RECIPER: A Dual-View Retrieval Pipeline for Procedure-Oriented Materials Question Answering

2026-04-13

AI Summary

This study addresses the challenge that critical synthesis procedure information in materials science papers is often scattered across lengthy texts, making it difficult for paragraph-level dense retrieval to capture effectively. To overcome this limitation, the work proposes a dual-view retrieval framework that, for the first time, incorporates large language model–generated procedural summaries as complementary signals alongside original paragraphs in a joint index. A lightweight lexical re-ranker then fuses results from both views. Evaluated with BGE-large-en-v1.5, the approach achieves notable improvements of 3.73, 2.85, and 3.13 percentage points in Recall@1, nDCG@10, and MRR, respectively—reaching a Recall@1 of 86.82%—demonstrating significant gains in early-stage ranking performance and downstream question-answering effectiveness.

Technology Category

• Natural Language Processing: Summarization
• Data Mining & Knowledge Management: Conversational Systems for Recommendation & Retrieval
• Computer Vision: Image and Video Retrieval

Application Category

• Search and Retrieval-Augmented AI: Large language models for search
• Graph Algorithms and Modeling for the Web: Querying, indexing, and retrieval in Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Retrieving procedure-oriented evidence from materials science papers is difficult because key synthesis details are often scattered across long, context-heavy documents and are not well captured by paragraph-only dense retrieval. We present RECIPER, a dual-view retrieval pipeline that indexes both paragraph-level context and compact large language model-extracted procedural summaries, then combines the two candidate streams with lightweight lexical reranking. Across four dense retrieval backbones, RECIPER consistently improves early-rank retrieval over paragraph-only dense retrieval, achieving average gains of +3.73 in Recall@1, +2.85 in nDCG@10, and +3.13 in MRR. With BGE-large-en-v1.5, it reaches 86.82%, 97.07%, and 97.85% on Recall@1, Recall@5, and Recall@10, respectively. We further observe improved downstream question answering under automatic metrics, suggesting that procedural summaries can serve as a useful complementary retrieval signal for procedure-oriented materials question answering. Code and data are available at https://github.com/ReaganWu/RECIPER.

Problem

• procedure-oriented retrieval
• materials science
• evidence retrieval
• synthesis details
• dense retrieval

Innovation

• dual-view retrieval
• procedural summarization
• materials science QA
• dense retrieval
• lexical reranking