![[2602.22456] Automating the Detection of Requirement Dependencies Using Large Language Models](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2602.22456] Automating the Detection of Requirement Dependencies Using Large Language Models
Summary
This article presents LEREDD, a novel approach utilizing Large Language Models to automate the detection of requirement dependencies in software engineering, achieving high accuracy in classification tasks.
Why It Matters
As software systems grow increasingly complex, understanding requirement dependencies is crucial for effective development. This study highlights the potential of LLMs to streamline this process, offering a solution to a traditionally manual and error-prone task, thereby enhancing software quality and efficiency.
Key Takeaways
- LEREDD leverages LLMs to automate the detection of requirement dependencies.
- The approach achieved an accuracy of 0.93 and an F1 score of 0.84 in classification tasks.
- LEREDD outperforms existing methods, particularly in identifying fine-grained dependency types.
- An annotated dataset of 813 requirement pairs is provided for reproducibility.
- This research underscores the role of AI in enhancing software engineering practices.
Computer Science > Software Engineering arXiv:2602.22456 (cs) [Submitted on 25 Feb 2026] Title:Automating the Detection of Requirement Dependencies Using Large Language Models Authors:Ikram Darif, Feifei Niu, Manel Abdellatif, Lionel C. Briand, Ramesh S., Arun Adiththan View a PDF of the paper titled Automating the Detection of Requirement Dependencies Using Large Language Models, by Ikram Darif and 4 other authors View PDF HTML (experimental) Abstract:Requirements are inherently interconnected through various types of dependencies. Identifying these dependencies is essential, as they underpin critical decisions and influence a range of activities throughout software development. However, this task is challenging, particularly in modern software systems, given the high volume of complex, coupled requirements. These challenges are further exacerbated by the ambiguity of Natural Language (NL) requirements and their constant change. Consequently, requirement dependency detection is often overlooked or performed manually. Large Language Models (LLMs) exhibit strong capabilities in NL processing, presenting a promising avenue for requirement-related tasks. While they have shown to enhance various requirements engineering tasks, their effectiveness in identifying requirement dependencies remains unexplored. In this paper, we introduce LEREDD, an LLM-based approach for automated detection of requirement dependencies that leverages Retrieval-Augmented Generation (RAG) and In-Conte...
