Llama-3.2-3B model achieves 92% accuracy in parsing blood donation requests

Researchers have developed the Cognitive Blood Request System (CBRS), a framework designed to efficiently filter and parse urgent blood donation requests from social media streams. This system utilizes a novel bilingual dataset of over 11,000 messages in Bengali and English, incorporating adversarial negatives to enhance robustness. CBRS achieves 99% accuracy in filtering and a 92% zero-shot accuracy in parsing using a LoRA finetuned Llama-3.2-3B model, significantly outperforming other large language models while reducing token usage. AI

影响 Improves efficiency and inclusivity for time-sensitive information extraction tasks on social media.

排序理由 This is a research paper detailing a new system and dataset for information extraction. [lever_c_demoted from research: ic=1 ai=1.0]

在 arXiv cs.CL 阅读 →

• arXiv
• Bengali
• English
• Gemini-2.0-Flash
• GPT-4o-mini
• Llama-3.2-3B

• 论文
• 其他

AI 生成摘要 · Google Gemini · 来自 1 个来源。 我们如何撰写摘要 →