As organizations increasingly rely on multi-modal information including text, images, audio, and structured databases, there is a critical need for intelligent systems that can seamlessly process, understand, and connect insights across different data modalities. This workshop brings together researchers and practitioners to explore cutting-edge approaches in multi-modal representation learning, neural architectures for semantic fusion, and the integration of large language models for enhanced understanding.

This workshop provides a unique forum for in-depth discussion. We welcome submissions that are aligned with the workshop's thematic. Papers, consisting of only a paper title and an abstract, should be sent to lwlmath@buaa.edu.cn. Please note that accepted papers are for presentation and discussion only and will not be included in the official conference proceedings.

9:00 – 12:00, February 2 (Monday), 2026

Chair: Wenling Li (Beihang University)

• Natural Language to SQL Translation through Understanding Data Systems

Zhenying He (Fudan University)

• Resilience Analysis of Multilayer Complex Networks via High-Order Failure Structures

Jianfeng Guan (Beijing University of Posts and Telecommunications)

• BAG: A Bound-Aware Graph Partitioned Index for Spatial Queries over Moving Objects on Road Networks

Ziqiang Yu (Yantai University)

• Efficient and Robust Deep Semantic Fusion and Representation Learning for Cross-modal Retrieval and Mining

Meiyu Liang (Beijing University of Posts and Telecommunications)

• Phase-Harmonized Transformer for Robust Synthetic Image Detection

Wenqi Ren (Sun Yat-sen University)