Selected work

Representative Papers

Opinion Mining and Analysis Using Hybrid Deep Neural Networks

Apr 28, 2025Technologies

Existing lexicon-based and traditional machine learning approaches for sentiment analysis suffer from limited contextual modeling capability, poor generalization, and inadequate handling of class imbalance—particularly for negative sentiment. To address these challenges, this paper proposes a hybrid deep neural network architecture integrating Bidirectional Gated Recurrent Units (BGRU) and Long Short-Term Memory (LSTM) networks. The model jointly captures long-range dependencies and bidirectional semantic context, thereby significantly enhancing negative sentiment detection. Evaluations on the IMDB and Amazon review datasets demonstrate a test accuracy of 95%, a negative sentiment recall rate of 96%, and a reduction in misclassification loss from 20.24% to 13.3%. Compared to state-of-the-art methods, the proposed model achieves superior performance in accuracy, robustness, and class-balanced prediction, making it especially suitable for fine-grained opinion mining in social media and e-commerce applications.

2 citationsRead paper

Spam and Sentiment Detection in Arabic Tweets Using MARBERT Model

Jun 24, 2026

This study addresses the challenge of simultaneously detecting sentiment and spam in Arabic tweets from customers of Saudi Telecom Company (STC), where fine-grained sentiment analysis is complicated by linguistic nuances and noisy user-generated content. For the first time, the authors apply the Transformer-based MARBERT pre-trained language model to a five-class sentiment classification task—encompassing positive, negative, neutral, sarcastic, and uncertain sentiments—within an end-to-end joint detection framework for sentiment and spam. Evaluated on a dataset of 24,513 tweets, the proposed approach significantly outperforms existing methods, achieving state-of-the-art results across multiple metrics including F1 score, precision, and recall. This work bridges a critical gap in Arabic natural language processing by demonstrating effective deployment in a real-world customer service context.

0 citationsRead paper

Explainable Ensemble-Based Machine Learning Models for Detecting the Presence of Cirrhosis in Hepatitis C Patients

Jun 24, 2026

This study addresses the challenge of early detection of cirrhosis in patients with hepatitis C by proposing a highly accurate and interpretable automated diagnostic approach based on ensemble learning. Leveraging 28 clinical features from 2,038 Egyptian patients, the work systematically evaluates several ensemble models—including Random Forest, Gradient Boosting Machine, XGBoost, and Extremely Randomized Trees—combined with feature selection and interpretability techniques. The optimal Extremely Randomized Trees model achieves exceptional performance using only 16 key features, yielding 96.92% accuracy, 94.00% recall, 99.81% precision, and an AUC of 96%, while maintaining strong clinical interpretability.

0 citationsRead paper

Absher: A Benchmark for Evaluating Large Language Models Understanding of Saudi Dialects

Jul 14, 2025

This study addresses the limited understanding of Saudi Arabic dialects by large language models (LLMs). We introduce Absher, the first systematic benchmark for evaluating LLMs on Saudi Arabic, comprising over 18,000 human-verified multiple-choice questions across six task categories—including semantic understanding, cloze completion, binary judgment, and contextual application—and pioneering evaluation dimensions such as cultural interpretation and regional identification. We evaluate both multilingual and Arabic-specific LLMs, revealing substantial deficiencies in dialectal semantic comprehension, cultural reasoning, and contextual inference. Absher fills a critical gap in Arabic dialect NLP evaluation and underscores the necessity of dialect-aware pretraining and culturally aligned assessment. It provides a reproducible, fine-grained evaluation framework and empirical evidence to support robust, real-world deployment of Arabic NLP systems.

0 citationsRead paper

A Comparative Review of Parallel Exact, Heuristic, Metaheuristic, and Hybrid Optimization Techniques for the Traveling Salesman Problem

May 23, 2025

This paper systematically evaluates parallelization paradigms for large-scale Traveling Salesman Problems (TSP)—a canonical NP-hard combinatorial optimization problem—with respect to scalability and practical deployability. It benchmarks parallel exact methods (e.g., branch-and-bound), classical heuristics, metaheuristics (genetic algorithms, ant colony optimization, simulated annealing), and emerging machine learning–enhanced approaches (including reinforcement learning and quantum-inspired methods). The study introduces a task-specific evaluation framework for hybrid and adaptive solvers, moving beyond conventional single-metric (accuracy/time) assessments. It rigorously characterizes trade-offs among solution quality, computational efficiency, problem-scale scalability, and robustness across paradigms. Key research gaps are identified, and a reproducible benchmarking infrastructure—designed for heterogeneous computing and distributed architectures—is proposed. This work provides methodological foundations for intelligent logistics and real-time route planning.

