Abstract: Requirement understanding and analysis is the most difficult and time-consuming task in software projects. Whether it can be carried out effectively is directly related to the success or failure of the project. Due to the software engineers’ lack of domain knowledge in general, the knowledge gap is the biggest challenge faced by traditional methods that rely heavily on them for requirement understanding. Large language models are believed to have memorized various knowledge and it is natural to think that they seem to have the potential to bridge this gap. However, the free conversational style of large language models is not suitable for requirement elicitation. How to enable large language models to interact with users in a targeted, strategic, and natural way to efficiently determine the most relevant information remains a key challenge. This talk will delve into these challenges. Based on this, this talk will further outline the high-level knowledge and skills that a requirements agent needs to possess, in order to support the efficient acquisition of project requirements.

Bio: Zhi Jin, IEEE/CCF/CAAI Fellow, is Hongyi visiting professor at Wuhan University and professor of computer science at Peking University. Her main research interest is AI for SE, with a long-term focus on domain knowledge-led requirements engineering. She has published over 300 scientific articles in refereed international journals, such as IEEE T-KDE, T-SE, ACM T-OSEM, and T-CHI, and high rank conferences, such as ICSE, FSE, ASE, ACL and RE. She has co-authored five books and has held more than 30 approved invention patents. She is six times recipient of ACM SIGSOFT Distinguished Paper Awards.