This study describes the model design of the NYCU-NLP system for the SemEval-2024 Task 2 that focuses on natural language inference for clinical trials. We aggregate several large language models to determine the inference relation (i.e., entailment or contradiction) between clinical trial reports and statements that may be manipulated with designed interventions to investigate the faithfulness and consistency of the developed models. First, we use ChatGPT v3.5 to augment original statements in training data and then fine-tune the SOLAR model with all augmented data. During the testing inference phase, we fine-tune the OpenChat model to reduce the influence of interventions and fed a cleaned statement into the fine-tuned SOLAR model for label prediction. Our submission produced a faithfulness score of 0.9236, ranking second of 32 participating teams, and ranked first for consistency with a score of 0.8092.

This study describes the model design of the NYCU-NLP system for the EXALT shared task at the WASSA 2024 workshop. We instruction-tune several large language models and then assemble various model combinations as our main system architecture for cross-lingual emotion and trigger detection in tweets. Experimental results showed that our best performing submission is an assembly of the Starling (7B) and Llama 3 (8B) models. Our submission was ranked sixth of 17 participating systems for the emotion detection subtask, and fifth of 7 systems for the binary trigger detection subtask.

This paper describes our proposed system design for the WASSA 2023 shared task 1. We propose a unified architecture of ensemble neural networks to integrate the original RoBERTa transformer with two sentiment-enhanced RoBERTa-Twitter and EmoBERTa models. For Track 1 at the speech-turn level, our best submission achieved an average Pearson correlation score of 0.7236, ranking fourth for empathy, emotion polarity and emotion intensity prediction. For Track 2 at the essay-level, our best submission obtained an average Pearson correlation score of 0.4178 for predicting empathy and distress scores, ranked first among all nine submissions.

This study describes the model design of the NCUEE-NLP system for the Chinese track of the SemEval-2022 MultiCoNER task. We use the BERT embedding for character representation and train the BiLSTM-CRF model to recognize complex named entities. A total of 21 teams participated in this track, with each team allowed a maximum of six submissions. Our best submission, with a macro-averaging F1-score of 0.7418, ranked the seventh position out of 21 teams.

This study describes our proposed system design for the SMM4H 2022 Task 8. We fine-tune the BERT, RoBERTa, ALBERT, XLNet and ELECTRA transformers and their connecting classifiers. Each transformer model is regarded as a standalone method to detect tweets that self-reported chronic stress. The final output classification result is then combined using the majority voting ensemble mechanism. Experimental results indicate that our approach achieved a best F1-score of 0.73 over the positive class.