Conference Program
08:45 - 09:30
Opening Session
09:30 - 10:30
10:30 - 11:00
Coffee Break
11:00 - 12:40
Parallel Sessions
Room 1 - Session 1A - KRR
Chair: To Be Announced
- MaxSAT-Based Feedback for Guiding Vision-Language Models in SudokuPedro Orvalho, Guillem Alenyà and Felip Manyà
- Revisiting Incremental Linearization for Nonlinear Integer ArithmeticMarek Dančo, Karel Chvalovský and Mikoláš Janota
- Modeling Constrained Routing in PDDL: Lessons from 11 Independent EncodingsArthur Mello and Bruno Ribas
- Modeling the Aquasplash Puzzle Game in PDDLAndré Silva and Bruno Ribas
Chair: To Be Announced
- Defending Federated Learning Against Poisoning Attacks: A Comparative Study of Unsupervised Anomaly Detection MethodsYohannes Biadgligne, Yassine Baghoussi, Kai Li and Alípio Jorge
- Error- and Prediction-Driven Motor Learning in the Cortico-Cerebellar LoopAna Carolina Filipe, Rui Ponte Costa and Claudia Soares
- Physics-Guided Frequency-Aware CNNs for Robust Physical Time Series ForecastingAmal Saadallah, Julia Tjus, Petra Wiederkehr and Wolfgang Rhode
- Fast and Interpretable Autoregressive Estimation with Neural Network BackpropagationAnaísa Lucena, Ana Martins, Armando Pinho and Sónia Gouveia
Chair: To Be Announced
- Large-Scale User Behavior Analysis in Multimodal AI-Assisted Manual Task ExecutionRafael Ferreira, Diogo Tavares, Diogo Glória-Silva, David Semedo and João Magalhães
- Auditing AI-Assisted Software: Development Traces, Code Quality, and Trustworthiness in a Remote-Sensing PipelineOraib Almegdadi, Miguel Monteiro and Nuno C.Marques
- Role-Conditioned LLMs for Diverse Synthetic Expert Panels: Expert Elicitation on Barriers to Direct Current AdoptionCatherine Ngirwa and Lucas Pereira
- RAG-DW: A Multi-Level Semantic Retrieval Architecture Integrated with Data Warehouses for Natural Language SQL GeneratioGabriela Albuquerque, Paulo Pinheiro, Pedro Martins, Maryam Abbasi and Luís Cavique
12:40 - 14:00
Lunch
14:00 - 15:40
Parallel Sessions
Room 1 - Session 2A - AI4IS
Chair: To Be Announced
- Hybrid Visual Recognition for Smart Tourism Augmented RealityGabriel da Silva Neto, Paulo Pinheiro and Rui Pedro Lopes
- Multi-Agent Reinforcement Learning for Procurement Auctions Based Distributed Energy Market GovernanceFrederico Metelo, Luís Pisco and Cláudia Soares
- A Three-Stage Cascade Pipeline for DDoS Detection and Zero-Day Anomaly DiscoveryMarco Ferreira, Marcos Freitas and Fábio Mendonça
- Evaluating PPO-driven Agents for Replenishment Optimization of Short-Life Cycle ProductsJosé Borges, Luís Carvalho, Sofia Machado and Luís Paulo Reis
Room 2 - Session 2B - AITD
Chair: To Be Announced
- A Comprehensive Analysis on Predictive Machine Learning for Win Probability in Live Football MatchesVicente Felino, Diogo Cordeiro, David Semedo and Nuno Correia
- A two-step sequential approach for hyperparameter selection in finite context modelsJosé Contente, Ana Martins, Armando Pinho and Sónia Gouveia
- Weakly Supervised Classification of Focal and Non-focal Sperm Whale Clicks: A Comparative StudyJosé Cunha, Sofia Cavaco, Cláudia Oliveira, Sergi Pérez-Jorge, Rui Prieto, Mónica Silva and Joaquim Silva
- Chronological classification in distributed RAG systems, with continuous data feedGonçalo Freitas and Manuel Rodrigues
Room 3 - Session 2C - Student Symposium
Chair: To Be Announced
- Towards Reliable Agentic Decision Support: Overcoming the Plausibility ParadoxAlexandre Godinho
- From Judicial Reading Practices to AI Legal Research Design - Evidence from International CourtsAnna Bugaeva
- From Elicitation to Interaction: LLMs in Asynchronous Group Expert Elicitation on MVDC and LVDC Technology AdoptionCatherine Ngirwa
15:40 - 16:10
Coffee Break
16:10 - 17:50
Parallel Sessions
Room 1 - Session 3A - AI4IS / AmIA Environments
Chair: To Be Announced
- Towards a data-centric KDD pipeline for natural-stone valorization under weak supervisionLudwig Krippahl, Cláudia Soares and João Leite
