Comparing Chunking and Embedding Strategies for Turkish RAG Systems
土耳其检索增强生成(RAG)系统的分块与嵌入策略比较
机构 * Data Science and Innovation Ata Technology Platforms(阿塔技术平台数据科学与创新部门)
专题命中 向量检索 :RAG(title,title_cn);retrieval-augmented generation(abstract);分类 cs.CL、cs.AI
AI总结 该研究针对土耳其语RAG系统,对比三种分块策略、五种嵌入模型和两种生成器LLM,发现布局感知分块可压缩嵌入模型差异,最优配置依内容类型而定,最佳单个组件无法组合出最优整体配置,准确率达87.0%。
Comments Accepted to INTCEC 2026. This is the author's pre-print version. The final authenticated version will be available through the conference proceedings