Research Article Open Access

Measuring Customers’ Satisfaction Using Sentiment Analysis: Model and Tool

Ahmed Alqurafi1 and Tawfeeq Alsanoosy1
  • 1 Department of Computer Science, College of Computer Science and Engineering, Taibah University, Saudi Arabia


Customer reviews are a valuable data resource for business owners and companies. Customers frequently write reviews on many platforms, such as eBay or Amazon. These reviews show, for instance, to what extent customers are satisfied with a service or a product. Therefore, these reviews can be used by companies and business owners to improve their services or products. However, numerous reviews make it difficult and time-consuming for companies to manually read, analyze, and classify every review. To tackle this issue, we proposed a sentiment analysis model that automatically analyses and classifies customer reviews. To build the model, six popular Machine Learning (ML) classifiers and a Deep Learning (DL) classifier were chosen. The six applied ML classifiers were implemented using three feature extraction techniques: Bag of Words (BoW), Term Frequency-Inverse Document Frequency (TF-IDF), and N-grams. The aim was to determine the most efficient classifiers and feature sets for analyzing customer reviews. To train the model, we used a large, public, and real-world dataset that consisted of 4 million customer reviews. The results of this study confirmed some of the published results and showed some considerable improvements compared to some of the existing sentiment analysis models. Moreover, the findings indicated that applying N-grams revealed better accuracy of almost all ML classifiers. Among the selected ML classifiers, the higher accuracy was achieved at 91.3% when using the Support Vector Machines (SVM) with TFIDF and a combination of Unigram, Bigram, and Trigram. The worst accuracy was 77.3% when applying the Decision Tree (DT). However, Long Short-Term Memory (LSTM) showed the highest accuracy at 93.3%. We also utilized a web-based tool to deploy the sentiment analysis model so it would be freely accessible. Our tool will help companies and business owners analyze their customer reviews automatically and display a set of statistics effortlessly at a low cost, thereby measuring customer satisfaction.

Journal of Computer Science
Volume 20 No. 4, 2024, 419-430


Submitted On: 2 September 2023 Published On: 7 February 2024

How to Cite: Alqurafi, A. & Alsanoosy, T. (2024). Measuring Customers’ Satisfaction Using Sentiment Analysis: Model and Tool. Journal of Computer Science, 20(4), 419-430.

  • 1 Citations



  • Natural Language Processing
  • Sentiment Analysis
  • Learning
  • Customer Reviews
  • Customer Satisfaction