The Role of Emotions for the Perceived Usefulness in Online Customer Reviews

Online customer reviews often express emotions. This can enable marketers to analyze the textual content of online reviews with the aim to understand the role of emotions and how they can affect other customers. In this paper, we present an approach to extracting emotion content from online reviews in order to measure the importance of various emotion dimensions within different product categories. The approach uses an emotion lexicon to extract emotion terms, while it also builds a classification model to measure the importance of emotion dimensions based on the quality of reviews. Review quality is measured based on the usefulness of online customer reviews, which are perceived and evaluated by other customers through their helpfulness ratings. This approach allows the identification of emotion dimensions that characterize qualitative reviews. The empirical evaluation in our study suggests that trust, joy, and anticipation are the most decisive emotion dimensions, although substantial variance across product categories can also be detected. Additionally, we compared two contrasting emotion dictionaries. One lexicon was crowd-funded and contained a large vocabulary, whereas the other was more focused and smaller, since it was created word-wise by an expert. Our empirical findings indicate that the crowd-funded solution outperforms its smaller counterpart in terms of classification precision. The main implication of this study is that it adds an emotional perspective to the broad set of existing tools that marketers employ to analyzing online reviews. Our contributions are: i) we are the first to analyze emotions' role in online customer reviews; ii) we demonstrate how to develop a big data model such as this, without external assistance; iii) we show how to interpret the results of the created model; and iv) we show which dictionary to prefer when creating the model.