Papers by Sanda Harabagiu
Tree-of-Counterfactual Prompting for Zero-Shot Stance Detection (2024.acl-long)
Copied to clipboard
| Challenge: | Stance detection is used to infer attitudes from human communications . stance decisions involve complex judgments generated by LLMs . |
| Approach: | They propose a method for stance detection which relies on a new prompting framework . it allows for more than one stance object type and no examples of stance attribution . |
| Outcome: | The proposed method outperforms fine-tuned stance detection systems. |
VaccineLies: A Natural Language Resource for Learning to Recognize Misinformation about the COVID-19 and HPV Vaccines (2022.lrec-1)
Copied to clipboard
| Challenge: | VaccineLies can detect misinformation about vaccines on Twitter without using language resources. |
| Approach: | They present a dataset of tweets propagating misinformation about two vaccines . authors propose novel methods to detect misinformation on Twitter and identify stance towards it . |
| Outcome: | VaccineLies can detect misinformation on Twitter and identify the stance towards it. |
Identification of Multimodal Stance Towards Frames of Communication (2023.emnlp-main)
Copied to clipboard
| Challenge: | Until now, determining whether an author is in favor of, against or has no stance towards a frame was performed only when processing texts. |
| Approach: | They propose to use a dataset to infer stance towards 113 different frames of communication in multimodal documents. |
| Outcome: | The proposed model improved the quality of identifying multimedia stance by 20% compared to previous methods, which only performed when processing texts. |
The Language of Brain Signals: Natural Language Processing of Electroencephalography Reports (2020.lrec-1)
Copied to clipboard
| Challenge: | Clinical electroencephalography (EEG) is an excellent tool for probing neural function. |
| Approach: | They propose to use EEG to capture brain signals and its correlations with pathologies by a corpus of EEG reports to provide examples of EMG-specific concepts. |
| Outcome: | The proposed method provides examples of EEG-specific and clinically relevant concepts and exemplifies a self-attention joint-learning model to predict similar annotations in the EEG report corpus. |
Discovering and Articulating Frames of Communication from Social Media Using Chain-of-Thought Reasoning (2024.eacl-long)
Copied to clipboard
| Challenge: | a method for Discovering and Articulating FoCs is proposed . 86.2% of the FoC encoded by communication experts were also uncovered . |
| Approach: | They propose a method for Discovering and Articulating FoCs that uses Chain-of-Thought prompting and In-Context Active Curriculum Learning to uncover FoC. |
| Outcome: | The proposed method uncovered 86.72% of the FoCs encoded by communication experts on the same reference dataset. |