Abstract: This study employs bibliometric methods and knowledge mapping to analyze 164 papers on panic buying from the Science Citation Index Expanded database. It examines the research landscape, identifies key hotspots, traces evolutionary pathways, and proposes future directions. The findings reveal that: 1) Annual publication output has transitioned through three distinct phases: a low-frequency exploratory period, an explosive growth period, and a stable development period. 2) A novel method for identifying research hotspot clusters—based on core keywords, cluster labels, and central terms—is proposed, leading to the identification of four major knowledge clusters: triggering contexts, psychological drivers, behavioral manifestations, and coping strategies. 3) Research topics exhibit clear evolutionary phases: an early exploratory stage (pre-2019), a pandemic-driven systematic research stage (2020–2022), and a stage of theoretical model refinement and innovation (post-2023). 4) Future research should progress systematically along four dimensions: theoretical foundations, content expansion, methodological innovation, and practical applications. Read More

Abstract: TikTok short videos are a product of advancements in internet technology and modern information technology. They provide users with a wide variety of content and knowledge. Due to their ease of use, convenience, entertainment value, and feedback interaction, among other advantages, the number of TikTok app users continues to grow. However, given the intense competition in the app market, it is particularly important to ensure user retention. Therefore, this study employs SPSS 16.0 and AMOS 17.0 software to empirically test the model of factors influencing users' willingness to use TikTok. The findings indicate: ease of use positively influences users' willingness to use TikTok, entertainment value positively influences users' willingness to use TikTok, and feedback interaction positively influences users' willingness to use TikTok. Among these, ease of use has the greatest impact on users' willingness, with a specific value of 0.33, followed by entertainment value with a value of 0.21, while feedback interaction has the smallest impact, with a value of 0.21. Read More

Abstract: Industrial supply chain networks are increasingly exposed to compound disruptions caused by facility failures and road interruptions. These disruptions may reduce supply capacity, block transportation links, and lead to unmet demand. To address this problem, this paper studies the resilient design of a regional industrial supply chain network under joint facility-road disruptions. A two-stage robust optimization model is developed. In the first stage, facility location, facility protection, and road protection decisions are made before disruptions occur. In the second stage, transportation reallocation and shortage recourse decisions are optimized after the worst-case disruption scenario is realized. To solve the resulting min-max-min model, a column-and-constraint generation algorithm is used. The algorithm decomposes the original problem into a master problem and a subproblem. The master problem determines the first-stage network design, while the subproblem searches for the worst-case joint disruption scenario. A case study based on the electronic product supply chain in Guangdong Province is conducted to test the model. The results show that the proposed method can identify critical facilities and key roads, improve post-disruption service performance, and reduce shortage losses under extreme disruption scenarios. This study provides a practical decision-making framework for resilient supply chain network design under compound disruption risks. Read More

Abstract: Port throughput is a key indicator of port operation efficiency and economic contribution. Accurate forecasting helps improve port management and risk control. This paper uses monthly container throughput data of Shanghai Port from 2013 to 2023 to compare the ARIMA model and Holt-Winters exponential smoothing method. After trend and stationarity tests, ARIMA (1,1,2), Holt-Winters additive, and multiplicative models are constructed. Model performance is evaluated using MAE, RMSE, and MAPE. The results show that the Holt-Winters multiplicative model performs best and is used to forecast throughput from January to June 2025. Read More

Abstract: Fresh products are difficult to manage because they are perishable, demand changes quickly, and losses are strongly affected by storage and purchasing decisions. Taking Xinyue Laike Supermarket as a case, this paper rewrites the original thesis into a concise academic article. The study first identifies the main problems in the supermarket, including traditional inventory management, subjective ordering, high fruit loss rates, weak demand forecasting, and insufficient supplier cooperation. It then proposes an optimization framework based on ABC classification, a perishable-goods EOQ pricing model, particle swarm optimization, exponential smoothing, and supplier collaboration. The rewritten framework shows that high-value fruits should be controlled strictly, medium-value fruits should be supported by demand forecasting, and low-value fruits should be managed through stable cold-chain cooperation and simplified replenishment rules. The research provides a practical and operable method for reducing fruit inventory cost and improving fresh-food inventory efficiency in small supermarkets. Read More

Abstract: This paper investigates decision-making issues of green supply chains under alternative power structures. Assuming that the manufacturer and two retailers are fully rational and risk-neutral decision-makers, three game-theoretic models are established corresponding to manufacturer-led dominance, retailer-led dominance and balanced bilateral power, whose equilibrium solutions are derived analytically. Comparative analysis and numerical simulation of the three game models are implemented to identify the impacts of disparate power structures on the decision-making behaviors and profits of upstream and downstream enterprises within the supply chain. The results reveal that the effects of power structures on green technology investment, order quantity and supply chain members’ profits are predominantly contingent on consumers’ green preference coefficient and the impact coefficient of green technology investment on wholesale price. Read More

Abstract: To address the dilemma of "daring not to transform, unwilling to transform, and unable to transform" faced by small and medium-sized enterprises (SMEs) in digital transformation, this paper applies perceived utility theory to construct a digital adaptability index measurement model for SMEs. The model defines positive utility as the weighted sum of belief values and attribute weights of perceived attributes, and negative utility as the product of perceived cost and cost weight. The net utility is defined as the adaptability index. Hierarchical thresholds are set to classify SMEs into three levels: high, medium, and low adaptability. Furthermore, based on importance-performance analysis (IPA), an obstacle identification model is constructed, where attribute weights serve as importance indicators and belief values serve as performance indicators. Key obstacles are identified and improvement priority scores are calculated to achieve ordered ranking. This paper aims to provide an operable model tool for the diagnosis and tiered intervention of digital transformation in SMEs. Read More

Abstract: As semiconductor devices scale and advanced packaging architectures such as 2.5D interposer-based integration, 3D chip stacking and heterogeneous chiplet assembly are adopted, local power densities have surpassed 600 W/cm2, which is beyond the capability of standard thermal management techniques. Thermal interface materials (TIMs) are strategically critical to the thermal path of packaged devices bridging the high resistance solid-solid contact surfaces between the die, heat spreader and heat sink. This review covers the physical mechanisms of the interfacial thermal resistance in advanced packaging, the classification and performance of the major classes of TIMs (carbon-based, metal-based and polymer composite systems), and the crucial reliability challenges (pump-out, thermal cycling degradation and 3D stacking heat path interruption) which prevent the widespread use of TIMs today. The review also examines the stimulus-responsive smart TIMs as the platform to meet the dynamic thermal management requirements of next-generation heterogeneous integration platforms. Read More