Bachelor and Master Theses

The thesis is not available now. If you are interested in doing a thesis about model checking, large language model, and machine learning, please contact Rong Gu directly. 

------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

In 2022, generative Artificial Intelligence (AI) such as ChatGPT has exploded in popularity due to its remarkable performance of natural language processing, as well as a more interactive and user-friendly interface to leverage the power of AI. The technology has been successfully applied to various applications, such as text translation and code generation.

One promising application of generative AI in product engineering is to improve the quality of requirements. Very often, requirements are written in natural language and ambiguity is unavoidable. Ambiguous requirements introduce misunderstandings into product development, which could lead to higher development costs or even project failure, and therefore should be prevented. Although some tools exist to help the engineer detect and even fix ambiguous requirements, they tend to overlook the “intention” of the requirements and the inconsistencies thereafter. They also lack interactive means that can provide intuitive guidance to and take feedback from the engineer.  

This thesis aims to explore the use of generative AI technology in requirement engineering in the industrial context of train software design and development. In general, the student is asked to answer three questions in this thesis:

-        How can generative AI help to achieve more accurate ambiguity and inconsistency detection in requirements written in natural language?

-        How can generative AI create higher-quality requirements, given the detected ambiguity and inconsistency?

-        How can the tool provide better interaction with the engineer, both in giving guidance and receiving feedback?

Methods such as prompt engineering are going to be investigated. A tool prototype is expected as part of the results.

The work will be in collaboration with Alstom in Västerås. The student will visit Alstom in Västerås and work at Alstom’s premises regularly.

Please check the following link for more information: Click.

- Project management

- Experience in requirement engineering, including requirement extraction and analysis (course projects count)

- Programming ability

- Basic knowledge of ChatGPT

Alstom in västerås