The impact of review sentiment complexity on perceived helpfulness: an information overload perspective

Purpose: This study investigates how the complexity of sentiment in online reviews affects perceived helpfulness. Analyzed over 730,000 reviews from Tripadvisor.com, the research explores how information overload and increased cognitive load impact consumer decision-making. Design/methodology/approach: This study applied the BERT deep learning model to analyze sentiment complexity in online reviews. Based on cognitive load theory, we examined two key factors: the number of attributes mentioned in a review and the variation in sentiment valence of across attributes to evaluate their impact on cognitive load and review helpfulness. Findings: The results show that a higher number of attributes and greater variation in sentiment valence increase cognitive load, reducing review helpfulness. Reviewers’ expertise and review readability further moderate these effects, with complex or expert-written reviews worsening the negative impact. Originality/value: This research introduces a method for measuring attribute-level sentiment complexity and its impact on review helpfulness, emphasizing the importance of balancing detail with readability. These findings provide a foundation for future studies on review characteristics and consumer behavior.