一句话：用三个互补 LLM 组成集成，通过结构化领域模型与并行完整性检查检测并修复 SRS 中的需求遗漏。
软件需求规格说明（SRS）中的需求遗漏会带来高昂的下游修复成本，而单个 LLM 的检测能力有限。本文提出一个工具，将 DeepSeek Chat、GPT-4o Mini 与 Claude Sonnet 4 三个互补 LLM 组成集成：先生成结构化领域模型，再以定制提示并行执行外部与内部完整性检查，发现缺失需求并给出补救建议，用户可选参与模型与聚合方式（多数、加权、元融合）。作者在四个不同领域的 SRS 上注入人工遗漏，系统比较各聚合策略：单模型召回率仅 0–52%，集成稳定超过 75% 乃至 100%，建议合理性达 95–100%。这表明多 LLM 集成可辅助而非替代分析师，支持交互式需求精化。
Ensuring complete software requirements specifications (SRS) is critical to preventing costly downstream errors. We present a tool that ensembles three complementary LLMs—DeepSeek Chat, GPT-4o Mini, and Claude Sonnet 4—to detect and suggest remedies for missing requirements. The tool generates a structured domain model and applies parallel external and internal completeness checks through tailored prompts. Users can select LLMs and aggregation methods (majority, weighted, or meta-fusion). Unlike prior single-model approaches, we systematically evaluate aggregation strategies across four diverse SRS domains. In experiments with seeded omissions, single models achieved only 0–52% recall, whereas our ensemble consistently exceeded 75%—reaching up to 100%—with 95–100% plausibility. These results demonstrate the feasibility of multi-LLM ensembles as practical aids—complementing rather than replacing human analysts—and supporting interactive refinement workflows.
DOI 原文 ·
@inproceedings{paperbot9379,
title = {Detecting and Repairing Incomplete Software Requirements with Multi-LLM Ensembles},
author = {Mohamad Kassab and Marwan AbdElhameed},
booktitle = {ASE},
year = {2025},
doi = {10.1109/ase63991.2025.00328}
}