Tuan Bui
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Tuan Bui has a total of 35 co-authors. Here are the top co-authors by number of shared publications:
- Thang Bui
Ho Chi Minh City University of Technology (HCMUT), Vietnam
8 shared publications - Tho Quan
Ho Chi Minh City University of Technology (HCMUT)
8 shared publications - Phat Thai
Vietnam National University - Ho Chi Minh City
5 shared publications - Minh Hua
Ho Chi Minh City University of Technology (HCMUT), Vietnam National University - Ho Chi Minh City
4 shared publications - Ngan Pham
Ho Chi Minh City University of Technology (HCMUT), Vietnam National University - Ho Chi Minh City
4 shared publications - Trong Le
Vietnam National University - Ho Chi Minh City
4 shared publications
Latest Publications
- Open Access
- 2 November 2026
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Intelligent Systems and Technology
Symbolic-JEPA: Where Symbolic AI Meets Joint-Embedding Predictive Architecture for NL–FOL Conversion
Bridging natural language (NL) and formal logic, such as first-order logic (FOL), remains a fundamental challenge for developing trustworthy and explainable artificial intelligence (AI) systems. Despite recent progress, current NL–FOL systems still lack unified semantic–logical representation learning; most adopt late-fusion architectures that loosely couple large language models (LLMs) with symbolic solvers. Furthermore, extended reasoning chains often introduce inefficiencies and lead to violations in logical consistency, structural validity, and formal constraints. In this paper, inspired by Neuro-Symbolic Type-4&5 architectures and the Joint Embedding Predictive Architecture (JEPA), we propose Symbolic-JEPA , a neuro-symbolic predictive representation model that unifies semantic and logical learning within a single framework. The model incorporates symbolic priors into JEPA, ensuring that the generated FOL expressions remain both semantically faithful and logically consistent. Symbolic-JEPA comprises four key components: a Symbolic Graphizer that unifies NL and FOL inputs into a shared graph space; a Structure-Aware Node & Path Encoder (SANE) that fuses linguistic and structural cues; a Structure-Guided Attention Transformer (SGAT) that injects structure-biased attention; and a Predictor with Micro-Attention that performs clause- and alignment-aware inference efficiently. To support paragraph-level conversion and evaluation, we further introduce ParaLogic , a large-scale NL–FOL dataset with inter-sentence logical annotations (9,344 samples, 28K expressions), along with a three-dimensional evaluation framework spanning parse-, conversion-, and reasoning-level metrics. Experimental results show that Symbolic-JEPA achieves consistently higher logical accuracy, structural well-formedness, and generalization capability compared to neural and LLM-based baselines.
- 2 November 2026
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Intelligent Systems and Technology
Open AccessSymbolic-JEPA: Where Symbolic AI Meets Joint-Embedding Predictive Architecture for NL–FOL Conversion
Bridging natural language (NL) and formal logic, such as first-order logic (FOL), remains a fundamental challenge for developing trustworthy and explainable artificial intelligence (AI) systems. Despite recent progress, current NL–FOL systems still lack unified semantic–logical representation learning; most adopt late-fusion architectures that loosely couple large language models (LLMs) with symbolic solvers. Furthermore, extended reasoning chains often introduce inefficiencies and lead to violations in logical consistency, structural validity, and formal constraints. In this paper, inspired by Neuro-Symbolic Type-4&5 architectures and the Joint Embedding Predictive Architecture (JEPA), we propose Symbolic-JEPA , a neuro-symbolic predictive representation model that unifies semantic and logical learning within a single framework. The model incorporates symbolic priors into JEPA, ensuring that the generated FOL expressions remain both semantically faithful and logically consistent. Symbolic-JEPA comprises four key components: a Symbolic Graphizer that unifies NL and FOL inputs into a shared graph space; a Structure-Aware Node & Path Encoder (SANE) that fuses linguistic and structural cues; a Structure-Guided Attention Transformer (SGAT) that injects structure-biased attention; and a Predictor with Micro-Attention that performs clause- and alignment-aware inference efficiently. To support paragraph-level conversion and evaluation, we further introduce ParaLogic , a large-scale NL–FOL dataset with inter-sentence logical annotations (9,344 samples, 28K expressions), along with a three-dimensional evaluation framework spanning parse-, conversion-, and reasoning-level metrics. Experimental results show that Symbolic-JEPA achieves consistently higher logical accuracy, structural well-formedness, and generalization capability compared to neural and LLM-based baselines.
