With the changes of the times, marketing methods are changing rapidly with the advancement of technology. Due to the rise of e-commerce and the prevalence of social media, information dissemination is very fast, and the huge amount of unstructured data generated by consumers have high research value, especially the unstructured evaluation of consumers after shopping. Therefore, this study uses google.com for text mining and semantic networking and then uses content analysis to reconstruct a large number of reviews about boutique consumption. The results of the analysis found that according to the stems collected in this study, the evaluation analysis can summarize five major orientations, namely aesthetic orientation, value orientation, quality orientation, communication orientation, and time orientation. The evaluations of various types of consumers are obviously different, and the differences are quite large.