Mining Customer Reviews to Identify Customer Opinions of Products

Abstract

With the internet and technology development, online shopping has become an essential part of modern life. As a result, more people are purchasing products every day through various websites, and these E-commerce sites are producing an enormous amount of customers feedback in different formats. Moreover, if appropriately utilized, customers feedback can help improve the product, services, or marketing campaigns. Therefore, this project contributes to the area of mining and analysing customers’ feedback. This project proposes a framework using Natural Language Processing (NLP) techniques to find customers preferences through mining Customer Reviews (CR) text. First, implement the LDA model using the Gensim package in Python to extract topics from CR. After that, find the overall sentiment for every review in each topic using the Vader sentiment library in Python. After that, upload a selected portion of the CR to Sketch Engine to extract terms and keywords. Lastly, interpret the results and generate helpful insights for brand managers. The Amazon products reviews data are used in this study, and we picked products from three different categories. The first category includes two sewing machines, while the second category includes two gaming mouses. The final category has groups of coconut oils products divided into sustainable products and unsustainable products. The findings of the proposed framework are promising as we were able to identify the most discussed topics in one product and a group of products and produce an assessment that provides information about the aspects that the customers are most satisfied with and that can be improved. However, the Vader sentiment tool did not achieve the expectation because of the complexity of CR. On the other hand, the framework used is adaptable, as evaluated with the text analyst expert, which provides room for improvements and expanding its usage. Finally, we believe these findings would become suggestions for brand managers for future product development and improvement, leading to increased customer satisfaction which results in more sales and profits.