Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Pengjun Wu has a total of 47 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Public Environmental Experience and Restorative Perception in a Mountain Tourism Village: Field Evidence from Residents and Tourists in Sanpo Town, China

    by Bingnan Lu, Baohui Deng, Ni Lin, Pengjun Wu, Yike Hu

    Public-space experiences in mountain tourism villages involve environmental appraisal, bodily comfort, and place-related affect, but their associations with restorative perception require examination based on user feedback. Using Sanpo Town, China, as a case study, this study asks whether Local Environmental Perception (LEP), Perceived Microclimate Comfort (PMC), and Place Attachment (PA), operationalized as shared place-related affect, are distinguishable constructs; how they are associated with Restorative Perception (RP); and whether the findings are consistent across analytical methods and resident–tourist groups. An on-site non-probability survey conducted in May–June 2025 yielded 591 questionnaires from 300 residents and 291 tourists. Robust maximum-likelihood structural equation modeling was supplemented by ordinal-item analyses, coefficient-product decomposition, composite-score robustness checks, and exploratory group comparisons. LEP, PMC, and PA each retained a positive conditional association with RP, with standardized coefficients of 0.329, 0.259, and 0.202, respectively; the model explained 50.7% of the variance in RP. HTMT and alternative-model comparisons provided qualified support for distinguishing LEP from PMC, although their correlation slightly exceeded the square root of PMC’s average variance extracted. The latent-path comparisons did not detect clear resident–tourist differences under the present data and model specifications; this should not be interpreted as evidence of group equivalence. Familiarity-adjusted composite-score analyses retained positive associations, while a nominal group difference in the LEP–PMC path did not survive correction across six comparisons. The findings support a multidimensional evaluation of the experiences of users encountered in the surveyed public spaces. They do not establish causal design priorities, objective microclimatic performance, population representativeness, or measured recovery relative to a pre-exposure baseline.

  • Open Access

    Residents’ and Tourists’ Visual Perception of Historic Urban Public Spaces: Evidence from Eye-Tracking and Subjective Evaluation in Zhengding Ancient City, China

    by Baohui Deng, Bingnan Lu, Yang Gao, Qing Zhang, Yike Hu, Pengjun Wu

    Historic urban public spaces simultaneously serve as settings for heritage representation and everyday urban life. However, it remains unclear whether residents and tourists allocate visual attention and evaluate the same historic urban public spaces differently. Taking Zhengding Ancient City, China, as a case study, this research integrates static panoramic image-viewing experiments, wearable eye-tracking, and subjective evaluation to investigate differences in visual attention and perceived attractiveness between residents and non-local tourists. Twenty-five representative public-space scenes were selected, and 32 residents and 32 tourists participated in the experiment. Areas of interest (AOIs) were classified into architecture-related AOIs and space-based AOIs. Total fixation duration and fixation count were recorded using Tobii Pro Glasses 3, and subjective evaluations of architecture-related attributes and everyday spatial attributes were collected using a five-point Likert scale. The results showed that residents allocated significantly greater visual attention to space-based AOIs than tourists, whereas tourists reported higher evaluation scores for architecture-related attributes. Conversely, residents assigned higher evaluation scores to everyday spatial attributes. No significant difference was observed in overall satisfaction, suggesting that the two groups differed mainly in the aspects of the historic environment that they evaluated more positively rather than in their overall evaluations. Overall, the findings demonstrate that residents and tourists showed different patterns of visual attention allocation and subjective evaluation when viewing the same static representations of historic urban public spaces. These findings provide empirical evidence for understanding differences in visual attention and subjective evaluation between different user groups and offer methodological support for integrating eye-tracking with subjective evaluation in research on historic urban public spaces.

  • Open Access

    Machine Learning-Driven Kinetic Optimization of Hydroxylamine-Modified Transition Metal Oxide/Peroxymonosulfate System for Antibiotic Degradation

    by Zhixuan Li, Jianwei Li, Ao Zeng, Xi Lian, Wenjun Zhou, Shuyi Xie, Pengjun Wu

    Hydroxylamine-modified transition-metal oxides (HA-TMOs) represent a promising class of catalysts for activating peroxymonosulfate (PMS) to degrade antibiotics. However, identifying energy-efficient operational conditions remains challenging. This study established a comprehensive dataset encompassing 600 experimental records from both in-house experiments and literature and systematically compared 12 machine learning algorithms for predicting the antibiotic degradation efficiency (%) in hydroxylamine-modified transition metal oxide/peroxymonosulfate (HA-TMO/PMS) systems. Among these models, CatBoost delivered the best generalization (test-set R2 = 0.8110, RMSE = 8.92, MAE = 6.15) across repeated 80/20 stratified splits with 5-fold cross-validation, outperforming other ensembles as confirmed by cumulative distribution plots and error-metric analyses. Moreover, the permutation importance analysis identified PMS dosage, HA level, pH, initial pollutant concentration, and catalyst loading as the dominant drivers governing the pollutant removal performance. The partial-dependence plots, incorporating two-variable interactions, elucidated the response surfaces and enabled the discovery of operating windows that jointly maximize degradation efficiency and minimize electrical energy per order (EE/O). ML-guided optimization yielded optimal conditions, which were experimentally verified with sulfamethoxazole (SMZ). The HA-Co3O4/PMS system achieved the highest degradation rate constant (k = 0.11 min−1) and the lowest EE/O value (6.84 kWh·m−3·order−1), markedly improving kinetics and reducing energy consumption compared with non-optimized runs. This work establishes an interpretable ML framework that connects catalyst properties and reaction conditions to degradation kinetics and mechanisms, providing a practical strategy for the screening and scale-up of energy-efficient HA-TMOs/PMS-based advanced oxidation processes (AOPs).

