This volumes constitute the proceedings of the Workshops KGMA, SemiBDMA, MADM, AIEDM and STBDM which were held in conjunction with APWeb-WAIM 2025 which took place in Shenyang, China, during August 2830, 2025.

The 28 full papers presented in these proceedings were carefully reviewed and selected from 49 submissions. Focus on the following topical sections: KGMA Workshop; SemiBDMAWorkshop; SRBDMA Workshop; GDMMA Workshop.

Inhalt

**.- KGMA Workshop.
.- Graph-Augmented Language Models for Strategic Reasoning:AHybrid Retrieval Framework.
.- GrowCap: Lightweight Image Captioning with Enhanced Hard Prompts.
.- Fine-Grained Prototype-Augmented Named Entity Recognition.
.- Deep Generative Replay for Continual Relation Extraction.
.- Efficient Subgraph Matching withAdaptive Filtering and Core-Set Prioritization.
.- Graph Neural Network-DrivenAnt Colony Optimization for Combinatorial Problems on Graphs.
.- CXLPowered Memory Intensive Graph Processing:AComprehensive Survey.
.- Design and Implementation of a Graph Neural Network Framework for Crop Variety Classification Based on Complex NetworkAnalysis.
.- SemiBDMA Workshop.
.- MSPD: Large Language Model Parallel Speculative Decoding Method with Multi-level Early Exiting.
.- ANewYiFont Generation using Diffusion Model.
.- AsAI Meets Gaokao: Fine-Tuning GPT-4o for Enhanced Performance in Objective Questions.
.- Scene-Aware Multi-Agents on Collaborative Retrieval Augmented Generation Model.
.- Cascaded dense networks for super-resolution of face images.
.- Image Super-Resolution via Dense Convolutional Neural Network in Curvelet Transform Domain.
.- Design and Practice of Ship Real-time CollisionAvoidance System Based onArtificial Intelligence.
.- AChinese Named Entity RecognitionApproach with Multi-HeadAttention.
.- AFeature-Clustering Algorithm Based on K-Medoids.
.- Tilt Correction of Italic Character Images.
.- SRBDMA Workshop.
.- DITS: Double Imputation for Enhanced Irregular Multivariate Time Series Classification.
.- Transmission Tower Inspection Drone Hub Deployment Optimization System Based on DeepSeek Large Language Model.
.- SGDA: ClassAugmentation for Imbalanced Graph Learning at Scale.
.- zkANN: Efficient and Verifiable ANN Search with Enhanced Data Privacy.
.- Text Classification Method Based on Hypergraph Convolutional Network and Attention Mechanism.
.- Enhancing Traffic Sign Detection with EEM-YOLO for Complex Scenarios.
.- Knowledge Hypergraph Link Prediction Based onAttention and Relational Contrast Learning.
.- GDMMA Workshop.**
.- Multi-hop Reasoning Method Oriented QuestionAnswering in Time-constrained Temporal Knowledge Graph.
.- M4-Greedy: Visualization-Oriented Representative Selection of Time Series Data.
.- MAD-GCN: Multi-Attention Diffusion based Graph Convolutional Network for Traffic Flow Forecasting.