Our Aim

Recommender systems are among the most widely deployed applications of machine learning and influence what people see and choose. Traditional recommender systems rely on a stable user profile learned from past behavior. In many real settings, however, users are not logged in, their needs depend on the situation, and they describe what they need in words rather than as product attributes. Large language models make it possible to elicit such needs in conversation, but their output is not always grounded in the actual catalog, they can reproduce and amplify bias, and their effects often show only over time.

We want to advance recommender systems that model the users’ situated and evolving needs, elicit preferences through grounded, bias-aware conversation, and are evaluated responsibly over time and across stakeholders.

At the CDL-RecSys at TU Wien, we address these questions together with partners from industry and research. Our work is organized around three topics, interaction, dynamics and responsibility, and is grounded in Digital Humanism.

Partners

YKMB GmbH

YKMB GmbH is a newly established startup on a mission to disrupt online fashion retail. It has been co-founded in late 2019 in Vienna by an experienced team of marketing, sales and computer science experts who realized that something very essential is missing in purchasing online: The social experience of shopping together with your friends, which provide instant feedback on what suits you and what to get. The results of the CD Laboratory will be used to improve the current product and help to develop new ones.

Preisvergleich Internet Services AG

Preisvergleich Internet Services AG hosts the price comparison plattform Geizhals, which operates in Austria, Germany, the UK and Poland. Geizhals is designed to streamline the shopping process by combining their advanced data analytics and extensive product database. The goal is to empower users to make informed purchasing decisions by effortlessly comparing prices from multiple online retailers. This service optimizes the shopping experience, enabling users to save both time and money in the process of finding the right product at the right place. The results of the CD Laboratory will be used to increase the knowledge about the customer preferences and introduce strategies to elicit them.

Industry Collaborations