<p>The purpose of this paper is to illustrate the importance of Arabic sign language (ArSL) recognition in facilitating communication among the Arab community of hard-of-hearing individuals. This task can be used in different domains, such as robotic systems guidance and intelligent communication paradigms. The present paper proposes a real-time sign language recognition using the MediaPipe framework for feature extraction and Logistic Regression (LR) for classification. The model utilizes the MediaPipe technique to extract 3D hand landmarks, facial expressions, and motion data from ArSL gesture recordings. A logistic regression classifier is then trained to learn the spatiotemporal features from the extracted landmarks to classify ArSL words with high accuracy. The proposed model has been evaluated on a different dataset after being trained and evaluated on a custom dataset of seven-word ArSL words. The trial results on a controlled, limited dataset of seven words show that the model operates efficiently with perfect (100%) accuracy, serving as a robust proof-of-concept with minimal computational cost. The system operates at 30 frames per second (FPS), making it ideal for real-time applications in assistive technologies, mobile devices, and embedded systems. The results demonstrate that our method offers a highly efficient and accurate alternative to complex deep learning models, and it is robust against variations in background noise, lighting conditions, and hand posture. The proposed paradigm allows for real-time translation, connectivity with intelligent devices and applications, and practical implications for improving communication within the Arab community of the deaf and hearing impaired.</p>