System improves extraction of breeding details from scientific texts
Retrieval-Augmented Multi-Prompt Ensemble for Minor-Grain Breeding Information Extraction
Computation and Language
Summary
Extracting specific information from scientific papers about minor grains is a difficult task because it involves many types of details and their connections. The researchers created a system called RAME that uses smart search combined with different question styles to get better answers from language models. Their system avoids extra training by combining many answers and picking the most common ones to increase accuracy. It performed better than previous methods in identifying breeding information from texts. Their approach helps make extracting complex scientific data easier and more reliable.
information extractionnamed entity recognition (NER)relation extraction (RE)large language models (LLMs)few-shot learningretrieval augmentationmajority votingBM25 retrievalprompt engineeringensemble methods
Authors
Hang Zhao, Jiahao Wang
Abstract
This paper presents our system for CCL2026-Eval Task 5: Minor-Grain Breeding Information Extraction (MGBIE), which jointly extracts 12 entity types and 6 relation types from minor-grain breeding literature. We propose RAME (Retrieval-Augmented Multi-Prompt Ensemble), a training-free framework that elicits multiple LLM outputs under controlled diversity and aggregates them by majority voting to obtain high-confidence predictions. RAME combines (i) retrieval-augmented few-shot selection via a hybrid BM25-embedding retriever, (ii) a three-prompt ensemble (Strict, Relaxed, Balanced) spanning the precision to recall spectrum, and (iii) large-scale repeated sampling with majority voting to filter noisy predictions. Built on DeepSeek-V4-Flash, RAME achieves a Total Score of 0.499 (NER 0.730, RE 0.346) on the leaderboard, ranking 1st and surpassing the official Track-A baseline powered by GPT-5.5 (0.448), representing an 11.4% relative improvement. Code is available at https://github.com/king-wang123/CCL26-RAME.