A MACHINE-LEARNING-BASED RECOMMENDER SYSTEM FOR HOTEL REVIEWS - Extended Abstract

Sep 17, 2026

Uploaded by Noah

With the rapid growth of online review platforms, selecting suitable accommodations has become increasingly challenging due to the overwhelming volume of user-generated content. This study presents a machine learning-based hotel recommendation system that processes Google Reviews for hotels in Perak, Malaysia. The system recommends hotels using user preferences such as star ratings and sentiment features extracted from review text. A dataset of 2,993 reviews from 58 hotels was collected, pre-processed, and evaluated across Naïve Bayes, Random Forest, and SVM, with SVM achieving the best accuracy at 74.46%.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages9

Reading Time23 min

AccessFree to download

Abstract

1. Introduction

2. Methodology

E-BOOK OF EXTENDED ABSTRACT THE 14th INTERNATIONAL INVENTION, INNOVATION & DESIGN COMPETITION 2025 Organized by: Office of Research, Industry, Community & Alumni Network UiTM Perak Branch © Unit Penerbitan UiTM Perak, 2025 All rights reserved. No part of this publication may be reproduced, copied, stored in any retrieval system or transmitted in any form or by any means; electronic, mechanical, photocopying, recording or otherwise; without permission on writing from the director of Unit Penerbitan UiTM Perak, Universiti Teknologi MARA, Perak Branch, 32610 Seri Iskandar Perak, Malaysia. Perpustakaan Negara Malaysia Cataloguing in Publication Data No e-ISBN: 978-967-2776-52-9 Cover Design: Dr. Mohd Khairulnizam Ramlie Typesetting : Georgia EDITORIAL BOARD Editor-in-Chief MUHD SYAHIR ABDUL RANI Managing Editors NUR FATIMA WAHIDA MOHD NASIR SYAZA KAMARUDIN NORASYIKIN ABDUL MALIK Copy Editors SHEEMA LIZA IDRIS AZURAWATI ZAIDI HALIMATUN SAADIAHABD MUTALIB HALIMATUSSAADIAH IKSANIZA FARADIBA MOHD PATEL MOHAMAD SAFWAT ASHAHRI MOHD SALIM MUHAMMAD WAJIHUDDIN JOHARI NAZIRUL MUBIN MOHD NOOR NORAZIAHAZIZAN NOOR AILEEN IBRAHIM NOOR FAZZRIENEE JZ NUN RAMLAN NOORLINDA ALANG NURAMIRA ANUARNURDIYANA MOHAMAD YUSOF NURSHAHIRAH AZMAN NURUL FARHANI CHE GHANI NURUL MUNIRAH AZAMRIONG ELLY PAUL GNANASELVAMSITI SYAIRAH FAKHRUDDIN WAN FARIDATULAKMAWAN MOHD RASHDIWAN NURUL FATIHAHWAN ISMAIL ZARLINA MOHD ZAMARIAMIRUL FARHAN AHMAD TARMIZI IMRAN TORIQ A MACHINE-LEARNING-BASED RECOMMENDER SYSTEM FOR HOTEL REVIEWS Nurul Hafizah Zainal Abidin1, Wong Wai On2, Norliana Muslim2, Rohani Bakar2, Yusrafizal Yusof3, Nur Idalisa4 1Faculty of Computer and Mathematical Sciences, Universiti Teknologi MARA, Perak Branch, Tapah Campus, 35400 Tapah Road, Perak, Malaysia 2Department of Computer and Communication Technology, Faculty of Information and Communication Technology, Universiti Tunku Abdul Rahman, 31900 Kampar, Perak, Malaysia 3Unit of Mathematics, Perak Matriculation College, 31600 Gopeng, Perak, Malaysia 4Faculty of Computer and Mathematical Sciences, Universiti Teknologi MARA, Terengganu Branch, Dungun Campus, 23000 Dungun, Terengganu, Malaysia [nurul354@uitm.edu.my](nurul354@uitm.edu.my) ABSTRACT With the rapid growth of online review platforms, selecting suitable accommodations has become increasingly challenging due to the overwhelming volume of user-generated content. This study presents a machine learning-based hotel recommendation system that processes Google Reviews for hotels in Perak, Malaysia. The system aims to enhance decision-making by recommending hotels based on user preferences, such as star ratings and sentiment features extracted from textual reviews. A total of 2,993 reviews from 58 hotels were collected using a Python-based scraper and pre-processed data to remove noise and tokenize content. Three machine learning models, namely Naïve Bayes, Random Forest, and Support Vector Machine, were evaluated using standard performance metrics. The SVM model achieved the highest accuracy (74.46%) and was selected for integration into a web-based application developed using Flask. The final system allows users to filter hotel recommendations by name, rating, and custom keywords, providing personalised suggestions and valuable insights. This project demonstrates the potential of machine learning to improve user experiences and decisionmaking in the hospitality industry. Keywords: data preprocessing, Google reviews, Naïve Bayes, random forest, support vector machine, machine learning, recommender system 1. INTRODUCTION The surge in popularity of online review platforms has revolutionised how customers evaluate goods and services, particularly in the hospitality industry. As Malaysia gains momentum as a top tourist destination, hotel reviews on platforms like Google have become central to decision-making, both for travellers and hotel operators. These reviews encapsulate customer

FAQs

What problem does this hotel recommendation system address?

It addresses the difficulty of choosing suitable accommodations when online review platforms contain a large volume of user-generated text. The system automates analysis to support better decision-making for travellers and hotel operators.

How was the Google review data collected and prepared?

Reviews for 58 hotels in Ipoh, Perak were collected using a Python-based scraper, producing 2,993 reviews. The text was pre-processed by removing stop words, cleaning, and tokenising, then converted into numerical features using CountVectorizer.

Which machine learning model performed best and how was it used?

Support Vector Machine (SVM) achieved the highest accuracy at 74.46% and was selected for integration into a web-based application developed with Flask. Users can filter recommendations by hotel name, rating, and custom keywords.

Sep 17, 2026

Uploaded by Noah