Zhanxi Xie
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Zhanxi Xie has a total of 12 co-authors. Here are the top co-authors by number of shared publications:
- Baili Lu
College of Animal Science and Technology, Zhongkai University of Agriculture and Engineering, Guangzhou 510225, China
1 shared publication - Edmund Loi
Macao Polytechnic University
1 shared publication - Frankie U. Kei Wong
Faculty of Humanities and Social Sciences, Macao Polytechnic University, Macau 999078, China
1 shared publication - Guanqiuyue Chen
Faculty of Humanities and Social Sciences, Macao Polytechnic University, China
1 shared publication - Junhao Wei
Macao Polytechnic University, China
1 shared publication - Ngai Cheong
Faculty of Applied Sciences, Macao Polytechnic University, Macao, China
1 shared publication
Latest Publications
- Open Access
- 10 September 2026
- Published by MDPI AG in Education Sciences
- Vol. 16 (9) , 1480
A Comparative Study of Generative AI Use Among Secondary and University Students in Macao: A Technology Acceptance Model Perspective
Generative AI (GenAI) is rapidly altering learning environments, yet comparative research spanning secondary and higher education remains scarce. Grounded in the Technology Acceptance Model (TAM), this cross-sectional study surveyed Macao secondary (n = 487) and university (n = 401) students in 2026 using convenience sampling; only respondents who had previously used GenAI were retained for analysis, and cross-group adoption mechanisms were evaluated with multi-group structural equation modeling. Results indicate high structural consistency across cohorts: the chain from perceived ease of use through perceived usefulness to behavioral intention holds robustly, with perceived usefulness acting as the primary mediator, and perceived popularity positively predicting behavioral intention in both groups. After correcting for multiple comparisons, the only robust cross-group difference among the 10 structural paths is the direct effect of perceived ease of use on behavioral intention, which is present among secondary students (partial mediation) but absent among university students (full mediation, primarily via perceived usefulness). Measurement-invariance testing further indicated that some perceived-ease-of-use and perceived-popularity items were not fully equivalent across learning stages; accordingly, and given the observational design, the estimated paths are interpreted as associations rather than confirmed causal effects. These findings clarify stage-specific GenAI adoption and yield actionable implications for differentiated educational guidance and digital literacy cultivation.
- 10 September 2026
- Published by MDPI AG in Education Sciences
- Vol. 16 (9) , 1480
Open AccessA Comparative Study of Generative AI Use Among Secondary and University Students in Macao: A Technology Acceptance Model Perspective
Generative AI (GenAI) is rapidly altering learning environments, yet comparative research spanning secondary and higher education remains scarce. Grounded in the Technology Acceptance Model (TAM), this cross-sectional study surveyed Macao secondary (n = 487) and university (n = 401) students in 2026 using convenience sampling; only respondents who had previously used GenAI were retained for analysis, and cross-group adoption mechanisms were evaluated with multi-group structural equation modeling. Results indicate high structural consistency across cohorts: the chain from perceived ease of use through perceived usefulness to behavioral intention holds robustly, with perceived usefulness acting as the primary mediator, and perceived popularity positively predicting behavioral intention in both groups. After correcting for multiple comparisons, the only robust cross-group difference among the 10 structural paths is the direct effect of perceived ease of use on behavioral intention, which is present among secondary students (partial mediation) but absent among university students (full mediation, primarily via perceived usefulness). Measurement-invariance testing further indicated that some perceived-ease-of-use and perceived-popularity items were not fully equivalent across learning stages; accordingly, and given the observational design, the estimated paths are interpreted as associations rather than confirmed causal effects. These findings clarify stage-specific GenAI adoption and yield actionable implications for differentiated educational guidance and digital literacy cultivation.
- Top Cited PapersOpen Access
MRBMO: An Enhanced Red-Billed Blue Magpie Optimization Algorithm for Solving Numerical Optimization Challenges
To address the limitations of the Red-billed Blue Magpie Optimization algorithm (RBMO), such as its tendency to get trapped in local optima and its slow convergence rate, an enhanced version called MRBMO was proposed. MRBMO was improved by integrating Good Nodes Set Initialization, an Enhanced Search-for-food Strategy, a newly designed Siege-style Attacking-prey Strategy, and Lens-Imaging Opposition-Based Learning (LIOBL). The experimental results showed that MRBMO demonstrated strong competitiveness on the CEC2005 benchmark. Among a series of advanced metaheuristic algorithms, MRBMO exhibited significant advantages in terms of convergence speed and solution accuracy. On benchmark functions with 30, 50, and 100 dimensions, the average Friedman values of MRBMO were 1.6029, 1.6601, and 1.8775, respectively, significantly outperforming other algorithms. The overall effectiveness of MRBMO on benchmark functions with 30, 50, and 100 dimensions was 95.65%, which confirmed the effectiveness of MRBMO in handling problems of different dimensions. This paper designed two types of simulation experiments to test the practicability of MRBMO. First, MRBMO was used along with other heuristic algorithms to solve four engineering design optimization problems, aiming to verify the applicability of MRBMO in engineering design optimization. Then, to overcome the shortcomings of metaheuristic algorithms in antenna S-parameter optimization problems—such as time-consuming verification processes, cumbersome operations, and complex modes—this paper adopted a test suite specifically designed for antenna S-parameter optimization, with the goal of efficiently validating the effectiveness of metaheuristic algorithms in this domain. The results demonstrated that MRBMO had significant advantages in both engineering design optimization and antenna S-parameter optimization.
Top Cited PapersOpen AccessMRBMO: An Enhanced Red-Billed Blue Magpie Optimization Algorithm for Solving Numerical Optimization Challenges
To address the limitations of the Red-billed Blue Magpie Optimization algorithm (RBMO), such as its tendency to get trapped in local optima and its slow convergence rate, an enhanced version called MRBMO was proposed. MRBMO was improved by integrating Good Nodes Set Initialization, an Enhanced Search-for-food Strategy, a newly designed Siege-style Attacking-prey Strategy, and Lens-Imaging Opposition-Based Learning (LIOBL). The experimental results showed that MRBMO demonstrated strong competitiveness on the CEC2005 benchmark. Among a series of advanced metaheuristic algorithms, MRBMO exhibited significant advantages in terms of convergence speed and solution accuracy. On benchmark functions with 30, 50, and 100 dimensions, the average Friedman values of MRBMO were 1.6029, 1.6601, and 1.8775, respectively, significantly outperforming other algorithms. The overall effectiveness of MRBMO on benchmark functions with 30, 50, and 100 dimensions was 95.65%, which confirmed the effectiveness of MRBMO in handling problems of different dimensions. This paper designed two types of simulation experiments to test the practicability of MRBMO. First, MRBMO was used along with other heuristic algorithms to solve four engineering design optimization problems, aiming to verify the applicability of MRBMO in engineering design optimization. Then, to overcome the shortcomings of metaheuristic algorithms in antenna S-parameter optimization problems—such as time-consuming verification processes, cumbersome operations, and complex modes—this paper adopted a test suite specifically designed for antenna S-parameter optimization, with the goal of efficiently validating the effectiveness of metaheuristic algorithms in this domain. The results demonstrated that MRBMO had significant advantages in both engineering design optimization and antenna S-parameter optimization.