GRASP: Guiding model with RelAtional Semantics using Prompt for Dialogue Relation Extraction

The dialogue-based relation extraction (DialogRE) task aims to predict the relations between argument pairs that appear in dialogue. Most previous studies utilize fine-tuning pre-trained language models (PLMs) only with extensive features to supplement the low information density of the dialogue by multiple speakers. To effectively exploit inherent knowledge of PLMs without extra layers and consider scattered semantic cues on the relation between the arguments, we propose a Guiding model with RelAtional Semantics using Prompt (GRASP). We adopt a prompt-based fine-tuning approach and capture relational semantic clues of a given dialogue with 1) an argument-aware prompt marker strategy and 2) the relational clue detection task. In the experiments, GRASP achieves state-of-the-art performance in terms of both F1 and F1c scores on a DialogRE dataset even though our method only leverages PLMs without adding any extra layers.

PDF Abstract COLING 2022 PDF COLING 2022 Abstract
Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Dialog Relation Extraction DialogRE GRASP_Large F1 (v1) 75.1 # 1
F1c (v1) 66.7 # 1
F1c (v2) 67.8 # 1
F1 (v2) 75.5 # 1
Dialog Relation Extraction DialogRE GRASP_Base F1 (v1) 69.2 # 2
F1c (v1) 62.4 # 2
F1c (v2) 61.7 # 5
F1 (v2) 69.0 # 4
Emotion Recognition in Conversation EmoryNLP GRASP_Large Weighted-F1 40.0 # 2
Emotion Recognition in Conversation MELD GRASP_Large Weighted-F1 65.6 # 10

Methods


No methods listed for this paper. Add relevant methods here