- 5 April 2026
- Published by Springer Nature
- p. 30-44
NeuroPath: Stepwise Knowledge Graph Reasoning with Neural Intelligence
- 5 April 2026
- Published by Springer Nature
- p. 30-44
NeuroPath: Stepwise Knowledge Graph Reasoning with Neural Intelligence
- 9 November 2025
- Published by Springer Nature
- p. 200-214
Text-JEPA: A Joint Embedding Predictive Architecture for the Conversion of Natural Language into First-Order Logic
- 9 November 2025
- Published by Springer Nature
- p. 200-214
Text-JEPA: A Joint Embedding Predictive Architecture for the Conversion of Natural Language into First-Order Logic
- Open Access
- 26 October 2025
- Published by Association for Computing Machinery (ACM)
- p. 19-27
Formal Reasoning for Intelligent QA Systems: A Case Study in the Educational Domain
- 26 October 2025
- Published by Association for Computing Machinery (ACM)
- p. 19-27
Open AccessFormal Reasoning for Intelligent QA Systems: A Case Study in the Educational Domain
- Open Access
Bridging LLMs and Symbolic Reasoning in Educational QA Systems: Insights from the XAI Challenge at IJCNN 2025
The growing integration of Artificial Intelligence (AI) into education has intensified the need for transparency and interpretability. While hackathons have long served as agile environments for rapid AI prototyping, few have directly addressed eXplainable AI (XAI) in real-world educational contexts. This paper presents a comprehensive analysis of the XAI Challenge 2025, a hackathon-style competition jointly organized by Ho Chi Minh City University of Technology (HCMUT) and the International Workshop on Trustworthiness and Reliability in Neurosymbolic AI (TRNS-AI), held as part of the International Joint Conference on Neural Networks (IJCNN 2025). The challenge tasked participants with building Question-Answering (QA) systems capable of answering student queries about university policies while generating clear, logic-based natural language explanations. To promote transparency and trustworthiness, solutions were required to use lightweight Large Language Models (LLMs) or hybrid LLM-symbolic systems. A high-quality dataset was provided, constructed via logic-based templates with Z3 validation and refined through expert student review to ensure alignment with real-world academic scenarios. We describe the challenge's motivation, structure, dataset construction, and evaluation protocol. Situating the competition within the broader evolution of AI hackathons, we argue that it represents a novel effort to bridge LLMs and symbolic reasoning in service of explainability. Our findings offer actionable insights for future XAI-centered educational systems and competitive research initiatives.
Open AccessBridging LLMs and Symbolic Reasoning in Educational QA Systems: Insights from the XAI Challenge at IJCNN 2025
The growing integration of Artificial Intelligence (AI) into education has intensified the need for transparency and interpretability. While hackathons have long served as agile environments for rapid AI prototyping, few have directly addressed eXplainable AI (XAI) in real-world educational contexts. This paper presents a comprehensive analysis of the XAI Challenge 2025, a hackathon-style competition jointly organized by Ho Chi Minh City University of Technology (HCMUT) and the International Workshop on Trustworthiness and Reliability in Neurosymbolic AI (TRNS-AI), held as part of the International Joint Conference on Neural Networks (IJCNN 2025). The challenge tasked participants with building Question-Answering (QA) systems capable of answering student queries about university policies while generating clear, logic-based natural language explanations. To promote transparency and trustworthiness, solutions were required to use lightweight Large Language Models (LLMs) or hybrid LLM-symbolic systems. A high-quality dataset was provided, constructed via logic-based templates with Z3 validation and refined through expert student review to ensure alignment with real-world academic scenarios. We describe the challenge's motivation, structure, dataset construction, and evaluation protocol. Situating the competition within the broader evolution of AI hackathons, we argue that it represents a novel effort to bridge LLMs and symbolic reasoning in service of explainability. Our findings offer actionable insights for future XAI-centered educational systems and competitive research initiatives.
