Aggregating Customer Review Attributes for Online Reputation Generation

Abstract

In this paper, we face the problem of generating reputation for movies, products, hotels, restaurants and services by mining customer reviews expressed in natural language. To the best of our knowledge, previous studies on reputation generation for online entities have primarily examined semantic and sentiment orientation of customer reviews, disregarding other useful information that could be extracted from reviews, such as review helpfulness and review time. Therefore, we propose a new approach that combines review helpfulness, review time, review attached rating and review sentiment orientation for the purpose of generating a single reputation value toward various entities. The contribution of the paper is threefold. First, we design two equations to compute review helpfulness and review time scores, and we fine-tune Bidirectional Encoder Representations from Transformers (BERT) model to predict the review sentiment orientation probability. Second, we design a formula to assign a numerical score to each review. Then, we propose a new formula to compute reputation value toward the target entity (movie, product, hotel, restaurant, service, etc). Finally, we propose a new form to visualize reputation that depicts numerical reputation value, opinion categories, top positive review and top negative review. Experimental results coming from several real-world data sets of miscellaneous domains collected from IMDb, TripAdvisor and Amazon websites show the effectiveness of the proposed method in generating and visualizing reputation compared to three state-of-the-art reputation systems.

Customized reputation generation of entities using sentiment analysis

“…This network could convert into a text-to-text format from any language. Benlahbib and Nfaoui (2020a) used the BERT model for sentiment classification, reputation generation and visualization, taking different attributes into account. In this study, we are inspired by Benlahbib and Nfaoui (2020a) and have used other attributes like review helpfulness for a more accurate reputation generation.…”

“…With the help of reviews, a user can make an informed decision. There is not much research done on the review extracted in the natural language of the e-commerce platforms (Benlahbib and Nfaoui, 2020a). The extracted knowledge from the reviews could help the seller to find the key features which users like or dislike helping the seller in increasing their sales.…”

“…There are mainly two existing techniques in the field of natural language review analysis. Summarization of the reviews is the first one in which the reviews are summarized based on their features (Hu and Liu, 2004; Zhuang et al , 2006) or other entities like the quality, customer experience and many more (Benlahbib and Nfaoui, 2020a). The second one is reputation generation, which is a score is generated based on the natural language customer reviews expressing the opinions (Benlahbib and Habib Nfaoui, 2019; Benlahbib and Nfaoui, 2020b; Yan et al , 2016).…”

“…This consideration of multiple attributes will help in better results and design customized reputation for the products and not just a generalized with no consideration of product demands. The main contributions of this paper are as follows: We propose a novel approach of combining two machine learning and one deep learning model (taking advantage of transfer learning) for the sentiment analysis; we have taken review helpfulness in the account being inspired by helping in a more accurate reputation score generation for book reviews (Benlahbib and Nfaoui, 2020a). We have used three classifier models BERT (Bidirectional Encoder Representations from Transformers) (Devlin et al , 2018), SVM (Steinwart and Christmann, 2008) and Naïve Bayes algorithm (Zhang and Li, 2007).…”

MTVRep: A movie and TV show reputation system based on fine-grained sentiment and semantic analysis

“…Over the last two decades, a few opinion summarizer systems Journal homepage: http://ijece.iaescore.com Ì ISSN: 2088-8708 have been proposed to produce a summary for product reviews [9], movie reviews [3], hotel reviews [1] and local service reviews [10]. Backing to reputation generation task, to the best of our knowledge, there are very few reputation systems that have been proposed to compute a single reputation value toward different entities based on fusing and mining user and customer reviews expressed in natural language [11][12][13][14][15]. Yan et al [11] applied opinion mining and fusion techniques on product reviews.…”