International Journal of Advanced Engineering and Management Research Vol. 9, No. 02; 2024 ISSN: 2456-3676 www.ijaemr.com Page 86 SPLIGN: Smart-belt Posture Monitoring System Based on AI-algorithms for Sitting Persons Ferdews Tlili 1 , Rim Haddad 2 , Ridha Bouallegue 1 1 Carthage University, Department of Applied Mathematics, Signals and Communication, Raoued Raoud, 2083 Ariana, Tunisia 2 LAVAL University, Department of Genie Electrique et Genie Informatique, 2325 Rue de l’Universite, Quebec, Canada doi.org/10.51505/ijaemr.2024.9207 URL: http://dx.doi.org/10.51505/ijaemr.2024.9207 Received: Feb 20, 2024 Accepted: Feb 29, 2024 Online Published: Mar 21, 2024 Abstract The back pain is the main common health problems on this decade because of the sitting for a long time tilted on the computer for working, studying or playing. The back problems are widely spread for different ages from young to old people. Last medical research demonstrates that sitting with the posture align can prevent and remedy many spine problems. In this paper, we propose ’SPLIGN’ which a posture monitoring system designed to help maintaining the good posture during sitting. The SPLIGN is a smart belt equipped with inertial sensors. A mobile and web applications are developed for monitoring and remind the user to correct posture. The proposed system is based on a detailed study of the machine learning algorithms in order to choose the best accurate algorithm for posture prediction. The main studied algorithms are Convolutional neural network (CNN), The K-nearest Neighbors (KNN), Support-vector machines (SVM), Decision tree classification, Random forest, Naive Bayes Classifier and Boosting algorithm. The test results demonstrate that The Random Forest algorithm has the best accuracy 99.67% compared to the other algorithms with appropriate processing time 67.7 ms for real time posture monitoring system. Keywords: Machine learning, monitoring system, inertial sensors, mobile application, posture monitoring 1. Introduction The humanity evolution has known a wide involving of the machine in different living fields such as agriculture, medical, learning, industry, etc. At the first time, the machines are designed to execute specific task. After that, the researchers are developed these machines in order to have an intelligent one which can be able to analyze data and aid for decision. The artificial intelligent is started in 1943with the proposal of the neural Network Model [1]. The main objective is to make the machine thinking and taking decision. The machine learning methodology is based on data learning or without data collected and classified data or unclassified ones. There are four