Thai Son (Jason) Chu

Dr. Thai Son (Jason) Chu

orcid

Research Focus & Metrics

Academic Career Length

3Years (2024-Present)

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

Representative Publications

View All Publications
12 May 2024
Verified
Journal Article
Open access
PDF file
MHMohammed Shahadat Hosen
Mohammed Shahadat Hosen
RIRaisul Islam
Raisul Islam
ZNZain Naeem
Zain Naeem
EFEsther Oluwabusayo Folorunso
Esther Oluwabusayo Folorunso
Thai Son (Jason) ChuThai Son (Jason) Chu
Thai Son (Jason) Chu
NONurudeen Olalekan Orunbon
Nurudeen Olalekan Orunbon
The paper explores the development of decision-making based on data in business intelligence strategies. It grows more and more significant in today's data-driven atmosphere. Business intelligence begins to become more prominent. Business link to existing databases with the help of business intelligence solutions. It can search through immense databases like a sea, searching for nuggets of crucial information.Business intelligence greatly improves modern database systems and boosts operational effectiveness. Artificial intelligence technologies are essential in providing business owners with pertinent information that enables them to outperform competitors. This includes present consumer trend data, which helps firms better understand their target market. It is crucial to decision-making, together with operational effectiveness requirements and market trends.The analytics is vital for assisting organizations in preserving their competitive advantage since it can forecast future market developments. Business intelligence incorporates large databases; it also promotes a culture where data is the primary tool for decision-making. In short, business analytics and intelligence tools assist firms in examining the world of data-driven decision-making. They allow innovation by exposing previously unexplored possibilities. They offer a variety of benefits and alter the way companies make decisions, but there are also obstacles that must be surmounted for deployment to be successful.

29 July 2025 in Knowledge
Verified
Journal Article
Open access
PDF file
Thai Son (Jason) ChuThai Son (Jason) Chu
Thai Son (Jason) Chu
MAMahfuz Ashraf
Mahfuz Ashraf
This paper shows that artificial intelligence is fundamentally transforming college curricula by enabling data-driven personalization, which enhances student outcomes and better aligns educational programs with evolving workforce demands. Specifically, predictive analytics, machine learning algorithms, and natural language processing were applied here, grounded in constructivist learning theory and Human–Computer Interaction principles, to evaluate student performance and identify at-risk students to propose personalized learning pathways. Results indicated that the AI-based curriculum achieved much higher course completion rates (89.72%) as well as retention (91.44%) and dropout rates (4.98%) compared to the traditional model. Sentiment analysis of learner feedback showed a more positive learning experience, while regression and ANOVA analyses proved the impact of AI on enhancing academic performance to be real. Therefore, the learning content delivery for each student was continuously improved based on individual learner characteristics and industry trends by AI-enabled recommender systems and adaptive learning models. Its advantages notwithstanding, the study emphasizes the need to address ethical concerns, ensure data privacy safeguards, and mitigate algorithmic bias before an equitable outcome can be claimed. These findings can inform institutions aspiring to adopt AI-driven models for curriculum innovation to build a more dynamic, responsive, and learner-centered educational ecosystem.

2 May 2025 in Preprints.org
Verified
Preprint
Open access
PDF file
Thai Son (Jason) ChuThai Son (Jason) Chu
Thai Son (Jason) Chu
MAMahfuz Ashraf
Mahfuz Ashraf
This paper presents evidence that Artificial Intelligence has fundamentally reshaped college curricula by promoting a data-driven personalization approach that improves student outcomes and aligns educational programs with workforce needs. Specifically, predictive analytics, machine learning algorithms, and natural language processing were applied here to evaluate student performance and identify at-risk students to propose personalized learning pathways. Results indicated that AI-based curriculum achieved much higher course completion rates (89.72%) as well as retention (91.44%) and dropout rates (4.98%) compared to the traditional model. Sentiment analysis of learner feedback showed a more positive learning experience, while regression and ANOVA analyses proved the impact of AI on enhancing academic performance to be real. Learning content delivery for each student was therefore continuously improved based on individual learner characteristics and industry trends by AI-enabled recommender systems and adaptive learning models. Its advantages notwithstanding, the study emphasizes the need to address ethical concerns, ensure data privacy safeguards, and mitigate algorithmic bias before an equitable outcome can be claimed. These findings can inform institutions aspiring to adopt AI-driven models for curriculum innovation to build a more dynamic, responsive, and learner-centered educational ecosystem.

