H-RAG在SemEval-2026任务8中的应用：多轮RAG对话中的层次父-子检索
H-RAG at SemEval-2026 Task 8: Hierarchical Parent-Child Retrieval for Multi-Turn RAG Conversations
AI总结：
H-RAG通过层次化父-子检索方法提升多轮RAG对话性能，结合密集-稀疏搜索与上下文重构，实现检索与生成的协同优化。
摘要：
我们提出了H-RAG，作为SemEval-2026任务8（MTRAGEval）的参赛方案，同时解决任务A（检索）和任务C（基于检索段落的生成）。任务A评估独立检索质量，而任务C评估多轮对话中端到端的检索增强生成（RAG）性能，需同时保证准确回答和忠实于检索证据。我们的方法实现了一种层次化父-子RAG流程，将细粒度子级检索与父级上下文重构分离。文档被分割为重叠的句子级子块，而完整文档作为父单元保留以提供连贯的上下文。检索结合混合密集-稀疏搜索、可调权重和基于嵌入的相似性重评分。检索证据在父级聚合并提供给指令微调的语言模型进行响应生成。H-RAG在任务A中达到nDCG@5得分为0.4271，在任务C中达到调和平均得分为0.3241（RB_agg: 0.2488，RL_F: 0.2703，RB_llm: 0.6508），凸显了检索配置和父级聚合在多轮RAG性能中的重要性。
英文摘要：
We present H-RAG, our submission to SemEval-2026 Task 8 (MTRAGEval), addressing both Task A (Retrieval) and Task C (Generation with Retrieved Passages). Task A evaluates standalone retrieval quality, while Task C assesses end-to-end retrieval-augmented generation (RAG) in multi-turn conversational settings, requiring both accurate answer generation and faithful grounding in retrieved evidence. Our approach implements a hierarchical parent-child RAG pipeline that separates fine-grained child-level retrieval from parent-level context reconstruction during generation. Documents are segmented into overlapping sentence-based child chunks, while full documents are preserved as parent units to provide coherent context. Retrieval combines hybrid dense-sparse search, tunable weighting, and embedding-based similarity rescoring over child chunks. Retrieved evidence is aggregated at the parent level and supplied to an instruction-tuned language model for response generation. H-RAG achieves an nDCG@5 score of 0.4271 on Task A and a harmonic mean score of 0.3241 on Task C (RB_agg: 0.2488, RL_F: 0.2703, RB_llm: 0.6508), underscoring the importance of retrieval configuration and parent-level aggregation in multi-turn RAG performance.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
