Re-Representation in Sentential Relation Extraction with Sequence Routing Algorithm

2025-08-28

AI Summary

This work addresses two key challenges in sentence-level relation extraction (RE): label noise inherent in distantly supervised data and insufficient semantic representation reconstruction. To this end, we propose a sequence routing algorithm based on dynamic routing in capsule networks. Inspired by the neuroscience concept of “re-representation”, we introduce a contrastive, noise-robust learning framework that jointly integrates dynamic routing with label quality analysis to enable adaptive optimization of relation representations. Our method achieves significant improvements over state-of-the-art approaches on benchmark datasets including TACRED and ReTACRED. Empirical analysis reveals the substantial performance degradation induced by Wikidata label noise and, for the first time, establishes a strong positive correlation between re-representation capability and model performance—thereby proposing a novel paradigm for representation learning in relation extraction.

Technology Category

• Reasoning under Uncertainty: Relational Probabilistic Models
• Machine Learning: Representation Learning
• Natural Language Processing: Sentence-level Semantics, Textual Inference, etc.

Application Category

• Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphs
• Search and Retrieval-Augmented AI: Web query analysis, representation and understanding
• Semantics and Knowledge: Representation, semantic annotation, enhancement, enrichments, access and/or integration of a variety of data on the Web

Abstract

Sentential relation extraction (RE) is an important task in natural language processing (NLP). In this paper we propose to do sentential RE with dynamic routing in capsules. We first show that the proposed approach outperform state of the art on common sentential relation extraction datasets Tacred, Tacredrev, Retacred, and Conll04. We then investigate potential reasons for its good performance on the mentioned datasets, and yet low performance on another similar, yet larger sentential RE dataset, Wikidata. As such, we identify noise in Wikidata labels as one of the reasons that can hinder performance. Additionally, we show associativity of better performance with better re-representation, a term from neuroscience referred to change of representation in human brain to improve the match at comparison time. As example, in the given analogous terms King:Queen::Man:Woman, at comparison time, and as a result of re-representation, the similarity between related head terms (King,Man), and tail terms (Queen,Woman) increases. As such, our observation show that our proposed model can do re-representation better than the vanilla model compared with. To that end, beside noise in the labels of the distantly supervised RE datasets, we propose re-representation as a challenge in sentential RE.

Problem

• Improving sentential relation extraction with dynamic capsule routing
• Investigating performance variations across different RE datasets
• Addressing label noise and re-representation challenges in RE

Innovation

• Dynamic routing in capsules
• Re-representation for similarity enhancement
• Noise identification in Wikidata labels