A Two Dimensional Feature Engineering Method for Relation Extraction

TL;DR

The paper tackles relation extraction (RE) when multiple relation instances overlap within a sentence by leveraging a two-dimensional (2D) sentence representation. It introduces a 2D feature engineering framework that injects carefully designed prior-knowledge features into the semantic plane, coupled with a feature-aware attention mechanism to connect these features with entities. The architecture comprises five modules: feature engineering, encoding, 2D sentence interaction, CFA, and classification, achieving state-of-the-art results on ACE05 Chinese, ACE05 English, and SanWen datasets and demonstrating strong ablations that validate the design choices. The work highlights that combining explicit prior knowledge with a 2D representation can rival large language models in tailored RE tasks and offers a publicly available implementation for reproducibility and further development.

Abstract

Transforming a sentence into a two-dimensional (2D) representation (e.g., the table filling) has the ability to unfold a semantic plane, where an element of the plane is a word-pair representation of a sentence which may denote a possible relation representation composed of two named entities. The 2D representation is effective in resolving overlapped relation instances. However, in related works, the representation is directly transformed from a raw input. It is weak to utilize prior knowledge, which is important to support the relation extraction task. In this paper, we propose a two-dimensional feature engineering method in the 2D sentence representation for relation extraction. Our proposed method is evaluated on three public datasets (ACE05 Chinese, ACE05 English, and SanWen) and achieves the state-of-the-art performance. The results indicate that two-dimensional feature engineering can take advantage of a two-dimensional sentence representation and make full use of prior knowledge in traditional feature engineering. Our code is publicly available at https://github.com/Wang-ck123/A-Two-Dimensional-Feature-Engineering-Method-for-Entity-Relation-Extraction

Figures (4)

• Figure 1: Middle: The structure of our model illustrated with an example input sentence.
• Figure 2: An example of Feature Injection
• Figure 3: Different methods of feature injection
• Figure 4: Visualization of semantic plane after feature engineering