ParsAVC: Appraising Voice of Customers based on Aspect-Based Analysis of Online Reviews Using Unsupervised Techniques

Abstract

With the increasing popularity of online shopping, customer reviews have become a valuable source of information for both businesses and consumers. This research proposes an unsupervised method for aspect-based sentiment analysis of customer reviews in online stores. This can help in enhancing the customer experience when shopping online. The method addresses the challenge of extracting and analyzing product aspects from reviews. It leverages rule-based grammatical structures to identify candidate aspects and then extracts explicit and implicit aspects related to these candidates. The evaluation shows that the proposed method achieves significant improvement in precision, recall, and F1-score compared to previous methods. This research contributes to the field of sentiment analysis by providing a novel method for extracting and analyzing product aspects from customer reviews and proposes an unsupervised method for aspect-based sentiment analysis of customer reviews in online stores. The method involves three steps: First, data preparation: Collecting customer reviews and categorize them by product. Second, aspect extraction: Identifying candidate aspects using rule-based grammatical structures. Third, sentiment analysis: Extraction of explicit and implicit aspects related to the identified candidate aspects. For evaluation using precision, recall, and F1-score shows improvement compared to previous methods. This research contributes to extracting and analyzing textual data from customer reviews to identify product aspects and user sentiment.