Influence of textual characteristics and reviewer expertise on the perception of usefulness of online reviews: application of natural language processing in Brazilian hotel reviews on TripAdvisor

Abstract

In the digital environment, characterized by pseudonymity and information overload, digital platforms encourage their users to rate the usefulness of online reviews (OR), promoting the filtering of relevant information, reducing cognitive overload, and promoting the veracity of information. In this article, we analyze the influence of detail, readability, sentiment, appeal, and adjectives of OR and the reviewer's expertise on perceived usefulness. For this purpose, a survey was conducted with 124,594 ORs taken from the TripAdvisor platform. The hypotheses were tested using a linear regression model. Our findings reveal that reviewer's expertise, readability, and detail of online reviews positively influence their perceived usefulness, while sentiment has a negative effect. The interaction between detail, emotional appeal, and adjectives reinforces this usefulness, increasing consumer trust. These findings contribute to a better understanding of online consumer behavior, highlighting how textual elements make reviews more trustworthy and useful for decision making.