Comprehensive Review of End-to-End Dependency Parsing with Auto-regressive Large Language Models
The result's identifiers
Result code in IS VaVaI
<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216208%3A11320%2F26%3A8MUEUWF4" target="_blank" >RIV/00216208:11320/26:8MUEUWF4 - isvavai.cz</a>
Result on the web
<a href="http://dx.doi.org/10.1007/978-3-031-96473-2_22" target="_blank" >http://dx.doi.org/10.1007/978-3-031-96473-2_22</a>
DOI - Digital Object Identifier
<a href="http://dx.doi.org/10.1007/978-3-031-96473-2_22" target="_blank" >10.1007/978-3-031-96473-2_22</a>
Alternative languages
Result language
angličtina
Original language name
Comprehensive Review of End-to-End Dependency Parsing with Auto-regressive Large Language Models
Original language description
This paper comprehensively reviews the application of large language models (LLMs) for dependency parsing, focusing on auto-regressive models such as LLaMA (Large Language Model Meta AI). Dependency parsing is a crucial task in natural language processing (NLP), essential for understanding the syntactic structure of sentences. This review traces the evolution of dependency parsing techniques and highlights the significant advancements brought by LLMs. We provide an in-depth analysis of LLaMA's performance, efficiency, and multilingual capabilities, comparing it with other state-of-the-art models like GPT-3 and BERT. Our findings reveal that LLaMA achieves state-of-the-art results with fewer computational resources, excels in multilingual contexts, and demonstrates robustness in parsing raw sentences. We also discuss the ethical considerations, challenges, and future directions in the field, providing valuable insights for researchers and practitioners. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.
Czech name
—
Czech description
—
Classification
Type
D - Article in proceedings
CEP classification
—
OECD FORD branch
10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)
Result continuities
Project
—
Continuities
—
Others
Publication year
2025
Confidentiality
S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů
Data specific for result type
Article name in the collection
Commun. Comput. Info. Sci.
ISBN
978-3-031-96472-5
ISSN
18650929
e-ISSN
—
Number of pages
18
Pages from-to
317-334
Publisher name
Springer Science and Business Media Deutschland GmbH
Place of publication
—
Event location
Bangalore
Event date
Jan 1, 2026
Type of event by nationality
WRD - Celosvětová akce
UT code for WoS article
—