UniRE: A Unified Label Space for Entity Relation Extraction
This addresses inefficiencies in joint extraction models for natural language processing, though it is incremental as it builds on existing methods by unifying label spaces.
The paper tackles the problem of joint entity and relation extraction by proposing a unified label space to improve information interaction between the two sub-tasks, resulting in a model that uses half the parameters, achieves competitive accuracy on benchmarks like ACE04, ACE05, and SciERC, and is faster.
Many joint entity relation extraction models setup two separated label spaces for the two sub-tasks (i.e., entity detection and relation classification). We argue that this setting may hinder the information interaction between entities and relations. In this work, we propose to eliminate the different treatment on the two sub-tasks' label spaces. The input of our model is a table containing all word pairs from a sentence. Entities and relations are represented by squares and rectangles in the table. We apply a unified classifier to predict each cell's label, which unifies the learning of two sub-tasks. For testing, an effective (yet fast) approximate decoder is proposed for finding squares and rectangles from tables. Experiments on three benchmarks (ACE04, ACE05, SciERC) show that, using only half the number of parameters, our model achieves competitive accuracy with the best extractor, and is faster.