An Intelligent Meeting Room Reservation System Based on Machine Learning Classification Techniques: A Case Study of Phranakhon Si Ayutthaya Rajabhat University

Abstract

This study aimed to (1) develop an intelligent meeting room reservation system that recommends rooms based on users' needs using machine learning classification techniques, (2) analyze and compare the performance of Decision Tree, Support Vector Machine (SVM), and K-Nearest Neighbors (KNN) models, (3) evaluate the system performance by experts, and (4) examine user acceptance based on the Technology Acceptance Model (TAM). The system was developed as a web application connected to the machine learning models through a RESTful API. The dataset comprised 3,677 actual meeting room reservation records from Phranakhon Si Ayutthaya Rajabhat University between 2014 and 2025. Four input features were engineered, and the target variable was the suitable meeting room for the booking, where the output was the name of the recommended meeting room ranked by suitability from a total of 10 rooms, framed as a multi-class classification task. The data were split into training and testing sets at an 80:20 ratio.

The results showed that the Decision Tree model achieved the best performance, with a test accuracy of 98.40%, weighted-average precision, recall, and F1-score of 0.99, 0.98, and 0.98, respectively, and the lowest gap value of 0.005, indicating model stability. In comparison, KNN and SVM achieved test accuracies of 93.80% and 90.80%, respectively. The system evaluation by five experts using black box testing was at a good level (= 4.39, S.D. = 0.57), and user acceptance among 30 target users was at the highest level (= 4.84, S.D. = 0.37). The findings indicate that the developed system is efficient, user-friendly, and applicable for supporting meeting room management in public sector organizations.