The impact of language consistency on the perceived quality of consumer reviews: A study based on information adoption model and automated annotation of language large models

Abstract

As online reviews play an increasingly important role in consumer decisions, understanding these influencing factors significantly improves review quality and user satisfaction. Based on the information adoption model, this study explores the impact of Language Style Matching (LSM) and Topic Matching on the perceived quality of user reviews. By analyzing 26,852 reviews on the Airbnb platform in California from January to May 2024, the LIWC tool was used to calculate the language style matching score of each review, and the LDA model was used to perform topic analysis to obtain the corresponding topic score. In addition, each review is scored through the GPT-3.5 model from four dimensions: specificity, clarity, emotional color, and usefulness. The research results show that language style matching and topic matching significantly affect the perceived quality of reviews, in which perceived credibility and perceived usefulness mediate in this process. Future research can use multiple language models to score reviews and calculate the scores, reducing the bias caused by a single model.