Article Preview
TopIntroduction
User-generated content (UGC) is vital for consumers, retailers, and managers, as customer opinions impact retail business significantly (Giachanou & Crestani, 2016). Online reviews have become essential for consumers when assessing the quality of products such as hotels and restaurants before making purchase decisions (Archak et al., 2011; Zhu & Zhang, 2010). Consumers rely on these reviews to learn from others’ experiences and evaluate product quality (Forman et al., 2008; Kim et al., 2006; Mudambi & Schuff, 2010).
Reviews play a significant role in consumers’ decision-making processes. Consumers rely on the polarity of reviews to assess product or service quality, aiding informed purchasing decisions (Pai et al., 2013). In a survey, 86% of respondents stated that reviews significantly influence their purchase decisions (PowerReviews, n.d.). Review sentiment is considered the second most crucial factor in evaluating consumer reviews (Paget, n.d.). By analyzing online reviews, merchants can understand consumer preferences, improve product quality, and cater to consumer needs (Tang et al., 2014).
However, the abundance of information in reviews can be overwhelming for consumers. To address information overload, three approaches are proposed: visualization, which involves summarizing important information in a graphical format; document summarization, which entails presenting key details; and review ratings (Chang et al., 2017; Lee et al., 2016; Falschlunger et al., 2016; Banerjee & Chua, 2016). Visualization techniques, such as graphic visualization, help reduce information overload and enhance decision-making by presenting complex data in a clear manner (Daniel et al., 2010; Falschlunger et al., 2016). Visualizations capitalize on the picture advantage effect, as people find pictures easier to understand than words or numbers (Paivio & Csapo, 1973; Kelleher & Wagener, 2011).
In addition, consumer preferences can change over time, making it crucial to explore review text and incorporate a time axis for calculating rating scores (Chang et al., 2017). Few studies have examined temporal factors and sentiment analysis, primarily focusing on product characteristics (Li et al., 2015). Charts can efficiently convey product positioning, aiding consumers in their decision-making process and helping companies develop new products (Lee et al., 2016). Timeline-based charts and charts in different orientations provide a quick overview of current hotel experience trends (Chang et al., 2017).
This study aims to develop a website with visual charts showing changes in sentiment over time in reviews. The goal is to determine if this website improves user efficiency compared to the original website, reducing search time and aiding decision-making. In particular, we try to answer the following hypotheses:
- 1.
The accuracy of using the new interface will be higher than the accuracy of using the old interface.
- 2.
The response time using the new interface will be less than the response time using the old interface.
- 3.
The time to find the correct answer and fill in the user answer varies for different facing interfaces.
- 4.
The order in which the interfaces were tested differed in the time it took for users to find the correct answer and user answer.
Historical data from TripAdvisor.com for 10 hotels was selected, and a dictionary of hotel characteristics was constructed. Review sentences were categorized based on these characteristics, and sentiment scores were calculated using the MPQA Corpus subjective dictionary. The study recruited 28 participants to evaluate the new interface’s accuracy, response time, and user experience compared to the original website.