Online customer reviews provide valuable information for restaurants to understand customer experiences, satisfaction and service issues. The problem is, KBB Steak Terengganu had to read and record the reviews on the internet manually, which makes it hard to monitor the reviews systematically and to look for recurring problems as well as to gain insights from the unstructured reviews data. Therefore, this project aimed to identify the current business process and problems faced by KBB Steak Terengganu in analyzing online customer reviews, implement sentiment analysis on the restaurant's online customer reviews and visualize the sentiment analysis results using a dashboard. The project was implemented following the Cross-Industry Standard Process for Data Mining (CRISP-DM) methodology and the collected customer reviews were 1,017 from Google Reviews and Google Forms. The reviews were prepared and the TF-IDF feature was extracted from the reviews before the models were developed using SVM, NB and K-NN and evaluated by stratified 5-fold, 10-fold and 15-fold cross validation. Finally, SVM, which had the highest accuracy of 88.00%, was chosen from the nine model configurations when 10-fold cross validation technique was used. Results were presented using the KBB Insight Dashboard (KBBID) created with Microsoft Power BI to aid review monitoring, sentiment analysis and management insights. Expert evaluation showed that the dashboard was useful, learnable, easy to use and satisfactory for presenting the insights gained from the customer reviews. In conclusion, the project shows that sentiment analysis using machine learning techniques and visualization of the information using Business Intelligence tools can help in understanding the meaning of the information in unstructured customer reviews, which can be used in systematic feedback monitoring and evidence-based decision-making. Future research could involve the incorporation of larger datasets from various data sources and could also try to use more advanced sentiment classification approaches to enhance the performance and generalizability of the model.
| Item Type: | Student Project |
|---|---|
| Creators: | Creators Email / ID Num. Mohd Azli, Aina Syafeeqa 2023637878 |
| Contributors: | Contribution Name Email / ID Num. Advisor Daud, Nur'aina ainadaud@uitm.edu.my. |
| Subjects: | H Social Sciences > HF Commerce > Consumer behavior. Consumers' preferences. Consumer research. Including consumer profiling H Social Sciences > HF Commerce > Consumer satisfaction H Social Sciences > HF Commerce > Customer services. Customer relations > Consumer complaints. Complaint letters |
| Divisions: | Universiti Teknologi MARA, Terengganu > Kuala Terengganu Campus > Faculty of Computer and Mathematical Sciences |
| Programme: | Bachelor Of Information Systems (Hons.) Business Computing |
| Keywords: | Sentiment analysis, Machine learning, Customer reviews |
| Date: | 2026 |
| URI: | https://ir.uitm.edu.my/id/eprint/149434 |
149434.pdf