- Real-Time Bayesian AI for Mission-Critical Sensor Fusion in Maritime Command and ControlMiguel Roque, Pedro Reis Nunes, Nuno Silva and João Galego
- Benchmarking Open-Source LLMs for Hybrid Educational Conversational Assistants Built with Rasa CALMAna Alves, Renata Magalhães, Dalila Duraes and Paulo Novais
- Towards Multimodal Physiological Stratification Using Self-Supervised Photoplethysmography Representation LearningGabriel Pinto, Rita Ribeiro, Goreti Marreiros and Luís Conceição
Room 2 - Session 3B - NLP-TeMA
Chair: To Be Announced
- On the Robustness of Medical Image Report Generation: Diagnosis Label Entailment Under Perturbed ConditionsFlávia Carvalhido, Henrique Lopes Cardoso, Vitor Cerqueira and Carlos Soares
- From Prototype to Platform: SaiLR 2.0 for AI-Assisted Systematic Literature ReviewsHeike da Silva Cardoso, Vítor Rocio and José Cravino
- Separating Human from Machine: Linguistic Analysis of Fake News in Brazilian PortuguesePedro Andrade, Renato Silva and Thiago Pardo
- PorTEXTO: A European Portuguese Benchmark for Visual Text ExtractionJoão Pereira, Diogo Glória-Silva, Manuel Letras da Luz, Rafael Ferreira, Diogo Tavares, David Semedo and João Magalhães
Room 3 - Session 3C - Student Symposium
Chair: To Be Announced
- Differentiable Estimation of ARMA Models via BackpropagationAnaísa Lucena
- LLM-driven Semantic Interoperability Framework in Electric Mobility DataspacesZakaria Moshi
19:00
Welcome Reception at Funchal City Hall
09:00 - 11:05
Parallel Sessions
Chair: To Be Announced
- Grooming-like Patterns in AI Chatbot Interactions with Children: A Novel Perspective?Beatrice Balzola and Florencia Valentina Maldonado
- Quality Assessment in Subgroup Discovery Using Uncertainty Targets: A Healthcare Use CaseEma Duarte, Ana Cravidão Pereira, Marília Barandas, Duarte Folgado, Carlos Soares, André Carreiro and Hugo Gamboa
- Gen AI for deceptive purposes in pre-trial investigations. Deepfakes vis-à-vis human and fundamental rightsLudovica Paseri and Serena Quattrocolo
- Automatic Legal Descriptor Generation for Portuguese Legal DocumentsGiovanni Maffeo de Medeiros, Catarina Silva and Hugo Gonçalo Oliveira
- Human Oversight in AI Governance: Ensuring Effectiveness Through By-Design Measures and Coordinated ResponsibilitiesElisa Puntil and Massimo Durante
Chair: To Be Announced
- New Results on Synthetic Speech Detection for European PortugueseRafael Geraldo dos Santos, André Miguel Viegas Oliveira, Helder Daniel and José Valente de Oliveira
- A Physics-Informed Neural Network for Muographic Density Estimation in ConcreteDaniele Corradetti and José Paulo Costa
- Overlap-controlled frame extraction for underwater photogrammetryDaniele Corradetti and José Paulo Costa
- Adapting SAM2-UNet for High-Resolution LULC Semantic Segmentation with OpenStreetMap Auxiliary SupervisionTiago Santos, Oraib Almegdadi, João Manso and Nuno C.Marques
- reBEN7-S2UNet: a Sentinel-2 U-Net for Seven-Class LULC Mapping in Portuguese LandscapesOraib Almegdadi, Rajchandar Padmanaban and Nuno C.Marques
Chair: To Be Announced
- NER Models for Portuguese Emergency Room Notes: Extracting Diagnoses, Medication Allergies, and Usual MedicationsNuno Guimarães, Tahsir Ahmed Munna, Ana Luísa Fernandes, Alexandra Mendes, Luísa Santa-Marinha, Francisco Bischoff, Ana Margarida Lopes, Luís Filipe Cunha, António Leal, Purificação Silvano and Alípio Jorge
- Beyond Downstream Accuracy: A Multi-Axis Framework for Evaluating Synthetic EEG in Seizure DetectionVanda Barata, Ana de Almeida and Luís Nunes
- A Transition-Aware Framework for Wearable Sleepiness Scoring in Real-Time and Near-Real-Time Sleep MonitoringHaochen Liu and Sam Nallaperuma-Herzberg
- NeurAFTs: Neural Accelerated Failure Time ModelsDavid Calhas, Ziga Crv, Lachlan Gilchrist, Petroula Proitsi, Dejan Georgiev, Vida Groznik and Aleksander Sadikov
- Evaluating the Privacy--Utility Trade-off of Differential Privacy in Machine Learning Models for Clinical Outcome PredictionMaryam Abbasi, Paulo Váz, José Silva, Maria Váz and Pedro Martins
11:05 - 11:30