- Open Access
Speaking in Words, Thinking in Logic: A Dual-Process Framework in QA Systems
Recent advances in large language models (LLMs) have significantly enhanced question-answering (QA) capabilities, particularly in open-domain contexts. However, in closed-domain scenarios such as education, healthcare, and law, users demand not only accurate answers but also transparent reasoning and explainable decision-making processes. While neural-symbolic (NeSy) frameworks have emerged as a promising solution, leveraging LLMs for natural language understanding and symbolic systems for formal reasoning, existing approaches often rely on large-scale models and exhibit inefficiencies in translating natural language into formal logic representations. To address these limitations, we introduce Text-JEPA (Text-based Joint-Embedding Predictive Architecture), a lightweight yet effective framework for converting natural language into first-order logic (NL2FOL). Drawing inspiration from dual-system cognitive theory, Text-JEPA emulates System 1 by efficiently generating logic representations, while the Z3 solver operates as System 2, enabling robust logical inference. To rigorously evaluate the NL2FOL-to-reasoning pipeline, we propose a comprehensive evaluation framework comprising three custom metrics: conversion score, reasoning score, and Spearman rho score, which collectively capture the quality of logical translation and its downstream impact on reasoning accuracy. Empirical results on domain-specific datasets demonstrate that Text-JEPA achieves competitive performance with significantly lower computational overhead compared to larger LLM-based systems. Our findings highlight the potential of structured, interpretable reasoning frameworks for building efficient and explainable QA systems in specialized domains.
Open AccessSpeaking in Words, Thinking in Logic: A Dual-Process Framework in QA Systems
Recent advances in large language models (LLMs) have significantly enhanced question-answering (QA) capabilities, particularly in open-domain contexts. However, in closed-domain scenarios such as education, healthcare, and law, users demand not only accurate answers but also transparent reasoning and explainable decision-making processes. While neural-symbolic (NeSy) frameworks have emerged as a promising solution, leveraging LLMs for natural language understanding and symbolic systems for formal reasoning, existing approaches often rely on large-scale models and exhibit inefficiencies in translating natural language into formal logic representations. To address these limitations, we introduce Text-JEPA (Text-based Joint-Embedding Predictive Architecture), a lightweight yet effective framework for converting natural language into first-order logic (NL2FOL). Drawing inspiration from dual-system cognitive theory, Text-JEPA emulates System 1 by efficiently generating logic representations, while the Z3 solver operates as System 2, enabling robust logical inference. To rigorously evaluate the NL2FOL-to-reasoning pipeline, we propose a comprehensive evaluation framework comprising three custom metrics: conversion score, reasoning score, and Spearman rho score, which collectively capture the quality of logical translation and its downstream impact on reasoning accuracy. Empirical results on domain-specific datasets demonstrate that Text-JEPA achieves competitive performance with significantly lower computational overhead compared to larger LLM-based systems. Our findings highlight the potential of structured, interpretable reasoning frameworks for building efficient and explainable QA systems in specialized domains.
- Open Access
- 10 June 2024
- Published by Association for Computing Machinery (ACM)
- p. 36-43
Cross-Data Knowledge Graph Construction for LLM-enabled Educational Question-Answering System: A Case Study at HCMUT
In today's rapidly evolving landscape of Artificial Intelligence, large language models (LLMs) have emerged as a vibrant research topic. LLMs find applications in various fields and contribute significantly. Despite their powerful language capabilities, similar to pre-trained language models (PLMs), LLMs still face challenges in remembering events, incorporating new information, and addressing domain-specific issues or hallucinations. To overcome these limitations, researchers have proposed Retrieval-Augmented Generation (RAG) techniques, some others have proposed the integration of LLMs with Knowledge Graphs (KGs) to provide factual context, thereby improving performance and delivering more accurate feedback to user queries. Education plays a crucial role in human development and progress. With the technology transformation, traditional education is being replaced by digital or blended education. Therefore, educational data in the digital environment is increasing day by day. Data in higher education institutions are diverse, comprising various sources such as unstructured/structured text, relational databases, web/app-based API access, etc. Constructing a Knowledge Graph from these cross-data sources is not a simple task. This article proposes a method for automatically constructing a Knowledge Graph from multiple data sources and discusses some initial applications (experimental trials) of KG in conjunction with LLMs for question-answering tasks.