  • Open Access

    System Design and Evaluation of RAG-Enhanced Digital Humans in Design Education: Analyzing Cognitive Load and Instructional Efficiency

    by Xiaofei Zhou, Shiru Zhao, Pengjun Wu, Yan Chen

    Design education involves complex historical knowledge structures that often impose a high extraneous cognitive load on students. This study proposes and evaluates an intelligent instructional system that integrates Retrieval-Augmented Generation (RAG) with anthropomorphic digital humans to function as scalable cognitive scaffolding. We developed a locally deployed architecture utilizing the Qwen3-30B Large Language Model (LLM) for reasoning, BGE-Large-Zh for high-precision semantic embedding, and LiveTalking for real-time audiovisual generation. To validate the system’s pedagogical efficacy, a multi-center randomized controlled trial (RCT) was conducted across three universities (N = 150). The experimental group utilized the RAG-enhanced digital human system, while the control group received traditional instruction. Quantitative results demonstrate that the system significantly improved learning outcomes (p<0.001, Cohen’s d=1.14) and classroom engagement (p<0.001, d=1.39). Crucially, measurements using the Paas Mental Effort Rating Scale revealed a significant reduction in mental effort (p<0.001, d=1.71) for the experimental group. Instructional efficiency analysis (E) confirmed that the system successfully converted reduced extraneous load into germane learning gains (Experimental E=+0.72 vs. Control E=−0.68). These findings validate the technical feasibility and educational value of combining localized RAG architectures with embodied AI, offering a replicable framework for reducing cognitive load in intensive learning environments.


  • Research network and trends in ecological architecture from 2005 to 2024: a visual analysis based on Web of Science Core Database and CiteSpace

    by Yao Liu, Shanshan Wang, Dandan Xu, Pengjun Wu

    Purpose: This study aims to systematically explore the research development, collaborative networks and emerging trends in the field of ecological architecture from 2005 to 2024, offering insights to support theoretical advancement and policy formulation. Design/methodology/approach: Using bibliometric analysis and visualization tools (CiteSpace 6.3.R3), 3,936 English-language articles from the Web of Science Core Collection were analyzed. The study employs co-authorship networks, keyword co-occurrence, cluster detection and burst term analysis to map the structure and evolution of ecological architecture research. Findings: The field has undergone exponential growth (R2 = 0.9743) and thematic shifts from early ecological theories to interdisciplinary applications integrating AI, urban planning, and biodiversity. The USA and China dominate in output, while European countries such as Spain and the UK act as collaborative hubs. Key research hotspots include green buildings, carbon neutrality, AI-assisted design and urban ecological adaptability. Originality/value: This study’s primary contribution is the construction of a comprehensive, 20-year knowledge map that reveals a unique three-dimensional research pattern: micro-mechanism analysis, macro-system optimization, and technology-driven innovation. Unlike previous reviews, our long-term analysis identifies critical evolutionary shifts in the field. It provides strategic direction for global collaboration, sustainable urban development and the digital transformation of ecological architecture.

  • Open Access

    The associational relationships between multidimensional digital literacy and wage income: evidence from urban–rural heterogeneity

    by Jiangwei Hu, Chunyun Tan, Yamei Jiao, Xu Mou, Pengjun Wu

    Amid the rapid development of the digital economy, digital literacy has become a crucial factor associated with labor market outcomes and wage levels, particularly in China, where urban–rural disparities and structural inequalities remain pronounced. Drawing on data from the 2022 China Family Panel Studies (CFPS 2022), this study constructs a multidimensional digital literacy index using the entropy–TOPSIS method, covering five behavioral dimensions: digital tool usage, online social interaction, online learning, entertainment consumption, and e-commerce activities. Stepwise OLS regressions are employed as the main analytical approach to examine how digital literacy is associated with wage income, while a supplementary SEM analysis is conducted to validate the robustness of the observed relationships and to explore potential mediating pathways. The results indicate that digital literacy is positively and significantly associated with wage income after controlling for key covariates, with a stronger association observed among urban workers, highlighting its structural role as a new form of human capital. However, the mediating association of non-farm employment is not statistically significant, suggesting that the linkage between digital skills and employment outcomes may be constrained by skill mismatches, institutional barriers, and unequal access to resources. Overall, this study contributes to a correlational understanding of the relationship between digital literacy and income distribution and provides empirical evidence and policy insights for enhancing digital skills, promoting labor mobility, and narrowing the urban–rural income gap.