hong, PeiJyun, "A Study Of Online Beauty Community Members' Voices: EWOM Text Mining" (2015). ICEB 2015 Proceedings. 7.