LLM-Based Dependency Parsing with Step-by-Step Instructions and a Simple Tabular Output

Q: What does this research mean for the field?

Combining step-by-step instruction prompts with a simplified tabular output improves the Labeled Attachment Score (LAS) of LLM-based dependency parsing across multiple languages compared to baseline methods. Novelty: ClaimNovelty.METHODOLOGICAL. Consensus alignment: ConsensusAlignment.NEUTRAL.

Hiroshi Matsuda; Chunpeng Ma; Masayuki Asahara

doi:10.5715/jnlp.33.920

January 1, 2026Journal of Natural Language ProcessingOpen Access

LLM-Based Dependency Parsing with Step-by-Step Instructions and a Simple Tabular Output

View Full Paper

Ask AI

Bookmark

Share

Key Points

This research aims to enhance dependency parsing accuracy using large language models (LLMs) by implementing step-by-step instructions.
Used large language models for dependency parsing;
Proposed a structured prompting approach that includes multi-step instructions;
Evaluated on data from 17 languages using CoNLL-U format.
Achieved Labeled Attachment Score (LAS) exceeding baseline by 0.4 points in 17 languages;
Outperformed baseline in LAS for 15 out of 17 languages;
No significant impact from test data contamination was found on parsing accuracy.

Implication

Demonstrates improved dependency parsing accuracy with LLMs in 17 languages, suggesting effective processing.

Abstract

No abstract yet. The journal hasn't published one we can show.

Discussion

No takes yet. Share an insight, caveat, or question.

Cite This Study

Matsuda et al. (2026) studied this question.

synapsesocial.com/papers/6a323915d50b63ecad204b6c https://doi.org/10.5715/jnlp.33.920

View Full Paper

Ask AI

Bookmark

Share

LLM-Based Dependency Parsing with Step-by-Step Instructions and a Simple Tabular Output

Key Points

Implication

Abstract

Discussion

Cite This Study

LLM-Based Dependency Parsing with Step-by-Step Instructions and a Simple Tabular Output

Key Points

Implication

Abstract

Discussion

Cite This Study

Keep reading

More related papers

More from Journal of Natural Language Processing

Keep reading

More related papers

More from Journal of Natural Language Processing