WISTERIA: Weak Implicit Signal-based Temporal Relation Extraction with Attention

TL;DR

WISTERIA introduces an attention-based framework for temporal relation extraction that focuses on pair-specific cues, improving interpretability and competitive accuracy across multiple datasets.

Contribution

It proposes a novel pair-conditioned top-K attention pooling method that isolates informative tokens for temporal relation classification, enhancing interpretability.

Findings

Achieves competitive accuracy on multiple datasets.

Provides interpretable rationales aligned with linguistic cues.

Demonstrates the effectiveness of pair-specific attention in TRE.

Abstract

Temporal Relation Extraction (TRE) requires identifying how two events or temporal expressions are related in time. Existing attention-based models often highlight globally salient tokens but overlook the pair-specific cues that actually determine the temporal relation. We propose WISTERIA (Weak Implicit Signal-based Temporal Relation Extraction with Attention), a framework that examines whether the top-K attention components conditioned on each event pair truly encode interpretable evidence for temporal classification. Unlike prior works assuming explicit markers such as before, after, or when, WISTERIA considers signals as any lexical, syntactic, or morphological element implicitly expressing temporal order. By combining multi-head attention with pair-conditioned top-K pooling, the model isolates the most informative contextual tokens for each pair. We conduct extensive experiments on TimeBank-Dense, MATRES, TDDMan, and TDDAuto, including linguistic analyses of top-K tokens. Results show that WISTERIA achieves competitive accuracy and reveals pair-level rationales aligned with temporal linguistic cues, offering a localized and interpretable view of temporal reasoning.