Grigorios Tsoumakas

Overview of ELCardioCC Task on Clinical Coding in Cardiology at BioASQ 2026

Loss-Based Active Learning for Neural Abstractive Summarization

This work proposes LOBSTER, a novel active learning framework designed specifically for abstractive summarization that improves performance by prioritizing unlabeled instances semantically similar to the model's current high-loss training examples, enabling the model to explicitly correct its specific weaknesses.

MAGNETIC: Multilayer Attention and Graph Neural Network Diffusion for Effective Drug-Target Interaction Prediction.

Results reveal that MAGNETIC consistently out performs baselines on both the area under the receiver operating characteristic curve (AUROC) and the area under the precision-recall curve (AUPRC), indicating improved identification and ranking of true interactions under class imbalance.