Shuo Xu

Technology-based startups (TSs) increasingly rely on venture capital (VC) to transform emerging technologies into scalable ventures, yet investors often face difficulties in identifying startups whose technological attributes fit their investment capabilities. Existing studies have advanced methods for identifying VC investors, but they often overlook region-specific industrial policy signals, underrepresent technological knowledge structures (e.g., patents and IPC classes), and offer limited decision interpretability. To address these limitations, we propose an AI-driven framework that integrates four modules. First, an LLM-based prompting pipeline extracts computable policy signals, which are then combined with a Regional Industrial Base Index (RIBI) and an Investment Capital Attractiveness Index (ICAI) to identify priority sectors. Second, a heterogeneous network is constructed by linking VCs, TSs, cities, patents, and IPC domains. Third, our proposed TECH-MAGNN model performs meta-path-based representation learning to capture geographic, social, and technological proximity for VC–startup link prediction. Fourth, an LLM explanation layer synthesizes multi-source structural evidence to verify predicted matches and identify boundary conditions, translating complex relational signals into decision-relevant narratives. Using Suzhou’s biopharmaceutical industry as an empirical case, we demonstrate how policy-guided targeting, technology-aware network learning, and explainable AI jointly improve regional startup–investor matching. By integrating industrial policy, technological knowledge, and investment relations into a unified analytical framework, this study shifts the analytical focus from how capital promotes innovation to how capital can be guided to identify and connect with innovation—an actionable perspective for regional policy.

Generative artificial intelligence (GenAI) has become a revolutionary tool in the quickly changing field of education, providing new avenues for improving educational outcomes. This study investigates the impact of generative artificial intelligence (GenAI) adoption on university students’ academic performance, with a focus on collaborative e-learning and effort expectancy. Based on theoretical underpinnings of SCT, a conceptual framework was created. PLS-SEM is utilized to test the conceptual framework empirically. Data was collected through the technique of a structured survey, and the sample size was 260. The target population for this study is students of Chinese higher educational institutions. The empirical findings reveal that GenAI significantly enhances academic performance with the positive mediation of collaborative e-learning. Moreover, effort expectancy demonstrates a positive moderation on the link between GenAI adoption and collaborative e-learning. This research focuses on the importance of lowering technological challenges in order to facilitate the wider acceptance and smooth utilization of GenAI technologies in higher educational settings.

The arrival of generative artificial intelligence (GAI) technologies marks a significant transformation in the educational landscape, with implications for teaching and learning performance. These technologies can generate content, simulate interactions, and adapt to learners' needs, offering opportunities for interactive learning experiences. In China's education sector, incorporating GAI technologies can address educational challenges, enhance teaching practices, and improve performance. This study scrutinises the impact of GAI technologies on learning performance in the education sector, focusing on the mediating roles of e‐learning competence (EC), desire for learning (DL), and beliefs about the future (BF), as well as the moderating role of facilitating conditions amongst Chinese educators. Data was collected from 411 teachers across various educational institutions in China using purposive sampling. PLS‐SEM and ANN were employed to assess the suggested structural model. The study results indicate that GAI technologies significantly influence learning performance by mediating EC, DL, and BF roles. Furthermore, facilitating conditions positively moderate the association amongst GAI technologies and EC, DL, and BF. This study underscores the critical role of self‐determination theory in shaping the effective incorporation of GAI technologies in education, offering valuable insights to improve teaching and learning outcomes in the Chinese education sector.