Abstract
Online review platforms are a popular way for users to post reviews by expressing their opinions towards a product or service, and they are valuable for other users and companies to find out the overall opinions of customers. These reviews tend to be accompanied by a rating, where the star rating has become the most common approach for users to give their feedback in a quantitative way, generally as a Likert scale of 1–5 stars. In other social media platforms like Facebook or Twitter, an automated review rating prediction system can be useful to determine the rating that a user would have given to the product or service. Existing work on review rating prediction focuses on specific domains, such as restaurants or hotels. This, however, ignores the fact that some review domains which are less frequently rated, such as dentists, lack sufficient data to build a reliable prediction model. In this paper, we experiment on 12 datasets pertaining to 12 different review domains of varying level of popularity to assess the performance of predictions across different domains. We introduce a model that leverages aspect phrase embeddings extracted from the reviews, which enables the development of both in-domain and cross-domain review rating prediction systems. Our experiments show that both of our review rating prediction systems outperform all other baselines. The cross-domain review rating prediction system is particularly significant for the least popular review domains, where leveraging training data from other domains leads to remarkable improvements in performance. The in-domain review rating prediction system is instead more suitable for popular review domains, provided that a model built from training data pertaining to the target domain is more suitable when this data is abundant.
Highlights
In recent years, the advent and the prevalent popularisation of social media has led to a change in users’ habits of surfing the Internet (Kaplan & Haenlein, 2010; Quan-Haase & Young, 2010; Perrin, 2017; Goss, 2017)
To facilitate and ensure the effectiveness of building a model that will generalise to other domains, we propose an approach for generating aspect phrase embeddings and polarised aspect phrase embeddings, where phrases that vary across domains can be brought to a common semantic space by using word embeddings
Our analysis shows that, while an in-domain review rating prediction system is better for popular domains, for the least popular domains our cross-domain review rating prediction system leads to improved performance when we make use of aspect phrase embeddings
Summary
The advent and the prevalent popularisation of social media has led to a change in users’ habits of surfing the Internet (Kaplan & Haenlein, 2010; Quan-Haase & Young, 2010; Perrin, 2017; Goss, 2017). Since the emergence of social media platforms, Internet users are no longer limited to browsing online contents as mere readers, but they can contribute by expressing and sharing their opinions (Salehan, Zhang & Aghakhani, 2017). Leveraging aspect phrase embeddings for cross-domain review rating prediction. These reviews present the subjective opinions of people on products or businesses, which are invaluable for consumers and companies (Sparks & Browning, 2011). To alleviate the cost of this task, there is a need to explore the development of automated review rating prediction systems
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have