What question did this study set out to answer?

The aim is to improve biomedical named entity recognition and relation extraction using advanced models and methodologies.

August 13, 2026Applied SciencesOpen Access

Biomedical Text Mining and Information Extraction Using Prompt-Enhanced and LoRA-Adapted Large Language Models

View Full Paper

Ask AI

Bookmark

Randomized trial demonstrates significant improvements in biomedical information extraction, indicating enhanced accuracy and efficiency.

The aim is to improve biomedical named entity recognition and relation extraction using advanced models and methodologies.
Developed a prompt-enhanced and QLoRA-adapted large language model framework for NER and RE.
Utilized a 4-bit quantized frozen backbone with the DeepSeek-R1-Distill-Qwen-7B model.
Conducted experiments on three NER and three RE datasets, employing seed-balanced paired bootstrap resampling.
Achieved a macro-average F1 of 0.909 ± 0.001 for NER and 0.787 ± 0.001 for RE across datasets.
Statistical analyses confirmed significant improvements over a simple-prompt baseline with adjusted p < 0.001.
Absolute F1 gains ranged from +0.091 to +0.131 across different datasets.

A 2026 study studied this question.

View Full Paper

Ask AI

Bookmark