Predicting emotions and emotional reactions during conversations and within texts poses challenges, even for advanced AI systems. The second iteration of the WASSA Empathy and Personality Shared Task focuses on creating innovative models that can anticipate emotional responses to news articles containing harmful content across four tasks.In this paper, we introduce our Fraunhofer SIT team’s solutions for the three tasks: Task 1 (CONVD), Task 2 (CONVT), and Task 3 (EMP).It involves combining LLM-driven data augmentation with fuzzy labels and fine-tuning RoBERTa models pre-trained on sentiment classification tasks to solve the regression problems. In the competition, our solutions achieved first place in Task 1, X in Task 2, and third place in Task 3.

This paper describes the system used to predict stances towards health orders and to detect premises in Tweets as part of the Social Media Mining for Health 2022 (SMM4H) shared task. It takes advantage of GPT-2 to generate new labeled data samples which are used together with pre-labeled and unlabeled data to fine-tune an ensemble of GAN-BERT models. First experiments on the validation set yielded good results, although it also revealed that the proposed architecture is more suited for sentiment analysis. The system achieved a score of 0.4258 for the stance and 0.3581 for the premise detection on the test set.