TaxoCritic: Exploring Credit Assignment in Taxonomy Induction with Multi-Critic Reinforcement Learning

Injy Sarhan, Bendegúz Toth, Pablo Mosteiro, Shihan Wang

Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

Abstract

Taxonomies can serve as a vital foundation for several downstream tasks such as information retrieval and question answering, yet manual construction limits coverage and full potential. Automatic taxonomy induction, particularly using deep Reinforcement Learning (RL), is underexplored in Natural Language Processing (NLP). To address this gap, we present TaxoCritic, a novel approach that leverages deep multi-critic RL agents for taxonomy induction while incorporating credit assignment mechanisms. Our system uniquely assesses different sub-actions within the induction process, providing a granular analysis that aids in the precise attribution of credit and blame. We evaluate the effectiveness of multi-critic algorithms in experiments regarding both accuracy and robustness performance in edge identification. By providing a detailed comparison with state-of-the-art models and highlighting the strengths and limitations of our method, we aim to contribute to the ongoing development of automatic taxonomy induction while exploring the usage of deep RL techniques in this field.

Original languageEnglish
Title of host publicationProceedings of the Workshop on DLnLD 2024
Subtitle of host publicationDeep Learning and Linked Data at LREC-COLING 2024 - Workshop Proceedings
EditorsGilles Serasset, Hugo Goncalo Oliveira, Giedre Valunaite Oleskeviciene
PublisherEuropean Language Resources Association (ELRA)
Pages14-30
Number of pages17
ISBN (Electronic)9782493814166
Publication statusPublished - 21 May 2024
Event2024 Workshop on Deep Learning and Linked Data, DLnLD 2024 - Torino, Italy
Duration: 21 May 2024 → …

Publication series

NameProceedings of the Workshop on DLnLD 2024: Deep Learning and Linked Data at LREC-COLING 2024 - Workshop Proceedings

Conference

Conference2024 Workshop on Deep Learning and Linked Data, DLnLD 2024
Country/TerritoryItaly
CityTorino
Period21/05/24 → …

Keywords

  • Actor-Critic
  • Credit Assignment
  • Reinforcement Learning
  • Taxonomy Induction

Fingerprint

Dive into the research topics of 'TaxoCritic: Exploring Credit Assignment in Taxonomy Induction with Multi-Critic Reinforcement Learning'. Together they form a unique fingerprint.

Cite this