Bilal Nakhal

Dr. Bilal Nakhal

orcid

Research Focus & Metrics

Academic Career Length

4Years (2023-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
Verified
Journal Article
Open access
PDF file
MNMoustafa Mohammad Al Najjar
Moustafa Mohammad Al Najjar
MGMohamad Gaber Ghanem
Mohamad Gaber Ghanem
RMRasha Mahboub
Rasha Mahboub
BNBilal Nakhal
Bilal Nakhal
This study investigates the impact of artificial intelligence (AI) on reducing accounting errors from two distinct angles: that of accounting software developers and of certified public accountants. We employ a questionnaire-based approach informed by prior research and validated through pilot testing. Our findings reveal significant benefits for software developers. AI effectively addresses various accounting errors, including tax rate discrepancies, cutoff period inaccuracies, principal violations, concealed transactions, mathematical mistakes, and manipulation errors. However, when considering users, AI’s effectiveness varies. While it successfully mitigates certain errors, such as those related to principles, it falls short in eliminating mathematical errors. This research contributes fresh insights into the role of AI in accounting within emerging markets, enhancing our understanding of its potential and limitations.

5 November 2025 in IoT
Verified
Journal Article
Open access
PDF file
HYHisham Abdul Karim Yassine
Hisham Abdul Karim Yassine
MSMohammed Al Saleh
Mohammed Al Saleh
BNBilal Nakhal
Bilal Nakhal
ACAbdallah El Chakik
Abdallah El Chakik
The exponential growth of Internet of Things (IoT) devices introduces significant security challenges due to their resource constraints and diverse attack surfaces. To address these issues, this paper proposes the Centralized Two-Tiered Tree-Based Intrusion Detection System (C2T-IDS), a lightweight framework designed for efficient and scalable threat detection in IoT networks. The system employs a hybrid edge-centralized architecture, where the first tier, deployed on edge gateways, performs real-time binary classification to detect anomalous traffic using optimized tree-based models. The second tier, hosted on a centralized server, conducts detailed multi-class classification to diagnose specific attack types using advanced ensemble methods. Evaluated on the realistic CIC-IoT-2023 dataset, C2T-IDS achieves a Macro F1-Score of up to 0.94 in detection and 0.80 in diagnosis, outperforming direct multi-class classification by 5–15%. With inference times as low as 6 milliseconds on edge devices, the framework demonstrates a practical balance between accuracy, efficiency, and deployability, offering a robust solution for securing resource-constrained IoT environments.


Verified
Journal Article
Open access
PDF file
MNMoustafa Mohammad Al Najjar
Moustafa Mohammad Al Najjar
RMRasha Mahboub
Rasha Mahboub
BNBilal Nakhal
Bilal Nakhal
MGMohamad Gaber Ghanem
Mohamad Gaber Ghanem
In recent years, artificial intelligence has increasingly been interesting for its role in improving accounting practices. This research investigates whether there is a significant difference in value-added tax (VAT) reporting quality between traditional methods and those assisted by artificial intelligence (AI) in emerging markets. The experiment introduces an AI intervention using ChatGPT-4 to analyze data for accounting errors. The results demonstrate that AI-assisted reporting significantly improves reporting quality, as the AI effectively identified accounting errors that were missed in traditional reporting. This study makes a valuable contribution by providing novel, practical insights into the role and capabilities of AI in tax reporting, employing a rarely used experimental methodology to explore this topic.

