Research on Customer Satisfaction Based on Multidimensional Analysis
- DOI
- 10.2991/ijcis.d.210114.001How to use a DOI?
- Keywords
- Online reviews; Sentiment analysis; Customer satisfaction; Kano model; Multidimensional analysis
- Abstract
Sentiment analysis has been extensively studied recently for developing methodologies to automatically extract information from online reviews, which is important for manufacturers to improve their products or services. Unfortunately, most of current studies don’t take several key factors (e.g., sentiment strength, background of customer) into account. In this study, after building feature, sentiment, and degree vocabulary from online reviews using part-of-speech and word similarity analysis, a review-feature sentiment value (R-FSV) matrix is developed and classified by a model combining the long short-term memory and gated recurrent unit ensemble. The R-FSV matrix and online rating associated with the review are analyzed by a multivariate linear regression model to derive customer satisfaction. Since the sentiment values for each feature of the product are calculated with consideration of sentiment strength, more accurate sentiment orientation is obtained for each review, which is used to develop a set of rules to identify customer requirements based on the Kano model. By taking into account of both product perspective (e.g., product upgrade) and customer perspective (e.g., background of customers), a multidimensional analysis model is proposed to further analyze customer requirements. This sheds light on the dynamic and diversity of customer satisfaction, which help manufacturers to gain insight on not only how the customer satisfaction correlates with product improvements, but also how to develop products for particular group of customers. The proposed method is deployed to study the online reviews of mobile phones from one of the main e-commerce companies in China (i.e., JD.com). The results show that our method is capable to identify the change of customer requirements over time and the preferences of different types of users (i.e., iOS or Android). Hence, the proposed method is more effective in extracting information from the online reviews for manufacturer to improve customer satisfaction efficiently.
- Copyright
- © 2021 The Authors. Published by Atlantis Press B.V.
- Open Access
- This is an open access article distributed under the CC BY-NC 4.0 license (http://creativecommons.org/licenses/by-nc/4.0/).
Download article (PDF)
View full text (HTML)
Cite this article
TY - JOUR AU - Rui Mu AU - Yujie Zheng AU - Kairui Zhang AU - Yufeng Zhang PY - 2021 DA - 2021/01/19 TI - Research on Customer Satisfaction Based on Multidimensional Analysis JO - International Journal of Computational Intelligence Systems SP - 605 EP - 616 VL - 14 IS - 1 SN - 1875-6883 UR - https://doi.org/10.2991/ijcis.d.210114.001 DO - 10.2991/ijcis.d.210114.001 ID - Mu2021 ER -