A Hybrid Approach for Aspect-based Sentiment Analysis: A Case Study of Hotel Reviews

Abstract

This study presents a method of aspect-based sentiment analysis for customer reviews related to hotels. The considered hotel aspects are staff attentiveness, room cleanliness, value for money and convenience of location. The proposed method consists of two main components. The first component is used to assemble relevant sentences for each hotel aspect into relevant clusters of hotel aspects using BM25. We developed a corpus of keywords called the Keywords of Hotel Aspect (KoHA) Corpus, and the keywords of each aspect were used as queries to assemble relevant sentences of each hotel aspect into relevant clusters. Finally, customer review sentences in each cluster were classified into positive and negative classes using sentiment classifiers. Two algorithms, Support Vector Machines (SVM) with a linear and a RBF kernel, and Convolutional Neural Network (CNN) were applied to develop the sentiment classifier models. The model based on SVM with a linear kernel returned better results than other models with an AUC score of 0.87. Therefore, this model was chosen for the sentiment classification stage. The proposed method was evaluated using recall, precision and F1 with satisfactory results at 0.85, 0.87 and 0.86, respectively. Our proposed method provided an overview of customer feelings based on score, and also provided reasons why customers liked or disliked each aspect of the hotel. The best model from the proposed method was used to compare with a state-of-the-art model. The results show that our method increased recall, precision, and F1 scores by 2.44%, 2.50% and 1.84%, respectively.