0 citationsRead paper

Recent publications

Latest Papers

Spam and Sentiment Detection in Arabic Tweets Using MARBERT Model

Jun 24, 2026

This study addresses the challenge of simultaneously detecting sentiment and spam in Arabic tweets from customers of Saudi Telecom Company (STC), where fine-grained sentiment analysis is complicated by linguistic nuances and noisy user-generated content. For the first time, the authors apply the Transformer-based MARBERT pre-trained language model to a five-class sentiment classification task—encompassing positive, negative, neutral, sarcastic, and uncertain sentiments—within an end-to-end joint detection framework for sentiment and spam. Evaluated on a dataset of 24,513 tweets, the proposed approach significantly outperforms existing methods, achieving state-of-the-art results across multiple metrics including F1 score, precision, and recall. This work bridges a critical gap in Arabic natural language processing by demonstrating effective deployment in a real-world customer service context.

0 citationsRead paper

Explainable Ensemble-Based Machine Learning Models for Detecting the Presence of Cirrhosis in Hepatitis C Patients

Jun 24, 2026

This study addresses the challenge of early detection of cirrhosis in patients with hepatitis C by proposing a highly accurate and interpretable automated diagnostic approach based on ensemble learning. Leveraging 28 clinical features from 2,038 Egyptian patients, the work systematically evaluates several ensemble models—including Random Forest, Gradient Boosting Machine, XGBoost, and Extremely Randomized Trees—combined with feature selection and interpretability techniques. The optimal Extremely Randomized Trees model achieves exceptional performance using only 16 key features, yielding 96.92% accuracy, 94.00% recall, 99.81% precision, and an AUC of 96%, while maintaining strong clinical interpretability.

0 citationsRead paper

Absher: A Benchmark for Evaluating Large Language Models Understanding of Saudi Dialects

Jul 14, 2025

This study addresses the limited understanding of Saudi Arabic dialects by large language models (LLMs). We introduce Absher, the first systematic benchmark for evaluating LLMs on Saudi Arabic, comprising over 18,000 human-verified multiple-choice questions across six task categories—including semantic understanding, cloze completion, binary judgment, and contextual application—and pioneering evaluation dimensions such as cultural interpretation and regional identification. We evaluate both multilingual and Arabic-specific LLMs, revealing substantial deficiencies in dialectal semantic comprehension, cultural reasoning, and contextual inference. Absher fills a critical gap in Arabic dialect NLP evaluation and underscores the necessity of dialect-aware pretraining and culturally aligned assessment. It provides a reproducible, fine-grained evaluation framework and empirical evidence to support robust, real-world deployment of Arabic NLP systems.

0 citationsRead paper

A Comparative Review of Parallel Exact, Heuristic, Metaheuristic, and Hybrid Optimization Techniques for the Traveling Salesman Problem

May 23, 2025

This paper systematically evaluates parallelization paradigms for large-scale Traveling Salesman Problems (TSP)—a canonical NP-hard combinatorial optimization problem—with respect to scalability and practical deployability. It benchmarks parallel exact methods (e.g., branch-and-bound), classical heuristics, metaheuristics (genetic algorithms, ant colony optimization, simulated annealing), and emerging machine learning–enhanced approaches (including reinforcement learning and quantum-inspired methods). The study introduces a task-specific evaluation framework for hybrid and adaptive solvers, moving beyond conventional single-metric (accuracy/time) assessments. It rigorously characterizes trade-offs among solution quality, computational efficiency, problem-scale scalability, and robustness across paradigms. Key research gaps are identified, and a reproducible benchmarking infrastructure—designed for heterogeneous computing and distributed architectures—is proposed. This work provides methodological foundations for intelligent logistics and real-time route planning.

0 citationsRead paper

Weaponizing Language Models for Cybersecurity Offensive Operations: Automating Vulnerability Assessment Report Validation; A Review Paper

May 07, 2025

This study addresses the high false-positive rate and low manual verification efficiency in vulnerability assessment (VA) reports. We propose and implement, for the first time, an LLM-driven paradigm for automatic VA report validation. Methodologically, we integrate critical literature review, semantic parsing of VA reports, and credibility reasoning to construct an LLM-based verification framework tailored to offensive security contexts. Unlike conventional rule-based or statistical models, our approach enables deep discrimination of technical vulnerability details, contextual consistency, and evidentiary sufficiency. Experimental results demonstrate a 37.2% average reduction in false positives, a 5.8× improvement in verification throughput, and an accuracy of 92.4%, substantially reducing human effort. This work bridges a critical research gap—applying LLMs to trustworthy, evidence-grounded validation in offensive cybersecurity tasks.

0 citationsRead paper