16 April 2026 in Journal of Imaging
Verified
Journal Article
Open access
PDF file
NSNada Elie Said
Nada Elie Said
BNBilal Nakhal
Bilal Nakhal
AEAli El-Zaart
Ali El-Zaart
LALama Ahmed Affara
Lama Ahmed Affara
Vanishing point detection is a fundamental step in computer vision that allows 3D scene understanding and autonomous navigation. Classical techniques have significant challenges when trying to understand scenes that are heavily cluttered and images containing multiple perspective cues, leading to poor or unreliable vanishing point determination. We present a Dual RANSAC with Rescue Midpoint-based Multi-Trend Vanishing Point Detection framework, which targets the simultaneous detection and fine-tuning of multiple, globally consistent vanishing points. The proposed framework introduces a novel Midpoint-based Multi-Trend Random Sample Consensus formulation that operates on line segment midpoints to infer dominant directional groups, thereby eliminating noisy or unstable midpoints and stabilizing subsequent vanishing point inference. The main novelty lies in using line segment midpoints to model the orientation variation as a linear regression in the midpoint–orientation space, which helps reduce sensitivity to endpoint instability. Candidate vanishing points are prioritized through inlier-based confidence ranking and subsequently optimized via an MSAC-based arbiter to resolve hypothesis conflicts and minimize geometric error. We evaluate our work against state-of-the-art techniques such as J-Linkage and Conditional Sample Consensus, over two of the current challenging public datasets that comprise the York Urban Dataset and the Toulouse Vanishing Point Dataset. The results show that the proposed framework achieves a recall of up to 95% and an image success rate of almost 84%, outperforming both J-Linkage and Conditional Sample Consensus, especially under tighter angular thresholds. This demonstrates the ability of the proposed framework to provide enhanced stability and localization accuracy.

2 April 2026
Verified
Thesis
Closed access
BNBilal Nakhal
Bilal Nakhal
Génération des intentions communicatives pour agents virtuels dans un environnement virtuel intelligent : application aux environnements d'apprentissage virtuels: La réalité virtuelle joue un rôle majeur dans le développement de nouvelles technologies de l’éducation, et permet de développer des environnements virtuels pour l’apprentissage, dans lesquels, des agents virtuels intelligents jouent le rôle de tuteur. Ces agents sont censés aider les utilisateurs humains à apprendre et appliquer des procédures ayant des objectifs d’apprentissage prédéfini dans différents domaines. Nous travaillons sur la construction d’un système temps-réel capable d’entamer une interaction naturelle avec un utilisateur dans un Environnement d’Apprentissage Virtuel (EAV). Afin d’implémenter ce modèle, nous proposons d’utiliser MASCARET (Multi-Agent System for Collaborative, Adaptive & Realistic Environments for Training) comme modèle d’Environnement Virtuel Intelligent (EVI) afin de représenter la base de connaissances des agents, et de modéliser la sémantique de l’environnement virtuel et des activités des utilisateurs. Afin de formaliser l’intention des agents, nous implémentons un module cognitif dans MASCARET inspiré par l’architecture BDI (Belief-Desire-Intention) qui nous permet de générer des intentions de haut-niveau pour les agents. Dans notre modèle, ces agents sont représentés par des Agents Conversationnels Animés (ACA), qui sont basés sur la plateforme SAIBA (Situation, Agent, Intention, Behavior, Animation). Les agents conversationnels de l’environnement ont des intentions communicatives qui sont transmises à l’utilisateur via des canaux de communication naturels, notamment les actes communicatifs et les comportements verbaux et non-verbaux. Afin d’évaluer notre modèle, nous l’implémentons dans un scénario pédagogique concret pour l’apprentissage des procédures d’analyse de sang dans un laboratoire biomédical. Nous utilisons cette application afin de réaliser une expérimentation et une étude pour valider les propositions de notre modèle. L’hypothèse de notre étude est de supposer que la présence d’un ACA dans un Environnement Virtuel (EV) améliore la performance du processus d’apprentissage (ou qu’au moins, ça ne le dégrade pas) dans le contexte de l’apprentissage d’une procédure spécifique. La performance de l’utilisateur est représentée par le temps requis pour l’exécution de la procédure, le nombre d’erreurs commises et le nombre de demande d’assistance. Nous analysons les résultats de cette évaluation, ce qui confirme partiellement l’hypothèse de l’expérience et affirme que la présence de l’ACA dans l’EV ne dégrade pas la performance de l’apprenant dans le contexte de l’apprentissage d’une procédure.