A Time and Opinion Quality-Weighted Model for Aggregating Online Reviews

Beyond mean rating: Probabilistic aggregation of star ratings based on helpfulness

“…In the same study mentioned earlier (McGlohon et al, ), the mean‐rating model performs well among the other selected statistical and heuristics functions. In another study (Shaalan & Zhang, ), the heuristic model perform better than the mean‐rating model.…”

“…However, the model was limited to reviews with at least 10 votes. More recently, Shaalan and Zhang () developed a heuristic aggregation model weighting star rating with a function containing time of post, reviewer and review votes. In their model, the votes for each reviewer are tallied to represent the reputation of reviewers.…”

“…Some existing studies on “helpfulness” examine the utility of “helpfulness” as a social mechanism (Otterbacher et al, ; Otterbacher, ; Ghose & Ipeirotis, ; Zhang & Varadarajan, ) to promote useful reviews and to represent reviewer reputation and expertise (Ghose & Ipeirotis, ; Shaalan & Zhang, ). Other studies examine factors affecting “helpful” votes (Yin, Bond, & Zhang, ; Baek, Ahn, & Choi, ; Cao, Duan, & Gan, ; Pan & Zhang, ; Zhang & Varadarajan, ) or predict the “helpfulness” of reviews (Ghose & Ipeirotis, ; Liu, Huang, An, & Yu, ; Zhang, ; Kim, Pantel, Chklovski, & Pennacchiotti, ).…”

“…It was shown when there exist sufficient review ratings (more than 100), the mean‐rating model is the best model among the statistical models and can tolerate some noise in ratings (McGlohon et al, ; Xie & Lui, ). Heuristic approaches often use extra information in addition to ratings to compute the collective opinion from reviews, such as textual contents, time of post, reviewer IDs and helpful votes (Shaalan & Zhang, ; Zhang et al, ). Aggregation models are often domain‐dependent, and especially the aggregation models using review texts for one domain may not work effectively for another domain.…”

Not All Reviews Are Equal: Towards Addressing Reviewer Biases for Opinion Summarization

“…Reviewer reputation can be modeled by the number of helpfulness votes and total votes received by the reviewer. A higher ratio of helpfulness votes to total votes suggests a better reputation (Shaalan and Zhang, 2016). In another reviewer reputation model, reviewers are modeled to have domain expertise which corresponds to the product categories that the reviewer reviewed on (Chen et al, 2011).…”