26 September 2025 in International Journal of Computational and Experimental Science and Engineering
Verified
Journal Article
Open access
PDF file
ABAl Bagiro
Al Bagiro
MAMasjidul Azad
Masjidul Azad
SKSyed Riazul Islam Karim
Syed Riazul Islam Karim
Thai Son (Jason) ChuThai Son (Jason) Chu
Thai Son (Jason) Chu
HKHumera Khan
Humera Khan
SSSamia Hasan Suha
Samia Hasan Suha
The goal of the research is to answer how deep learning techniques affect the accuracy of threat detection, threat response automation and the security of the digital infrastructure in general. These days, security measures can hardly defend against increasingly sophisticated cyber-attacks, including ransomware, phishing and even zero-day vulnerabilities. The rapid digitalization of modern-day industries, the increasing complexities of cyber threats have made the utilization of AI in cybersecurity imperative. As a set of traditional measures is unable to defend against these kinds of attacks, new emerging technologies such as deep learning are set to make an impact. AI and deep learning in particular, is set to make strides in the active, preventive and mitigative measures of cyber-attack threats. These goals achieved by employing a systematic literature review hybridized with deep learning model experimentation concerning the field of cybersecurity. Recent developments in the integration of AI in security measures reviewed, along with numerous neural networks CNNs, RNNs and GANs. It is implementing an experimental design and evaluation with a benchmark dataset pertaining to cybersecurity. The conclusion highlights that deep learning greatly improves threat detection mechanisms through automation in cybersecurity. Compared to conventional security systems, machine learning models offer higher mastery in identifying anomalies and produce fewer false alarms. There are still some obstacles like the complexity of computations, opposing threats and privacy issues. These findings suggest that AI-powered cybersecurity solutions greatly enhance the protection of national critical assets and infrastructures in a rapidly changing cyber environment.

26 September 2024
Verified
Journal Article
Open access
PDF file
Thai Son (Jason) ChuThai Son (Jason) Chu
Thai Son (Jason) Chu
MAMahfuz Ashraf
Mahfuz Ashraf
Distributed Denial of Service (DDoS) attacks could lead to enormous problems, which raises important security issues in many organizations and corporations. By using Bi-directional Long Short-Term Memory Recurrent Neural Network (BLSTM RNN), this paper presents a novel deep learning technique for detecting DDoS attacks within the networks. To achieve highly effective intrusion detection models, we first train our deep learning neural network with multiple layers using a benchmark data set, CICDDOS2019. We examine the binary classification of normal and attack patterns on the web. The result of the experiment shows that our proposed model achieves excellent outcomes in precision, recall, F-1 score, and FPR. Our proposed BLSTM RNN model can detect attacks with over 98% accuracy on practical data sets, better than other Kumar models (31) and even our previous model (30).

13 October 2026 in Engineering Proceedings
Verified
Conference Paper
Open access
PDF file
Thai Son (Jason) ChuThai Son (Jason) Chu
Thai Son (Jason) Chu
NSNitirajsingh Sandu
Nitirajsingh Sandu
BGBhargavi Goswami
Bhargavi Goswami
SKSHAKIR KARIM
SHAKIR KARIM
Artificial Intelligence (AI) systems are increasingly being applied to complex social and environmental challenges, but their implementation across different application domains requires appropriate combinations of computational techniques, data sources, deployment environments, user interfaces, and responsible-AI mechanisms. This study proposes an AI Integration Framework for examining and configuring AI-enabled social innovation systems in support of the Triple-Zero Agenda: zero poverty, zero unemployment, and zero net carbon emissions. The framework is derived from a qualitative multi-case analysis of five social enterprises operating in agriculture, healthcare, education, clean energy, and employment. The analysis examines the AI technologies employed, their development and deployment environments, accessibility mechanisms, data-collection approaches, privacy mechanisms, explainability practices, human-in-the-loop controls, ethical review mechanisms, and measurable impact indicators. The cases demonstrate the use of heterogeneous AI configurations, including predictive machine learning in mobile and cloud environments, natural language processing in application and SMS-based systems, adaptive learning algorithms in web platforms, AI-based energy optimization integrated with IoT and edge devices, and AI-driven skill-job matching through mobile applications. The cross-case analysis indicates that effective AI deployment depends not only on the choice of AI technique but also on the configuration of supporting data, infrastructure, interaction, privacy, explainability, and human-oversight mechanisms. Based on these observations, the study organizes these dimensions into an AI Integration Framework that provides a structured basis for analyzing and configuring AI systems for socially oriented applications. The framework also connects system configurations with operational, social, economic, and environmental outcomes, thereby providing a computer science-oriented perspective on the design, deployment, and scalability of AI-powered social innovation.