Purpose: Under the guidance of the concept of openness and development, the paper grasps the mechanism of technology spillover in developed countries and analyzes how to better absorb advanced manufacturing technology based on empirical analysis so as to point out the path for the transformation and development of China’s digital manufacturing industry.Design/methodology/approach: The paper constructs the panel data model and further analyzes the impact of international technology spillovers on the transformation and development of the digital manufacturing industry.Findings: This paper measures the level of technology spillover in the Yangtze River Delta region and finds that foreign direct investment (FDI) technology spillover and import trade technology spillover among four provinces and cities show a growth trend from 2010 to 2017. But after 2017, there is a certain degree of decline.Originality/value: With the advent of industry 4.0, the digital manufacturing industry of all countries in the world is developing with a new attitude, the global technology spillover methods are diverse and the spillover channels have changed greatly, which will affect the transformation and upgrading of China's digital manufacturing industry.

This study is aimed to promote China’s support for innovation and entrepreneurship in the context of the rapid development of the artificial intelligence industry. Considering the advance law-based governance implemented in this country, the relationship between the psychological understanding and behavior choice of entrepreneurs in the law industry and artificial intelligence industry was explored. The investigation was conducted in the form of questionnaire survey, and the data were collected by sending emails. In addition, the model was established to make the investigation direction clearer. The reliability analysis of the questionnaire was carried out by means of Cronbach α reliability coefficient. The validity analysis of the questionnaire was carried out by Kaiser-Meyer-Olkin (KMO) coefficient. The statistical software was used to analyze the correlation between the factors in the investigation. The results show that there is a positive correlation between behavior speed in the psychological understanding of entrepreneurs in the legal industry as well as the entrepreneurial experience and entrepreneur behavior. The concentration of entrepreneurs’ behavior is negatively correlated with the entrepreneur’s industrial experience. The behavior speed and concentration of entrepreneurs in the artificial intelligence industry are negatively correlated with the age of entrepreneurs, but their entrepreneurial behavior and psychological understanding do not show correlation. The behavior speed and concentration of entrepreneurs in the artificial intelligence industry are positively correlated with product innovation, technological innovation, and market innovation. Therefore, the behavior choice of entrepreneurs in the legal industry are highly correlated with psychological understanding. The entrepreneurs’ psychological understanding and behavior choice in the field of artificial intelligence do not show a direct correlation but indirect correlation through indirect factor innovation opportunities. The investigation on entrepreneurs’ psychological understanding and behavior choice can provide important theoretical support for entrepreneurs in China’s legal industry and artificial intelligence industry.

The objective of question generation from knowledge graphs (KGQG) is to create coherent and answerable questions from a given subgraph and a specified answer entity. KGQG has garnered significant attention due to its pivotal role in enhancing online education. Encoder-decoder architectures have advanced traditional KGQG approaches. However, these approaches encounter challenges in achieving question diversity and grammatical accuracy. They often suffer from a disconnect between the phrasing of the question and the type of the answer entity, a phenomenon known as semantic drift. To address these challenges, we introduce LEMON, a knowledge-enhanced, type-constrained, and grammar-guided model for KGQG. LEMON enhances the input by integrating entity-related knowledge using heuristic rules, which fosters diversity in question generation. It employs a hierarchical global relation embedding with translation loss to align questions with entity types. Additionally, it utilizes a graph-based module to aggregate type information from neighboring nodes. The LEMON model incorporates a type-constrained decoder to generate diverse expressions, and improves grammatical accuracy through a syntactic and semantic reward function via reinforcement learning. Evaluations on benchmark datasets demonstrate LEMON's strong competitiveness. The study also examines the impact of question generation quality on question-answering systems, providing guidance for future research endeavors in this domain.

Question answering (QA) and question generation (QG) over knowledge graphs (KGs) are inherently complementary, as they both entail the mapping of natural language to relations in KGs and the formation of reasoning paths. Existing studies show that KGQG can enhance the KGQA systems through data augmentation. An interesting discussion is whether KGQA can also benefit KGQG models. It can be postulated that the joint learning of KGQA and KGQG may result in mutual improvement, given that the two tasks are symmetrical in form. However, the effectiveness of their joint learning has not been rigorously investigated. This paper presents a modular dual learning framework for KGQA and KGQG, which fosters dual learning through an adaptively chosen and self-organized modular sharing mechanism to boost the performance of both tasks. The framework comprises three core modules: the KGQA module, the KGQG module, and the modular sharing network. Specifically, for a given input, the model initially generates a routing selector via a discrete latent variable in a mixture of experts fashion. Subsequently, the selector produces a routing that flexibly selects appropriate network layers from the shared network collection and organizes them into a network adapted to the current task. The shared neural layers are identical in structure and parameter size. In the construction of the new network, the shared layers can be positioned at various points along the sequential path and subsequently reused. As a result, these layers can contribute to the final output at distinct stages of processing. To enhance the current module by incorporating a dual task inductive bias, an assessment procedure is recommended to determine if the shared network should be included in the present computation. The proposed framework has been demonstrated to substantially boost the performance of KGQA and KGQG in experiments conducted on two widely-utilized benchmarks. The utilization of dual signals within the shared network has the potential to enhance the model's generalization capabilities. Furthermore, when the trained module is substituted with a randomized network, the modular shared network still demonstrates notable advancement of the new model with slight fine-tuning, which suggests the transferability of knowledge.

Urban traffic congestion remains a persistent challenge in smart cities, where complex spatiotemporal dependencies and limited interpretability hinder effective prediction and control. In law-governed road systems, effective prediction and intervention further depend on the accurate, automatic recognition of traffic laws (e.g., speed limits, lane-use rules, turn restrictions, signal timing plans), which provides the legal-operational constraints for intelligent control. Traditional models often fail to capture evolving causal relationships across road networks and lack the capability to simulate the impact of traffic interventions. To address these limitations, we propose DAST-CI, a Dual-Attention Spatiotemporal Causal Inference network that integrates dynamic causal graph learning, multi-modal feature fusion, and interpretable intervention reasoning.