Local large language models to simplify requirement engineering documents in the automotive industry

Abstract

In automotive engineering, requirements management is crucial for determining the functional and technical qualities of a vehicle and ensuring reproducibility and uniformity throughout the development process. This paper presents a novel and innovative Local GPT Q&A retrieval solution for requirement engineering in the automotive industry. The study demonstrates that leveraging massive language models can significantly simplify the requirements analysis process, providing a more efficient and effective approach to handle complex requirement documents. The evaluation of various language models reveal their exceptional performance in answering evaluation questions, showcasing their potential for automating and enhancing requirement engineering tasks.

Keywords

• Large Language Models
• requirements engineering
• natural language processing