Neuro-Symbolic Methods in Natural Language Processing: A Review

Mst Shapna Akter, Md Fahim Sultan, Alfredo Cuzzocrea, Alfredo Cuzzocrea

2025

Abstract

Neuro-Symbolic (NeSy) techniques in Natural Language Processing (NLP) combine the strengths of neural network-based learning with the clear interpretability of symbolic methods. This review paper explores recent advancements in neurosymbolic NLP methods. We carefully highlight the benefits and drawbacks of different approaches in various NLP tasks. Additionally, we support our evaluations with explanations based on theory and real-world evidence. Based on our review, we suggest several potential research directions. Our study contributes in three main ways: (1) We present a detailed, complete taxonomy for the Neuro-Symbolic methods in the NLP field; (2) We provide theoretical insights and comparative analysis of the Neuro-Symbolic methods; (3) We propose future research directions to explore.

Paper Citation

in Harvard Style

Akter M., Sultan M. and Cuzzocrea A. (2025). Neuro-Symbolic Methods in Natural Language Processing: A Review. In Proceedings of the 14th International Conference on Data Science, Technology and Applications - Volume 1: DATA; ISBN 978-989-758-758-0, SciTePress, pages 274-282. DOI: 10.5220/0013453100003967

in Bibtex Style

@conference{data25,
author={Mst Akter and Md Sultan and Alfredo Cuzzocrea},
title={Neuro-Symbolic Methods in Natural Language Processing: A Review},
booktitle={Proceedings of the 14th International Conference on Data Science, Technology and Applications - Volume 1: DATA},
year={2025},
pages={274-282},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0013453100003967},
isbn={978-989-758-758-0},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 14th International Conference on Data Science, Technology and Applications - Volume 1: DATA
TI - Neuro-Symbolic Methods in Natural Language Processing: A Review
SN - 978-989-758-758-0
AU - Akter M.
AU - Sultan M.
AU - Cuzzocrea A.
PY - 2025
SP - 274
EP - 282
DO - 10.5220/0013453100003967
PB - SciTePress