H-RAG at SemEval-2026 Task 8: Hierarchical Parent-Child Retrieval for Multi-Turn RAG Conversations

TL;DR

H-RAG introduces a hierarchical retrieval-augmented generation approach for multi-turn conversational tasks, emphasizing parent-child document segmentation, hybrid retrieval, and parent-level evidence aggregation to improve response accuracy.

Contribution

The paper presents a novel hierarchical parent-child RAG pipeline that enhances multi-turn conversational retrieval and generation by separating fine-grained retrieval from context reconstruction.

Findings

Achieved an nDCG@5 score of 0.4271 on retrieval task.

Secured a harmonic mean score of 0.3241 on generation task.

Highlighted the importance of retrieval configuration and parent-level aggregation.

Abstract

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.