Coffee Break
11:30 - 12:30
12:30 - 14:00
Lunch
14:00 - 15:15
Parallel Sessions
Room 1 - Session 5A - AIL
Chair: To Be Announced
- Adversarial Attacks on Legal NLP: Evaluating the Robustness of Contract Analysis ModelsMaryam Abbasi, Paulo Váz, José Silva, Maria Váz and Pedro Martins
- Beyond Single-Step Judgments: Recursive Evaluation of Legal Summaries with Language ModelsDiego Silva, Allberson Bruno de Oliveira Dantas and Vládia Pinheiro
- VeLex: A Neuro-Symbolic Pipeline for Compliance-Matrix Extraction from Swiss Financial RegulationJulius Kooistra, Dirk Spiegel, Lucia Gomez Teijeiro and Branka Hadji Misheva
Chair: To Be Announced
- Spatial Agency Beyond the Prompt: An in-the-wild evaluation of Generative AI Literacy in Architectural EducationMichela Barosio, Enrico Pupi, Piergiuseppe Rechichi, Francesco Manigrasso and Filippo Gabriele Pratticò
- Integrating Keypoint Modality into Large Language Model for Continuous Sign Language TranslationTiago Gonçalves, Pedro Jorge, Rui Jesus and David Semedo
- Agentic AutoML: LLM-driven Semantic Auditing for Model Integrity in Automated Feature SelectionJoão Gois and Filipe Portela
Room 3 - Session 5C - AITS
Chair: To Be Announced
- Robotic Navigation of a Delivery Robot Based on Behavior TreesMiguel Cruzeiro, Eurico Pedrosa and Nuno Lau
- A Robust Pipeline for Container Code Recognition Using YOLOv8 and TrOCRGonçalo Santos, Jacinto Estima and Catarina Silva
- Generative Counterfactual Explanations for Electric Vehicle Eco-DrivingLeonardo Carneiro Feltran, Paulo Ribeiro, Luan Lopes dos Santos, Lucas Airam Castro de Souza, Marcelo Luiz Drumond Lanza, Robson Francisco Dias, Luís Henrique Maciel Kosmalski Costa, Miguel Elias Mitre Campista and André Carlos Ponce de Leon Ferreira de Carvalho
15:15 - 15:45
Coffee Break
15:45 - 16:45
Special Invited Talk: João Pavão Martins, Instituto Superior Técnico / SISCOG
16:45 - 18:00
APPIA General Assembly
19:00
Conference Dinner at Fortress of São Tiago
09:30 - 11:10
Parallel Sessions
Room 1 - Session 6A - AITMA
Chair: To Be Announced
- Object Detection as a Differentiable Interpretability BottleneckRicardo Cruz and Jaime dos Santos Cardoso
- A CNF-Style PDDL Encoding for Akari, and What It CostsBruno Ribas and Fabiano Silva
- GeneNet-DARTS: Differentiable Architecture Search for the Computational Evolution of Gene CircuitsJefferson Maxwell Ataide de Almeida, José Victor Brito Costa, Guilherme Ribeiro, Matheus Menezes, Lewis John Grozinger, Alexandre De Oliveira and Ángel Goñi-Moreno
- AdamX: Cosine similarity meets gradient descentFrancisco M. Caldas, Ruben Belo and Cláudia Soares
Room 2 - Session 6B - GAAI
Chair: To Be Announced
- AMALIA-VL: A Native European Portuguese Open-Source Vision and Language ModelDiogo Glória-Silva, João Pereira, Manuel Letras da Luz, Afonso Simplício, Gonçalo Vinagre, Diogo Tavares, Rafael Ferreira, Inês Vieira, Inês Calvo, David Semedo and João Magalhães
- Large Language Models Are Not (Yet) Robust in Understanding Code Against Semantics-Preserving MutationsPedro Orvalho and Marta Kwiatkowska
- On the Limits of Similarity-Based Policy Detection in Course-Specific RAG: A Bilingual Ablation StudyLuís Lamy, Nelson Chibeles-Martins and Nuno C. Marques
- Sensitivity Ridges in Conditioned Generative ModelsMoritz Schwind
Room 3 - Session 6C - AIM
Chair: To Be Announced
- Stability-Aware Graph Neural Networks For Functional NeuroimagingCláudia Campos, Daniel Agostinho, Marco Simões and Bruno Direito
- Musiquence AI Toolset: A Serious Game Ecosystem for Health ProfessionalsJoão Cupido, Luís Ferreira and Sergi Bermúdez I Badia
- Dealing with Long-Form Clinical Audio for Portuguese VeterinariansAlice Arnaud Mota, Carolina Gomes, Luisa Coheur and Paula Tilley
- Assessing Cognitive Performance in Language Models: A Framework for Cognitive Impairment SimulationJonathan Guerne, Hatem Ghorbel and Paul Cotofrei
11:10 - 11:40
Coffee Break
11:40 - 12:40
Keynote: Michael Schroeder, TU Dresden
12:40 - 13:00
Closing Session
13:00 - 14:00