- 10 June 2024
- Published by Association for Computing Machinery (ACM)
- p. 36-43
Open AccessCross-Data Knowledge Graph Construction for LLM-enabled Educational Question-Answering System: A Case Study at HCMUT
In today's rapidly evolving landscape of Artificial Intelligence, large language models (LLMs) have emerged as a vibrant research topic. LLMs find applications in various fields and contribute significantly. Despite their powerful language capabilities, similar to pre-trained language models (PLMs), LLMs still face challenges in remembering events, incorporating new information, and addressing domain-specific issues or hallucinations. To overcome these limitations, researchers have proposed Retrieval-Augmented Generation (RAG) techniques, some others have proposed the integration of LLMs with Knowledge Graphs (KGs) to provide factual context, thereby improving performance and delivering more accurate feedback to user queries. Education plays a crucial role in human development and progress. With the technology transformation, traditional education is being replaced by digital or blended education. Therefore, educational data in the digital environment is increasing day by day. Data in higher education institutions are diverse, comprising various sources such as unstructured/structured text, relational databases, web/app-based API access, etc. Constructing a Knowledge Graph from these cross-data sources is not a simple task. This article proposes a method for automatically constructing a Knowledge Graph from multiple data sources and discusses some initial applications (experimental trials) of KG in conjunction with LLMs for question-answering tasks.
- Open Access
- 1 January 2023
- Published by Viet Nam National University Ho Chi Minh City in Science & Technology Development Journal - Engineering and Technology
- Vol. 6 (3) , 1924-1936
Towards formal verification of smart grids: An effective modelling approach and some first experimentation
Nowadays, smart grids are used widely around the world when they enable detecting, reacting and pro-acting to changes in usage and many concerns with the power system, as well as having self-healing capabilities. Some smart grids have recently been created and operational in Vietnam. To ensure the effectiveness of the smart grids, the correctness of the system designs must be studied carefully before the explosion of the use of smart grids, particularly in developing countries such as Vietnam. As formal methods, including formal verification and model checking, recently play more important role in verifying smart grid properties such as load balancing and fault resilience, the effectiveness of formal verification depends mostly on system modeling and verification techniques. Recent researches have shown the feasibility of applying model-checking tools in smart grid verification, and also the inability to test complex properties and systems. In our opinion, the inability could be come from the complicating of the models of the system-under-test. In this study, we suggested a new method for representing smart grids using Colored Petri Net (CPN), a formal representa- tion language. In comparing to the current modeling approach, the new model allows engineers transforms complex grids into simple models. Moreover, based on the advantages of the “color” aspect of the CPN, the result model can be easily upgraded to adapt to the changes of the original smart grids without re- modeling. Also in this study, a basic case study for representing a smart grid, which consists of multiple power sources and multiple consumers, will be shown. The model will be configured to capture some problems that may happen in the grid such as the changing of the capacity of the power sources. The verification experimentation conducted on the case study shows the usefulness of the proposed method.
- 1 January 2023
- Published by Viet Nam National University Ho Chi Minh City in Science & Technology Development Journal - Engineering and Technology
- Vol. 6 (3) , 1924-1936
Open AccessTowards formal verification of smart grids: An effective modelling approach and some first experimentation
Nowadays, smart grids are used widely around the world when they enable detecting, reacting and pro-acting to changes in usage and many concerns with the power system, as well as having self-healing capabilities. Some smart grids have recently been created and operational in Vietnam. To ensure the effectiveness of the smart grids, the correctness of the system designs must be studied carefully before the explosion of the use of smart grids, particularly in developing countries such as Vietnam. As formal methods, including formal verification and model checking, recently play more important role in verifying smart grid properties such as load balancing and fault resilience, the effectiveness of formal verification depends mostly on system modeling and verification techniques. Recent researches have shown the feasibility of applying model-checking tools in smart grid verification, and also the inability to test complex properties and systems. In our opinion, the inability could be come from the complicating of the models of the system-under-test. In this study, we suggested a new method for representing smart grids using Colored Petri Net (CPN), a formal representa- tion language. In comparing to the current modeling approach, the new model allows engineers transforms complex grids into simple models. Moreover, based on the advantages of the “color” aspect of the CPN, the result model can be easily upgraded to adapt to the changes of the original smart grids without re- modeling. Also in this study, a basic case study for representing a smart grid, which consists of multiple power sources and multiple consumers, will be shown. The model will be configured to capture some problems that may happen in the grid such as the changing of the capacity of the power sources. The verification experimentation conducted on the case study shows the usefulness of the proposed method.
- 23 August 2022
- Published by Springer Nature
- p. 310-319
Context Graph Alignment Using Adversarial Learning for Air Pollution Detection on IoT Sensor Systems
- 23 August 2022
- Published by Springer Nature
- p. 310-319
Context Graph Alignment Using Adversarial Learning for Air Pollution Detection on IoT Sensor Systems