ReqInOne: A Large Language Model-Based Agent for Software Requirements Specification Generation

2025-08-13

AI Summary

Existing SRS auto-generation approaches either rely heavily on manual analysis or suffer from LLM hallucinations and poor controllability. To address these limitations, this paper proposes a modular generation framework that decomposes SRS construction into three sequential subtasks: document summarization, requirement extraction, and requirement classification. The framework employs customized, stepwise prompt engineering to orchestrate multiple foundation models—including GPT-4o, LLaMA 3, and DeepSeek-R1—mimicking the logical workflow of professional requirements engineers. This design significantly improves structural consistency and semantic fidelity of generated SRS artifacts. Experimental evaluation demonstrates that the generated SRS outperforms end-to-end baseline methods and junior engineers’ outputs in completeness and compliance with standards. Notably, the requirement classification component achieves performance on par with or exceeding current state-of-the-art models. Overall, this work establishes a novel paradigm for high-assurance, interpretable, and automated software requirements specification generation.

Technology Category

• Natural Language Processing: Generation
• Machine Learning: Large Multimodal Models (LMMs)
• Constraint Satisfaction and Optimization: Satisfiability Modulo Theories

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

Software Requirements Specification (SRS) is one of the most important documents in software projects, but writing it manually is time-consuming and often leads to ambiguity. Existing automated methods rely heavily on manual analysis, while recent Large Language Model (LLM)-based approaches suffer from hallucinations and limited controllability. In this paper, we propose ReqInOne, an LLM-based agent that follows the common steps taken by human requirements engineers when writing an SRS to convert natural language into a structured SRS. ReqInOne adopts a modular architecture by decomposing SRS generation into three tasks: summary, requirement extraction, and requirement classification, each supported by tailored prompt templates to improve the quality and consistency of LLM outputs. We evaluate ReqInOne using GPT-4o, LLaMA 3, and DeepSeek-R1, and compare the generated SRSs against those produced by the holistic GPT-4-based method from prior work as well as by entry-level requirements engineers. Expert evaluations show that ReqInOne produces more accurate and well-structured SRS documents. The performance advantage of ReqInOne benefits from its modular design, and experimental results further demonstrate that its requirement classification component achieves comparable or even better results than the state-of-the-art requirement classification model.

Problem

• Automating SRS generation to reduce manual effort and ambiguity
• Addressing hallucinations and controllability in LLM-based SRS methods
• Improving accuracy and structure of generated SRS documents

Innovation

• Modular architecture for structured SRS generation
• Tailored prompt templates for LLM quality
• Comparative evaluation with expert benchmarks