Domain-specific fine-tuning of an LLM for NER-RE on human-smuggling court texts yields 15.5% and 31.46% absolute F1 gains over a larger baseline, with reduced noise, duplication, and runtime.
Relation extraction with fine-tuned large language models in retrieval augmented generation frameworks,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
FineREX: Fine-Tuned NER-RE for Human Smuggling Knowledge Graphs
Domain-specific fine-tuning of an LLM for NER-RE on human-smuggling court texts yields 15.5% and 31.46% absolute F1 gains over a larger baseline, with reduced noise, duplication, and runtime.