The six-volume proceedings set LNAI 15919, 15920, 15921, 15922, 15923 and 15924 constitutes the refereed proceedings of the 18th International Conference on Knowledge Science, Engineering and Management, KSEM 2025, held in Macao, China during August 47, 2025.

The 106 papers and 66 short papers are included in these proceedings were carefully reviewed and selected from 354 submissions. They focus on all aspects of the exchange of research in artificial intelligence, data science, knowledge engineering, AI safety, large language models, and related frontier areas.

Inhalt

.- Global Discovery: A Global Graph-RAG Approach for Query-Focused Multimodal Summarization Across Multiple PDF Papers.
.- Metric Embedding Initialization-Based Differentially Private and Explainable Graph Clustering.
.- M³Net: Multimodal-Feature-masked Networks for Fake News Detection.
.- TB-DML4HS: A Task-Based Modeling and Causal Effect Estimation Method Using DML for Heterogeneous UAV Swarm.
.- WebGCN: Web Information Extraction Algorithm Based on Graph Neural Networks.
.- Federated Rank Learning with Dimensionality Reduction and Clustering for Electricity Load Forecasting.
.- Context-Aware Vectors: A New Method Integrating Personality into LLMs for Enhanced Sentiment Analysis.
.- Research on Detection and Reconstruction of Multiple Types of Anomalies in Wind Speed-Power Data of Wind Farms.
.- An LLM-enabled data augmentation framework for low-resource scenarios.
.- Lightweight Remote Sensing Tiny Object Detection Model Based on YOLOv8n Architecture.
.- Causal Encoding Generative Model Based on Attention and KAN.
.- TRVP: Transformer-VAE Framework for 3D Point Cloud Instance Segmentation.
.- A Metapath-based Neighborhood Reconstruction network for Graph Anomaly Detection.
.- DarkFusionNet: A Fusion Neural Network Based Architecture for Darknet Text Classification.
.- Key Nodes Evaluation for Human Proximity Networks Based on Gravity Model.
.- SRViT-MCNet: An IoT malware classification model.
.- MultiTEmb: Multi-Scale Embeddings for Temporal KG Completion.
.- Education Distillation: Let the Model Learn in the School.
.- A Multi-Source Temporal Graph Approach for Reliable Market Forecasting with LLM Synergy.
.- Memory-Enhanced Transformer Adaptive Graph Convolutional Recurrent Network for traffic flow forecasting.
.- Investigation into Auto-Scaling Mechanisms in Cloud Computing.
.- Multi-Receptive-Field Feature Fusion Knowledge Graph Embedding for Link Prediction.
.- RAG with Visual Alert: Boosting Multimodal Language Models for Enhanced Visual Question Answering.
.- Dynamic Heterogeneous Graph Neural Network for Personality Detection in Chinese Social Media Texts.
.- DualCBR: Cross-Modal Collaborative Filtering with Bidirectional Alignment for Long-Tail Recommendation.
.- Semantic Information Extraction with Language Models for Zero-Day Attack Detection.
.- A Review of Optimization Techniques for Large Language Model Inference.
.- GeoER: A Challenging Benchmark for Geometric Element Recognition.
.- SocioSupplyAlert: Comprehensive Supply Chain Crisis Prediction Using LLMs and Social Media Data.
.- SKG-LLM: Enhancing Large Language Models with Sentiment Knowledge Graphs for Fine-Grained Sentiment Analysis.
.- Harnessing Heterogeneous Social Networks for Better Group Recommendations: An Integrated Approach Towards Cold-start Problem.
.- An Agent-Based Cybersecurity Framework Enhanced by Large Language Models: Integrating Retrieval-Augmented Generation and Monte Carlo Tree Search.
.- Constrained Multi-agent Reinforcement Learning Approach on Wireless Charging Scheduling.
.- A Dual-Module System Design and Application for Digital Payment Fraud Detection.