TY - GEN
T1 - Best of Both Worlds
T2 - 2024 Findings of the Association for Computational Linguistics: NAACL 2024
AU - Vacareanu, Robert
AU - Alam, Fahmida
AU - Islam, Md Asiful
AU - Riaz, Haris
AU - Surdeanu, Mihai
N1 - Publisher Copyright:
© 2024 Association for Computational Linguistics.
PY - 2024
Y1 - 2024
N2 - This paper introduces a novel neuro-symbolic architecture for relation classification (RC) that combines rule-based methods with contemporary deep learning techniques. This approach capitalizes on the strengths of both paradigms: the adaptability of rule-based systems and the generalization power of neural networks. Our architecture consists of two components: a declarative rule-based model for transparent classification and a neural component to enhance rule generalizability through semantic text matching. Notably, our semantic matcher is trained in an unsupervised domain-agnostic way, solely with synthetic data. Further, these components are loosely coupled, allowing for rule modifications without retraining the semantic matcher. In our evaluation, we focused on two few-shot relation classification datasets: Few-Shot TACRED and a Few-Shot version of NYT29. We show that our proposed method outperforms previous state-of-the-art models in three out of four settings, despite not seeing any human-annotated training data. Further, we show that our approach remains modular and pliable, i.e., the corresponding rules can be locally modified to improve the overall model. Human interventions to the rules for the TACRED relation org:parents boost the performance on that relation by as much as 26% relative improvement, without negatively impacting the other relations, and without retraining the semantic matching component.
AB - This paper introduces a novel neuro-symbolic architecture for relation classification (RC) that combines rule-based methods with contemporary deep learning techniques. This approach capitalizes on the strengths of both paradigms: the adaptability of rule-based systems and the generalization power of neural networks. Our architecture consists of two components: a declarative rule-based model for transparent classification and a neural component to enhance rule generalizability through semantic text matching. Notably, our semantic matcher is trained in an unsupervised domain-agnostic way, solely with synthetic data. Further, these components are loosely coupled, allowing for rule modifications without retraining the semantic matcher. In our evaluation, we focused on two few-shot relation classification datasets: Few-Shot TACRED and a Few-Shot version of NYT29. We show that our proposed method outperforms previous state-of-the-art models in three out of four settings, despite not seeing any human-annotated training data. Further, we show that our approach remains modular and pliable, i.e., the corresponding rules can be locally modified to improve the overall model. Human interventions to the rules for the TACRED relation org:parents boost the performance on that relation by as much as 26% relative improvement, without negatively impacting the other relations, and without retraining the semantic matching component.
UR - http://www.scopus.com/inward/record.url?scp=85191648921&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85191648921&partnerID=8YFLogxK
U2 - 10.18653/v1/2024.findings-naacl.165
DO - 10.18653/v1/2024.findings-naacl.165
M3 - Conference contribution
AN - SCOPUS:85191648921
T3 - Findings of the Association for Computational Linguistics: NAACL 2024 - Findings
SP - 2576
EP - 2594
BT - Findings of the Association for Computational Linguistics
A2 - Duh, Kevin
A2 - Gomez, Helena
A2 - Bethard, Steven
PB - Association for Computational Linguistics (ACL)
Y2 - 16 June 2024 through 21 June 2024
ER -