Applied Artificial Intelligence
Technische Hochschule Köln
Around 50 participants came to the TH Cologne and we had an intellectually well-filled evening with two really exciting presentations that offered a lot of interesting information from the world of LLM and personalizing carousel ranking.
Venue
Agenda
Fabian Haak – Research assistant at TH Köln:
Quantifying Subjective Phenomena: Simulation, Detection, and Synthetic Training Data with LLMs
“There are many aspects of different forms of media that are relevant to measure as subjective. Is a text easy to read? Is a comment offensive? Is a document relevant? Is a decision morally just? The inherent problem with measuring these aspects is that different humans would judge differently based on a range of often intransparent item properties, as well as the person’s personal experiences, attitudes, and preexisting knowledge. Despite that, effective means of measuring these aspects and creating training data representative of a typical user or a certain target audience are essential for ensuring quality, fairness, and relevance in media consumption and production. We at CIR, Cologne Information Retrieval Group at TH Köln, use a multi-stage approach leveraging large language models to quantify subjective phenomena and construct synthetic training data, that are on par with human-annotated datasets.”
Marcel Kurovski – Senior Data Scientist at Wolt (Steffen Klempau – ML Engineer at Wolt):
Personalizing Carousel Ranking on Wolt’s Discovery Page: A Hierarchical Multi-Armed Bandit Approach
“Personalized carousel ranking presents a major recommendation challenge across many domains like content streaming, ecommerce or quick commerce. We present a hierarchical multi-armed bandit (MAB) solution for personalizing the ranking of carousels on Wolt’s Discovery page. The Discovery page serves as the primary gateway for millions of weekly users exploring diverse cuisines and products. First, we illustrate the specific challenges of an (almost) everything online delivery platform and our goals for Wolt’s Discovery page. Second, we summarize how the problem of page personalization looks like in other domains and how others tackle it. We then present the approach and architecture of our Discovery Vertical Content Ranker (DVCR). Our approach hierarchically combines city-, context- and user-specific implicit feedback data to rank carousels on Wolt’s Discovery page. We illustrate the architecture to make this solution resilient, scalable and adaptive. We’re leveraging mlflow for tracking and lineage, Flyte for ML workflows, Redis for serving features, and Seldon Core for serving user requests online fast and reliably. We will wrap up with our learnings and an outlook.”
Networking
The two exciting presentations offered a lot of interesting information and questions to exchange ideas over hot pizza and cold drinks afterwards into the evening:
- Quantifying subjective phenomena: simulation, detection and synthetic training data with LLMs
- Personalizing carousel ranking on the discovery side of Wolt: A hierarchical multi-armed bandit approach
Videos
Here are some video impressions of the 31th Meetup, documenting the insightful presentations and lively discussions of the participants:
Dr. Cedric Reuter
Fabian Haak
Marcel Kurovski
Dr. Fabian Hadiji
Slides & publications
Fabian Haak
- Slides: de.slideshare.net/secret/amNm6szY5llbRW
- Google Schoolar: https://scholar.google.com/citations
- Publication: dl.acm.org/doi/abs/10.1145/3630744.3658415
Marcel Kurovski
Further Information
CAIML aims to bring together people interested in AI and machine learning. For more information visit & join us:
Retrospective
Don’t miss the previous meetups of the CAIML community in Cologne: