Zhaoxu Yu’s scientific contributions

Based on para-social interaction (PSI) theory and social identity perspective, this study explores the mechanisms driving consumers’ impulse buying in social live shopping. It examines how live content design, namely information comprehensiveness (INFCOM) and interactivity (INT), affects consumer cognition and affective experiences, namely perceived usefulness (PU), PSI, and sense of belonging (SOB), to generate the influence of the urge to buy impulsively (UBI), and further explores the moderating role of the consumer–broadcaster para-social relationship (PSR) between live content design and consumer experience. Findings indicate that in an accumulative social live shopping environment, comprehensive information and strong interactivity enhance consumer social identity, reduce shopping hesitations and obstacles, and encourage UBI. Forming a close consumer–broadcaster relationship is crucial for promoting social identity and increasing UBI. Even without interactive engagement, consumers who feel a close connection with the broadcaster still experience interaction and SOB. PSR influences impulse buying by enhancing consumer perceptions and thereby promoting UBI. This study advances the understanding of impulse buying from a social identity perspective and suggests that merchants and livestream designers can improve quality and sales by providing comprehensive product information and incorporating diverse interactive elements in live broadcasts.

In today’s increasingly competitive globalization, innovation is crucial to technological development, and original innovations have become the high horse in the fight for market dominance by enterprises and governments. However, extracting original innovative technologies from patent data faces challenges such as anomalous data and lengthy analysis cycles, making it difficult for traditional models to achieve high-precision identification. Therefore, we propose a Multi-Dimensional Robust Stacking (MDRS) model to deeply analyze patent data, extract leading indicators, and accurately identify cutting-edge technologies. The MDRS model is divided into four stages: single indicator construction, robust indicator mining, hyper-robust indicator construction, and the pioneering technology analysis phase. Based on this model, we construct a technological development matrix to analyze core 3D-printing technologies across the industry chain. The results show that the MDRS model significantly enhances the accuracy and robustness of technology forecasting, elucidates the mechanisms of technological leadership across different stages and application scenarios, and provides new methods for quantitative analysis of technological trends. This enhances the accuracy and robustness of traditional patent data analysis, aiding governments and enterprises in optimizing resource allocation and improving market competitiveness.