The rise of artificial intelligence (AI) technologies, particularly generative models for text, image, audio, and video, is radically transforming the creation and dissemination of digital content. These technologies provide significant opportunities, but also pose major challenges for copyright protection in terms of traceability, attribution, license management, and anti-counterfeiting measures. Traditional protection methods, such as digital watermarking, fingerprinting, and Digital Rights Management (DRM) systems, often prove insufficient in the face of AI’s ability to generate or transform works based on existing data. At the same time, information retrieval (IR) and blockchain emerge as promising tools for detecting, verifying, and certifying digital works. This paper analyzes the impact of artificial intelligence on copyright, reviews recent work, describes technical protection methods, examines the role of information retrieval and blockchain in safeguarding these rights, and identifies upcoming technical, legal, and ethical challenges.
Access to clean drinking water remains a critical public health concern worldwide, as chemical and microbial contamination continues to affect millions. Machine learning (ML) provides a promising approach for water quality assessment, enabling rapid and data-driven predictions based on complex physicochemical measurements. This study evaluates the performance of multiple ML algorithms—including logistic regression, support vector machines, decision trees, random forests, gradient boosting, and neural networks—on a standard water quality dataset. Experimental results reveal that ensemble tree-based methods, such as Random Forest and Gradient Boosting, consistently outperform simpler models in accuracy and F1-score, while deep learning models offer limited benefits on small tabular datasets. The study highlights the importance of proper data preprocessing, handling of missing values, and techniques to address class imbalance. These findings underscore the potential of machine learning as a reliable tool for automated potable water detection and suggest directions for improving predictive accuracy and interpretability in real-world water management applications.
The rapid spread of misleading Arabic news stories threatens societal harmony and facts-based choices, yet determining what is truthful remains complicated because of the language’s distinctive linguistic qualities. Current techniques generally focus solely on surface textual patterns, neglecting the combined potential of merging advanced sentiment metrics with hybrid machine learning and deep learning frameworks. We introduce an innovative system for identifying deceptive Arabic news that combines sentiment analysis and several approaches, including both machine learning (RF, Naive Bayes, SVM, LR) as well as deep learning (LSTM, RNN). Experiments on a comprehensive corpus of over 6100 Arabic news show that our sentiment-analysis enhanced hybrid model achieves notably superior performance, with SVM attaining 96
