In the ever-expanding domain of Internet of Things (IoT) networks, Distributed Denial of Service (DDoS) attacks represent a significant challenge, compromising the reliability of these systems. Traditional centralized detection methods struggle to cope effectively in the widespread and diverse environment of IoT, leading to the exploration of decentralized approaches. This study introduces a Federated Learning-based approach, named Federated Learning for Decentralized DDoS Attack Detection (FL-DAD), which utilizes Convolutional Neural Networks (CNN) to efficiently identify DDoS attacks at the source. Our approach prioritizes data privacy by processing data locally, thereby avoiding the need for central data collection, while enhancing detection efficiency. Evaluated using the comprehensive CICIDS2017 dataset and compared with conventional centralized detection methods, FL-DAD achieves superior performance, illustrating the potential of federated learning to enhance intrusion detection systems in large-scale IoT networks by balancing data security with analytical effectiveness.

In the ever-evolving landscape of smart devices, understanding user sentiments is crucial for refining technology and enhancing user experiences. This research presents a novel aspect-based sentiment analysis dataset in the domain of smart devices. The dataset compiles user reviews from diverse USA-based benchmark websites like Amazon, Target, and Walmart. The dataset contains 2370 reviews and offers a well-balanced sentiment distribution with 842 positive, 800 negative, and 728 neutral reviews. To identify key aspects for sentiment analysis, a consultative approach was employed, engaging both industry professionals and end-users. By employing a consultative approach, key aspects such as 'Clock,' 'Connectivity,' and 'Sound' along with other distinctive aspects are identified in user reviews. Covering 10 different smart devices and encompassing approximately 30 distinct aspects, this dataset not only provides comprehensive insights into user sentiments but also serves as valuable training data for machine learning and deep learning models in the realm of sentiment analysis. This research contributes a foundational resource for future research into the landscape of user experiences with smart devices.

The rapid growth of Industrial IoT (IIoT) has led to large volumes of sensor data, increasing the need for scalable and data-driven maintenance solutions. This paper presents a replay-based streaming framework for predictive maintenance that integrates Apache Kafka for data ingestion, Apache Spark Streaming for distributed processing, and an LSTM-based model for fault prediction. The NASA CMAPSS FD001 dataset is used in a controlled streaming simulation to evaluate the feasibility of processing multivariate degradation data under different workload levels. To reduce data leakage, engine units are split before sliding-window sequence construction. Experimental results show that the LSTM model achieves 90.4% accuracy, 89.2% precision, 87.6% recall, and an F1-score of 88.4% on the test set. Under the controlled replay conditions, the observed post-Spark prediction-stage latency remained below 600 ms across the evaluated workload levels. Since the Kafka and Spark timing components and the original micro-batch interval were not retained separately, these measurements should not be interpreted as end-to-end pipeline latency. The results demonstrate the feasibility of integrating stream-processing technologies with deep learning for predictive maintenance, while full end-to-end latency and horizontal scalability remain subjects for future evaluation.