Guangdong University Of Finances and Economics

ABSTRACT Amid global economic fluctuations and domestic transformation pressures, household economic decision-making is critical to the rational allocation of family wealth. As a core competency for optimizing such decisions, digital literacy is playing an ever more prominent role in helping families navigate complex economic environments and achieve efficient resource allocation. This study employs five waves of China Family Panel Studies(CFPS) data (2014–2022) to investigate the impact of Digital Literacy(DL) on household current consumption, savings, and financial asset allocation, as well as the underlying mechanisms. Findings reveal that DL boosts household current consumption, lowers household current savings, promotes investment in risky financial products, and increases holdings of such products. These conclusions hold after addressing endogeneity concerns and passing robustness tests. Heterogeneity analyses were performed by grouping samples according to urban/rural areas, household income, education level, and mortgage status. Family social networks, credit accessibility, entrepreneurial preferences, and cognitive abilities serve as mediating variables between digital literacy and household economic decision-making. Based on this, it is proposed that policies should be introduced to enhance digital literacy nationwide, with a focus on rural and disadvantaged groups, to support balanced development in residents’ entrepreneurship and consumption.

Against the backdrop of the deep integration of globalization and the digital economy, cross-border e-commerce has emerged as a core driving force for the growth of global trade. As the core linguistic carrier in cross-border consumption, the mechanism by which English proficiency influences consumers' purchase intention remains insufficiently explored. This paper constructs a chain mediating model of "English Proficiency - Perceived Risk - Self-Efficacy - Cross-Border Purchase Intention" with perceived risk and self-efficacy as mediating variables, to investigate the internal path of how English proficiency affects cross-border purchase intention. A questionnaire survey was conducted to collect 342 valid samples, and SPSS 26.0 and AMOS 24.0 were used to conduct reliability and validity tests, structural equation model analysis, and Bootstrap mediating effect tests. The results show that: English proficiency has a significant direct positive impact on cross-border purchase intention; English proficiency negatively affects perceived risk and positively affects self-efficacy, perceived risk exerts a negative influence on both cross-border purchase intention and self-efficacy, and self-efficacy has a significant positive impact on cross-border purchase intention; perceived risk and self-efficacy respectively play a partial mediating role between English proficiency and cross-border purchase intention, and the two form a significant chain mediating effect, accounting for 14.2% of the total effect. This paper incorporates language proficiency into the research framework of cross-border consumption, improves the theoretical system of consumers' psychological decision-making in cross-border e-commerce, and provides targeted practical suggestions for cross-border e-commerce platforms and merchants to optimize language services, reduce consumers' perceived risk and enhance their self-efficacy.

As the most representative general technology at present, artificial intelligence is profoundly reshaping the organizational form, operating model and operating mechanism of enterprises, and bringing unprecedented impact to the income distribution structure within enterprises. Therefore, based on the panel data of China's Shanghai and Shenzhen A-share non-financial listed companies from 2010 to 2022, this paper explores whether the development of AI will trigger new changes in the interest pattern between corporate profits and labor compensation. Based on basic theories such as capital-labor substitution principle and factor reward theory, this paper explores how AI can promote enterprises to adopt different income distribution modes by improving marginal output of capital and substituting low-skilled labor from the perspective of technology bias. At the same time, the key factor of financing constraint is considered to hinder the enterprise's choice of technology level, which leads to the change of its corresponding distribution effect. Finally, the group regression is carried out from the two perspectives of the ownership structure and the industry to find the different responses of different types of enterprises to the income distribution changes brought by this new technology. This study attempts to outline the basic picture of the evolution of enterprise income distribution mechanism in the era of artificial intelligence, and also provides certain theoretical support and practical evidence for coordinating the relationship between technological innovation process and social distribution justice.

Introduction Rapid urbanization in the Greater Bay Area has precipitated a “spatial decoupling” between historic built environments and local social practices, posing challenges to sustainable land use. This study investigates the re-integration of these dimensions through “Spatial Anchoring.” Methods Drawing on Lefebvre’s production of space theory, we employ a convergent mixed-methods design focused on Shawan Ancient Town. Data were triangulated from 25 stakeholder interviews, policy documents, and a computational sentiment analysis of 1,281 online reviews using a BERT deep learning model. Results Qualitative findings identify two key mechanisms: “Commercial Anchoring” and “Governance Anchoring.” Quantitative results confirm a significant positive interaction effect between tangible and intangible heritage presence on human-environment sentiment (β = 0.098, p < 0.001). Conclusion The study concludes that sustainable urban regeneration relies on the curated re-assembly of spatial relationships (“Assembled Authenticity”) rather than static preservation. It provides a scalable framework for optimizing the social sustainability of historic districts.

The escalating global demand for mental health services highlights the potential of Large Language Models (LLMs) in psychological counseling. However, current LLM-based approaches, particularly fine-tuned models, are constrained by data distribution biases, leading to limited therapeutic diversity and personalization. Crucially, they often lack anticipatory empathetic reasoning, struggle to foresee patient emotional responses beyond immediate dialogue history, and incur substantial computational costs. To address these limitations, we propose PsyPARSE, a novel training-free framework for psychological counseling that emulates the deliberate and empathetic reasoning of human counselors. PsyPARSE integrates Multi-Therapy Retrieval-Augmented Generation (RAG) to overcome data biases and provide highly personalized therapeutic approaches tailored to individual patient attributes. Pioneering the first multi-stage slow-thinking engine in mental health LLMs, PsyPARSE employs Multi-Turn Rollouts to identify optimal therapeutic paths and through anticipating patient reactions, optimizes empathetic responses, thereby ensuring genuinely empathetic and impactful responses in complex, long-dialogue interactions. Operating as a plug-and-play solution, PsyPARSE avoids the computational burden of fine-tuning. We establish a comprehensive LLM-based patient-therapist agent simulation framework for evaluation. Extensive experiments demonstrate that PsyPARSE significantly enhances the capabilities of various LLM baselines, achieving superior personalization and deeper empathy compared to both fine-tuned and other training-free methods. This work offers an efficient, adaptable, and scalable solution to advance mental health support.

This study examines how classroom climate and feedback quality/internalization shape the development of entrepreneurial creativity among undergraduates in China, and whether narrative re-authoring practice amplifies this process. We surveyed students enrolled in entrepreneurship courses across four innovation ecosystems-Shenzhen, Hangzhou, Wuhan, and Xi'an-and paired surveys with expert ratings of artifact novelty and usefulness across course milestones. After distributing 634 questionnaires, 496 valid responses were retained (in-person proctoring; tokenized QR to prevent duplicates). Established, Chinese-adapted scales showed strong measurement quality (loadings ≥0.70; α=0.84-0.92; CR=0.89-0.94; AVE=0.61-0.71; SRMR=0.058; HTMT <0.85). Using SmartPLS 4 with 5000 bootstrap resamples, we found that supportive classroom climate directly promoted entrepreneurial creativity development (β=0.12, p=.009) and strongly enhanced feedback quality/internalization (β=0.58, p<.001). Feedback quality/internalization, in turn, predicted creativity growth (β=0.41, p<.001). The indirect effect of classroom climate on entrepreneurial creativity via feedback was significant and substantial (β_indirect=0.24, p<.001; partial mediation), and narrative re-authoring practice positively moderated the feedback→creativity path (β=0.09, p=.007). Framed by Self-Determination Theory and the Componential Theory of Creativity, the findings identify an actionable mechanism-ignite motivation with climate, steer progress with high-quality feedback, and stabilize iteration with narrative practice-offering theory-driven guidance for scaling China's innovation-oriented entrepreneurship education.