Columbia University（哥伦比亚大学）

2504.02275 2026-02-09 cs.LG

Enhancing Customer Contact Efficiency with Graph Neural Networks in Credit Card Fraud Detection Workflow

利用图神经网络提升信用卡欺诈检测流程中的客户联系效率

Menghao Huo, Kuan Lu, Qiang Zhu, Zhenrui Chen

机构 * School of Engineering（工程学院）； Department of Mechanical and Aerospace Engineering（机械与航空航天工程系）； University of Houston（休斯顿大学）； School of Electrical and Computer Engineering（电气与计算机工程学院）； Cornell University（康奈尔大学）； Fu Foundation School of Engineering and Applied Science（工程与应用科学学院）； Columbia University in the City of New York（纽约市哥伦比亚大学）

AI总结本文提出利用RGCN提升信用卡欺诈检测流程中客户联系效率，通过关系图结构提高识别准确性和减少误报。

Comments Published in Proceedings of the 2025 IEEE 7th International Conference on Communications, Information System and Computer Engineering (CISCE), pp. 320-324. DOI: 10.1109/CISCE65916.2025.11065245

Journal ref Proceedings of 2025 IEEE 7th International Conference on Communications, Information System and Computer Engineering (CISCE), pp. 320-324

2602.06134 2026-02-09 cs.HC cs.AI

Hear You in Silence: Designing for Active Listening in Human Interaction with Conversational Agents Using Context-Aware Pacing

在沉默中倾听：利用情境感知节奏设计人与对话代理的主动倾听

Zhihan Jiang, Qianhui Chen, Chu Zhang, Yanheng Li, Ray LC

机构 * The University of Hong Kong Hong Kong, SAR China Columbia University New York United States ； Renmin University of China Beijing China ； City University of Hong Kong\ for Narrative Spaces Hong Kong, SAR China Guangdong University of Technology Guangzhou China ； City University of Hong Kong\ for Narrative Spaces Hong Kong, SAR China ； The University of Hong Kong ； Columbia University ； Renmin University of China ； City University of Hong Kong\ for Narrative Spaces ； Guangdong University of Technology

AI总结本文提出通过情境感知节奏策略提升对话代理的主动倾听能力，通过实验验证其在人际关系和职业支持场景中的有效性。

Comments 29 pages, 10 figures. Conditionally Accepted to CHI '26

2512.17873 2026-02-09 cs.CV

Preserving Spectral Structure and Statistics in Diffusion Models

在扩散模型中保持谱结构和统计信息

Baohua Yan, Jennifer Kava, Qingyuan Liu, Xuan Di

机构 * Department of Civil Engineering and Engineering Mechanics, Columbia University, NY, United States（土木工程与工程力学系，哥伦比亚大学，纽约，美国）

AI总结 PreSS通过在谱空间中保持信息性先验和结构信号，提升扩散模型的生成质量和效率。