This year’s Nordic Personalisation Day, held on September 16-17 in Gothenburg, Sweden, brought together around 60 participants, mostly from Scandinavia, to explore the ethics and responsibilities of AI-driven recommendation systems. The event was a great opportunity for networking, knowledge sharing, and sparking fresh ideas.
SFI MediaFutures made a strong contribution with presentations from PhD candidates Jia-Hua Jeng and Bilal Mahmood, as well as from Associate Professors Mehdi Elahi and Alain D. Starke and Professor Dietmar Jannach. PhD candidate Khadiga Seddik also presented a poster, alongside Jeng and Mahmood.
Seddik’s poster showcased her upcoming research paper, while Alain Starke took on the role of panel moderator for a lively discussion on the main theme of the conference. The panel included industry experts such as Peter Knees, Marko Tkcalcic, Lubos from TV2, and Simen Buobb from VG.
The event kicked off with a keynote by Peter Knees, Associate Professor at TU Wien, who highlighted the need for academia to strengthen policies around research-industry collaborations. He emphasized the importance of universities prioritizing ethical research and ensuring their work serves the public interest.
The second keynote speaker, Martin Tegner, Principal Data Scientist at IKEA, shared insights into how IKEA uses data to shape online customer experiences, blending 80 years of home-furnishing expertise with the latest in AI technology.
Attendees mostly came from Norway and Sweden, with some participants from other parts of Europe. The event fostered deep discussions on topics such as polarization, long-term evaluation of systems, and the ethical considerations around recommendation algorithms.
Overall, it was a successful event, packed with meaningful conversations and ideas that could drive future collaborations and innovations.
We want to thank all organisers and sponsors of the event and look forward to next years gathering.
All pictures are taken by Alan Said – alansaid.com, CC BY-SA-NC 4.0 license.
Program
Monday, 16. September 2024
| Time | Title |
| 11:00-12:00 | Registration desk opens |
| 12:00-13:00 | Lunch |
| 13:00-13:10 | Opening and introduction |
| 13:10-14:20 | Keynote: Peter Knees – AI needs ethics, but we need to change the game. |
| 14:20-14:40 | Jia-Hua Jeng – Negativity Sells? Using an LLM to Affectively Reframe News Articles in a Recommender System |
| 14:40-15:00 | Benjamin Kille – Large Language Models and their Potential for Personalization |
| 15:00-15:20 | Christoph Schmitz – Content recommendation in a News Setting |
| 15:20-15:40 | Bilal Mahmood – A Supervised Machine Learning Approach for Supporting Editorial Article Selection |
| 15:40-15:45 | Poster Madness |
| 15:45-17:00 | Poster session and coffee break |
| 17:00-19:00 | Reception |
Tuesday, 17. September 2024
| Time | Title |
| 09:00-09:10 | Welcome |
| 09:10-09:30 | Dietmar Jannach – Understanding longitudinal effects of recommender systems. |
| 09:30-09:50 | Alan Said – The Carbon Footprint of Recommender Systems. |
| 09:50-10:10 | Lovisa Annerwall – Recommender systems and firm performance: A systematic review of the past to prepare for future developments |
| 10:10-10:30 | Alain Starke – A Primer on Normative Design and Thinking in Recommender Systems: A Brief Research Agenda for the Food Domain. |
| 10:30-11:00 | Coffee break |
| 11:00-11:50 | Keynote: Martin Tegner – Keynote: Bringing a life-at-home experience to you. |
| 11:50-12:10 | Mehdi Elahi – Recommender Systems and the Challenge of Popularity Bias: An Overview. |
| 12:10-12:30 | Marko Tkalcic – Optimizing Recommender Systems for Political Depolarization |
| 12:30-13:30 | Lunch |
| 13:30-15:00 | Panel |
| 15:00-15:30 | Closing |
| 15:30-16:00 | Coffee break and good bye |