Data Science: Foundations and Applications

The two-volume set LNAI 15875 + 15876 constitutes the proceedings of the 29th Pacific-Asia Conference on Knowledge Discovery and Data Mining, PAKDD 2025 Special Session, held in Sydney, NSW, Australia, during June 10-13, 2025.

The 68 full papers included in this set were carefully reviewed and selected from 696 submissions. They were organized in topical sections as follows: survey track; machine learning; trustworthiness; learning on complex data; graph mining; machine learning applications; representation learning; scientific/business data analysis; and special track on large language models.

. - Survey Track.
. - Large Language Models for Cybersecurity Education: A Survey of Current Practices and Future Directions.
. - A Comprehensive Survey on Deep Learning Solutions for 3D Flood Mapping.
. - A Survey of Foundation Models for Environmental Science.
. - A Survey on Efficient Graph Reachability Queries.
. - Machine Learning.
. - Disentangled Representation Learning for Geospatial-temporal Data Modeling.
. - Treatment Effect Estimation for Graph-Structured Targets.
. - Dynamic DropConnect: Enhancing Neural Network Robustness through Adaptive Edge Dropping Strategies.
. - The Brownian Integral Kernel: A New Kernel for Modeling Integrated Brownian Motions.
. - Fed-ARIMA-OPARBFN: An Ensemble Model for Cross-Domain Crop Yield Time Series Prediction Based on Federated Learning.
. - S-CPD: Topological Smoothing-Based Change Point Detection.
. - VDASI: VAE-Enhanced Degradation-Aware System Identification Using Constrained Latent Spaces.
. - Disentangled Mode-Specific Representations for Tensor Time Series via Contrastive Learning.
. - PFformer: A Position-Free Transformer Variant for Extreme-Adaptive Multivariate Time Series Forecasting.
. - Advancing Long-Term High-Frequency Dissolved Oxygen Forecasting for Australian Rivers.
. - CNO-former: Chaotic Neural Oscillatory Transformer for Social Media Text Generation.
. - Multilingual Non-Factoid Question Answering with Answer Paragraph Selection
. - Turning Uncertainty to Information by Intervals in Ensemble Classifiers.
. - Determining the Need for Multi-Label Classifiers by Measuring Unexplained Covariance.
. - Evaluating Generative Vehicle Trajectory Models for Traffic Intersection Dynamics.
. - Trustworthiness.
. - Inversion Triplet - A Contrastive Backdoor Mitigation Method for Self-Supervised Vision Encoders.
. - Beyond Uniformity: Robust Backdoor Attacks on Deep Neural Networks with Trigger Selection.
. - Defence Against Multi-target Multi-trigger Backdoor Attack.
. - How to Backdoor Consistency Models? .
. - Multi-granularity Policy Explanation of Deep Reinforcement Learning Based on Saliency Map Clustering.
. - FACROC: A Fairness Measure for Fair Clustering Through ROC Curves.
. - Learning on Complex Data.
. - Action Sequence Analysis Using Temporal Commonsense Knowledge.
. - Foundation Model for Lossy Compression of Spatiotemporal Scientific Data.
. - CANTER: A Novel Causal Model for Tourism Demand Forecasting.
. - Time-Aware Complex Attention Space for Temporal Knowledge Graph Completion.
. - Adaptive Extraction of Variable-Length Subsequence Patterns in Noisy Time Series.
. - Hunting Inside N-Quantiles of Outliers (Hino).
. - Fast Approximation Algorithm for Euclidean Minimum Spanning Tree Building in High Dimensions.
. - ShuttleSHAP: A Turn-Based Feature Attribution Approach for Analyzing Forecasting Models in Badminton.
. - Offline Map Matching Based on Localization Error Distribution Modeling.