This book constitutes the proceedings of the 21st International Symposium on Intelligent Data Analysis, IDA 2022, which was held in Louvain-la-Neuve, Belgium, during April 12-14, 2023.The 38 papers included in this book were carefully reviewed and selected from 91 submissions.

 Contextual Word Embeddings Clustering through Multiway Analysis: A Comparative Study.- Transferable Deep Metric Learning for Clustering.- Spatial Graph Convolution Neural Networks for Water Distribution Systems.- Data-Centric Perspective on Explainability versus Performance Trade-off.- Towards Data Science Design Patterns.- Diverse Paraphrasing with Insertion Models for Few-Shot Intent Detection.- LEMON: Alternative Sampling for More Faithful Explanation through Local Surrogate Models.- GASTeN: Generative Adversarial Stress Test Networks.- Learning Permutation-Invariant Embeddings for Description Logic Concepts.- Diffusion Transport Alignment.- Mind the Gap: Measuring Generalization Performance Across Multiple Objectives.- Effects of Locality and Rule Language on Explanations for Knowledge Graph Embeddings.- Shapley Values with Uncertain Value Functions. -Revised Conditional t-SNE: Looking Beyond the Nearest Neighbors.- On the Change of Decision Boundary and Loss in Learning with Concept Drift.- AID4HAI: Automatic Idea Detection for Healthcare-Associated Infections from Twitter, A Framework based on Active Learning and Transfer Learning.- Explanations for Itemset Mining by Constraint Programming: A Case Study using ChEMBL data.- Translated Texts Under the Lens: From Machine Translation Detection to Source Language Identification.- Geolet: An Interpretable Model for Trajectory Classification.- An investigation of structures responsible for gender bias in BERT and DistilBERT.- Discovering diverse top-k characteristic lists.- Online Influence Forest for Streaming Anomaly Detection.- APs: a proxemic framework for social media interactions modeling and analysis.- User Authentication via Multifaceted Mouse Movementsand Outlier Exposure.- Explaining Black Box Reinforcement Learning Agents Through Counterfactual Policies.- A GNN-based Architecture for Group Detection from spatio-temporal Trajectory Data.- Discovering Rule Lists with Preferred Variables.- Don’t Start Your Data Labeling from Scratch: OpSaLa - Optimized Data Sampling Before Labeling.- The Other Side of Compression: Measuring Bias in Pruned Transformers.- Dropping incomplete records is (not so) straightforward.- Meta-Learning for Automated Selection of Anomaly Detectors for Semi-Supervised Datasets.- Should We Consider On-Demand Analysis in Scale-Free Networks?.- ROCKAD: Transferring ROCKET to whole time series anomaly detection.- Out-of-Distribution Generalisation with Symmetry-Based Disentangled Representations.- Forecasting Electricity Prices: an Optimize then Predict-based approach.- A Similarity-Guided Framework for Error-Driven Discovery of Patient Neighbourhoods in EMA Data.- QBERT:Generalist Model for Processing Questions.- On Compositionality in Data Embedding.