Enhanced Sentiment Analysis Using Hybrid IFAE-IADM Model: A Comparative Study of Machine Learning Algorithms

Abstract

Vehicular Ad Hoc Networks (VANETs) facilitate smooth communication between vehicles and infrastructure, which is essential for intelligent transportation systems. However, the existence of anomalies like cyberattacks, rogue nodes, and sensor failures makes guaranteeing the security and dependability of VANETs extremely difficult. In order to solve this problem, this research presents the Integrated Anomaly Detection Model (IFAE-IADM), a hybrid framework that uses a majority voting mechanism to effectively detect anomalies in real time by utilizing Isolation Forest and Autoencoder models. While the Autoencoder model detects irregularities based on reconstruction errors, the Isolation Forest model effectively isolates anomalous data points by partitioning the dataset. A synthetic dataset that mimics vehicle characteristics, such as movement patterns, speed, communication behavior, and data traffic, is used to train and assess the suggested framework. Standard metrics like accuracy, precision, recall, and F1-score are used to evaluate performance. Furthermore, the hybrid model's efficacy is contrasted with more conventional anomaly detection techniques like Support Vector Machine (SVM) and K-Means Clustering. According to experimental results, the IFAE-IADM model performs noticeably better than individual models and other approaches, providing improved detection accuracy and robustness in the complex and dynamic VANET environment. In order to strengthen VANET security and advance the creation of safer and more dependable vehicular networks, this research demonstrates the potential of hybrid machine learning-based anomaly detection.