Innovation in Science and Technology

The CTP and MELD Scores: Both Are Used for the Prediction of Mortality for the Patients with Decompensated Liver Cirrhosis

Haradhan Kumar Mohajan — 2025-12-09

The Child-Turcotte-Pugh (CTP) score (range: 5-15) and the Model for End Stage Liver Disease (MELD) score (range: 6-40) have been used for predicting the survival of liver cirrhosis patients. Both scores are general indicators of the severity of liver failure in patients with end-stage liver disease, and have been validated for short- and long- term survival in cirrhosis. The liver cirrhosis is the development of hepatic fibrosis and regenerative nodules as a result of chronic liver injury that may lead to portal hypertension and end-stage liver diseases, such as chronic viral hepatitis, cirrhosis, hepatocellular carcinoma (HCC), and liver failure. The aim of this study is to evaluate and compare the aspects of CTP classification and MELD scores for predicting the severity of the cirrhosis to the liver transplantation patients.

“Capital Injection + Post-Investment Empowerment”: The Dual Engines of Fintech Industry Development

Zhile Tan — 2025-12-26

Despite the rapid expansion of financing scale in the fintech industry, only 32% of enterprises can break through the growth bottleneck. The core contradiction lies in the disconnection between capital injection and post-investment empowerment. Based on a sample of 327 global fintech companies from 2019 to 2023, this paper constructs a coupling coordination model of “capital injection intensity (CD) – post-investment empowerment depth (ED)” and employs methods such as two-way fixed effects and threshold regression to explore the impact mechanism of the dual engines on enterprise development efficiency. The results show that: (1) A 0.1 increase in the coupling coordination degree of the dual engines significantly enhances the enterprise revenue growth rate by 0.85 percentage points and increases the number of patent applications by 0.56 items, an effect far exceeding that of single capital injection. (2) There is a significant threshold effect. When the capital injection intensity CD≥0.35 and the post-investment empowerment depth ED≥0.42, the marginal effect of the dual engines on the revenue growth rate jumps from 0.52 percentage points to 0.98 percentage points. After breaking through the threshold, the two form a “complementary and synergistic” pattern. (3) The synergistic effect exhibits heterogeneity in fields and regions. Blockchain companies are most sensitive to the dual engines, while payment companies are the least sensitive. North American companies have significantly higher collaborative efficiency than Asian companies. This study quantifies the critical standards for the synergy of “capital - empowerment” for the first time, providing theoretical support and practical guidelines for equity investment institutions to optimize resource allocation, fintech companies to accurately connect with elements, and regulatory authorities to formulate differentiated policies.

The Application Boundaries and Risk Management of AI in Financial Transactions: An Empirical Study

Jun Xin — 2025-12-26

This paper focuses on quantifying the application boundaries of AI in financial transactions, identifying the cross-institutional risk spillover and transmission patterns, and constructing a multi-dimensional governance framework. Based on a dataset of 986 daily observations from 123 product accounts with a total asset value of 360 billion yuan from 2020 to 2024, combined with in-depth interviews with 15 leading asset management (AM) AI executives and comparative case studies, we develop a two-dimensional “scene-fit-risk tolerance” boundary quantification model and an “AI-securities firm-bank-asset management” risk transmission chain model. We propose the “3% boundary rule” and a three-tier governance framework of “technology-process-regulation.” Empirically validated by a century-old insurance asset management company, this framework reduced the AI transaction risk loss rate from 0.85% to 0.18% while maintaining a 35% improvement in transaction efficiency. It effectively addresses the core pain points of large-scale asset management institutions, such as blind AI application, concealed risk transmission, and an incomplete governance system, providing a solution with both theoretical support and practical value for the industry.

Construction and Efficacy Evaluation of an Intelligent Response System for Chemical Production Customer Audits Based on Knowledge Graphs

Xinshun Liu — 2025-12-26

Chemical production customer audits face intractable challenges, including heterogeneous audit standards, inefficient manual responses, and inadequate handling of complex cross-standard issues. Traditional manual response models have struggled to meet the evolving demands of high-stakes supply chain audits. This study integrates 236 heterogeneous audit standards from 127 core customers—including industry leaders such as Contemporary Amperex Technology Co. Limited (CATL) and Tesla—to construct a ternary knowledge graph (TKG) centered on “process parameters-quality indicators-compliance clauses.” An NLP-driven intelligent response system was developed to enable rapid semantic understanding, precise knowledge retrieval, and standardized response generation for audit queries. Comprehensive validation, including laboratory testing and 12 months of industrial application, demonstrates that the system achieves a question matching accuracy of 91.3%, reduces response time from 48 hours (traditional manual) to 15 minutes, and supports 27 customer audits with a 100% pass rate. The complex issue resolution rate reaches 89.6%, significantly reducing enterprise audit costs and compliance risks. The proposed technical framework effectively addresses the core pain points of multi-customer heterogeneous standard integration and intelligent audit response, providing a replicable technical pathway for audit management in the chemical industry and offering practical insights for the application of knowledge graphs and NLP in industrial compliance scenarios.