SM451/CS710: Seminar on Fairness in Recommendation and Ranking (HWS 2026)

Ranking and recommendation systems govern access to information by determining which candidates are shortlisted for recruiters, which sellers are found at the top of a marketplace, and which artists are heard. There is no single definition of fairness in information access. Position determines exposure, so small differences in scores can have a large impact on visibility. At the same time, providers, users, and platform compete for that exposure. Personalization of results means that there is no single output to inspect, and the user responses feeds back into the next model. Satisfying one notion of fairness can therefore mean violating another.

This seminar asks for whom a ranking is fair. It examines how fairness is defined and measured, how exposure is allocated among competing stakeholders, and how industry applies these approaches.

The schedule contains all relevant deadlines of this seminar. Schedule and room tba.

• This seminar is organized by Dr. Andreea Iana.
• Available for up to 8 Master students (2 SWS, 4 ECTS) and up to 4 Bachelor students (2 SWS, 5 ECTS).
• Prerequisites: Solid background in machine learning (MSc students), Wirtschaftsinformatik IV (BSc students)

In this seminar, you will

• Read, understand, and explore scientific literature
• Summarize a current research topic in a concise report (10–15 single-column pages + references), using the Latex template
• Deliver two presentations about your topic (3 minutes flash presentation, 15 minutes final presentation)
• Moderate a scientific discussion about the topic of one of your fellow students
• Give constructive feedback on fellow students' report drafts

• Please register via Portal² until the official deadline.
• Note that this seminar is an in person event, so that you should only select it if you are planning to be in Mannheim in the fall term
• If you are accepted into the seminar, attend the kickoff (see schedule) and provide at least 4 topic areas of interest of your preference (your own and / or example topic areas; see below) via email to Ezgi Yilmaz until the deadline (see schedule). 
• The actual topic assignment takes place soon afterwards; we will notify you via email. Our goal is to assign one of your preferred topic areas to you.

You will be assigned a topic area and a specific topic based on your preferences. You may propose your own topic of interest provided it aligns with the seminar's theme and objectives.

The report should demonstrate your understanding of the paper and of the area it belongs to, with an emphasis on critical synthesis and contextualization. In the report, you should:

1. Provide a concise overview of the topic area (1/4). Here you are expected to discuss the main goals, challenges, and relevance of the area. A survey paper, book chapter, or recent research paper are good starting points. 
2. Present one topic within the area in depth (3/4). Here you are expected to explain the paper's main contributions, position it against related work, and provide your own critical assessment. A recent of highly influential paper is a good place to start.

Topics that are suited just for BSc or MSc students are marked as such.

Topic areas

• Defining and measuring fairness
• Item- and provider-side fairness
• Consumer- and user-side fairness
• Multi-stakeholder fairness
• Fairness in practice
• Fairness in LLM-based recommendation and RAG