Artificial Intelligence and Software Engineering

Course goals

After successfully completing this course, you

1. understand the theoretical aspects of the use of AI to support software and information systems engineering
2. are familiar with the broad landscape of AI techniques, from automated reasoning to machine and deep learning, that can be adopted
3. are able to design and execute an AI×SE research project in a rigorous manner
4. are able to report on an AI×SE research project according to state-of-the-art guidelines
5. are able to orally present research in a sub-field of AI×SE

We will henceforth refer to the above-mentioned learning objectives as LO1, …, LO5.

We aim to facilitate the attainment of these learning outcomes through an interactive and discussion-oriented seminar.
Active class participation is required to maximize your learning.

Assessment
The course assessment is based on five components:

A. regular and continued effort, via periodic project updates (individual) (20% of the final mark) [LO3, LO5]
B. oral presentation of a research paper (individual) (20%) [LO2, LO4, LO5]
C. research prototype and replication package (team) (20%) [L01, LO3, LO4]
D. research project pitch (team) (10%) [LO1, LO2, LO5]
E. final paper (team) (30%) [LO1, LO2, LO3, LO4]

Each partial grade needs to be at least 5.5.
The students who receive a grade below 5.5 for the prototypical research artefact (C) and/or final paper (E) will get a repair opportunity.
There is no repair option for the project updates (A), the oral presentations (B), and the research project pitch (D).

To qualify for a repair of the final result the mark needs to be at least a 4, or “AANV”.
The use of platforms for automated text generation based on user prompts like ChatGPT must be declared, and is allowed (given the declaration) for all assessment parts, except the final paper. Assisted code-production tools such as Co-Pilot is allowed but must be explicitly declared as well.

Content

Artificial Intelligence (AI) is revolutionizing the way software is being developed. Code generation is a prime example: AI assistants are significantly reducing the number of lines of code that human developers write. However, AI has the potential of being applied to many other software engineering activities, including requirements engineering, architecting, design, testing, and maintenance & evolution.

The course is centred around the theory and application of artificial intelligence techniques to support the construction of software, including information systems. We will refer to this intersection as Artificial Intelligence for Software Engineering (AI×SE).

The core of the course is the conduction of a small research project in the area of AI×SE.

Given its structure and aims, the course is designed for students enrolled in the Master’s programme in Business Informatics, also welcoming students in Artificial Intelligence, Game & Media Technology, and Data Science.

Course form

The overall aim of this course is to provide the students with foundations for conducting research that applies AI to aid in the development of software and information systems.

The course runs over nine weeks, where each week has its own theme, including ‘Agents in artificial intelligence and overview of AI techniques’, ‘Qualitative and quantitative evaluation of an AI×SE research artefact’, ‘Conduction of an AI×SE research project’.

The introductory weeks will illustrate the fundamental notions of artificial intelligence from a software and information systems engineering perspective. Thereupon, guidelines, criteria, and focus sessions on the conduction of an AI×SE research project weeks will be the main subject. Weeks dedicated to the presentation of research related to the students' project will follow. Project pitches by students will conclude the course's quarter. classes will be split in two sessions: a first one, in which lecturers or students will present and discuss scientific literature and methods; a second one, in which students will show the current course of action for their projects, including how the theories, techniques and approaches described in the literature can contribute to their own research endeavour.