D-REX: Dialogue Relation Extraction with Explanations
It addresses the lack of explainability in dialogue relation extraction, which is important for applications like conversational AI, though it is incremental as it builds on existing relation extraction methods.
The paper tackles the problem of cross-sentence relation extraction in multi-party dialogues by proposing D-REX, a model-agnostic framework that extracts explanations for relations using partially labeled data, achieving a 13.5% improvement in F1 score over existing methods and having its explanations preferred by human annotators 90% of the time.
Existing research studies on cross-sentence relation extraction in long-form multi-party conversations aim to improve relation extraction without considering the explainability of such methods. This work addresses that gap by focusing on extracting explanations that indicate that a relation exists while using only partially labeled data. We propose our model-agnostic framework, D-REX, a policy-guided semi-supervised algorithm that explains and ranks relations. We frame relation extraction as a re-ranking task and include relation- and entity-specific explanations as an intermediate step of the inference process. We find that about 90% of the time, human annotators prefer D-REX's explanations over a strong BERT-based joint relation extraction and explanation model. Finally, our evaluations on a dialogue relation extraction dataset show that our method is simple yet effective and achieves a state-of-the-art F1 score on relation extraction, improving upon existing methods by 13.5%.