Probing Representations for Document-level Event Extraction
Submission Type: Regular Short Paper
Submission Track: Information Extraction
Submission Track 2: Interpretability, Interactivity, and Analysis of Models for NLP
Keywords: Information Extraction, Interpretability, Embedding, Probing, Document-Level Information Extraction
TL;DR: Probing event information and more from the encodings of documents used in IE.
Abstract: The probing classifiers framework has been employed for interpreting deep neural network models for a variety of natural language processing (NLP) applications. Studies, however, have largely focused on sentencelevel NLP tasks. This work is the first to apply the probing paradigm to representations learned for document-level information extraction (IE). We designed eight embedding probes to analyze surface, semantic, and event-understanding capabilities relevant to document-level event extraction. We apply them to the representations acquired by learning models from three different LLM-based document-level IE approaches on a standard dataset. We found that trained encoders from these models yield embeddings that can modestly improve argument detections and labeling but only slightly enhance event-level tasks, albeit trade-offs in information helpful for coherence and event-type prediction. We further found that encoder models struggle with document length and cross-sentence discourse.
Submission Number: 5300
Loading