Name: 18th Recommender Systems Netherlands meetup at JADS
Start: 2024-04-25T17:30:00+02:00
End: 2024-04-25T20:30:00+02:00
Location: Universiteit Jheronimus Academy of Data Science

After a long break, RecSysNL is back to live!

We look forward to welcoming you to the 18th RecSysNL meetup, hosted by the beautiful JADS in Den Bosch. The doors open at 17:30 and, the talks start at 18:15. For the planning details, just scroll down below.

**This time we'll have the great opportunity of having Alan Said (University of Gothenburg:** [www.alansaid.com](www.alansaid.com) **) giving a keynote.** Alan is one of the very first organizers of the recsysnl who kicked off these meetups in the Netherlands in CWI.

Keynote: **Recommender systems and trust in e-commerce**

This talk explores the role of recommender systems in building consumer trust within the fashion e-commerce industry. With the Trust Building Model (TBM) as a foundation, it demonstrates how aspects such as recommendation quality, user experience, and authenticity can affect trust. Highlighting the significance of competence belief in the recommendations provided, the talk suggests strategies for improving the effectiveness and reliability of these systems. Through actionable insights on personalizing user experiences and ensuring the authenticity of recommendations, the discussion aims to show how recommender systems can be utilized to establish stronger, trust-based relationships with consumers in the fashion e-commerce sector.

Next to this keynote, we'll have two other interesting talks from **bol.com** and **nobl.ai**:

Talk 1: **Pfeed: Generating near real-time personalized feeds using precomputed embedding similarities by Binyam Gebre**

In personalized recommender systems, embeddings are often used to encode customer actions and items, and retrieval is then performed in the embedding space using approximate nearest neighbor search. However, this approach can lead to two challenges: 1) user embeddings can restrict the diversity of interests captured and 2) the need to keep them up-to-date requires an expensive, real-time infrastructure. In this talk, I will present a method that overcomes these challenges in a practical, industrial setting. The method dynamically updates customer profiles and composes a feed every two minutes, employing precomputed embeddings and their respective similarities. The method has been tested and deployed to personalize promotional items at Bol. It improved engagement and increased conversions by 4.9%.

Bio: **Binyam Gebre** has a background in computer science and artificial intelligence. Since earning his PhD in 2015, he has worked on various industry projects involving natural language processing and computer vision. Currently, he is working as a lead data scientist at Bol, focusing on applying deep learning to recommender and search systems. Prior to his current role at Bol, he worked as a deep learning scientist for Philips Research in Eindhoven.

Talk 2: **The Nobl solution: Combining probabilistic modelling and session-based recommendations for e-recruitment by Alexandru Mara**
Recommender systems for the labour market face a number of unique challenges that range from data sourcing and quality issues to overwhelming amounts of (unstructured) textual information and interaction data, job-seeker and recruiter preferences, inherent knowledge and skills related to different occupations, etc. Simultaneously, the high-risk classification of these types of AI systems implies that higher levels of explainability and trustworthiness are expected. In this talk, we will explore the modular hybrid recommender system proposed by Nobl.ai to address these challenges. The modular structure ensures trustworthiness and explainability while the probabilistic representation learning model at its core can identify and compensate for any biases in the data. Simultaneously, a session-based module ensures real-time adaptation to user inputs.

Bio: **Alexandru Mara** is the Co-Founder and CEO of Nobl.ai, a tech startup focused on delivering trustworthy AI for the labour market. Simultaneously, Alexandru holds a part-time appointment as a postdoctoral researcher in the Artificial Intelligence and Data Analytics group at Ghent University. His research interests revolve around recommender systems and representation learning as well as their applications to the labour market. Over the past decade, Alexandru has built an international academic career in AI and data mining obtaining his PhD from Ghent University in Belgium, MSc from Aalto University in Finland and BSc from the Technical University of Madrid in Spain. He was also part of the ERC projects ONTIC and COGNET.

17:30 Doors open, pizzas
18:15: Welcome and introduction by Martijn Willemsen (5 min)
18:20-18:30: About the REM lab and DPG Media by Martijn Willemsen (10min)
18:30-19:00: Keynote Alan Said with Q&A
19:00-19:25: 1st talk by bol.com
19:25-19:50: 2nd talk by noble.ai
19:50-20:45: Drinks

Soode Fazeli

Martijn Willemsen

Babak Loni

Feng

Manel Slokom

Recommender Systems Netherlands - RecSysNL

Technology

Artificial Intelligence

Data Science

Data Analytics

Machine Learning

Big Data

New Technology

Education & Technology

18th Recommender Systems Netherlands meetup at JADS

Universiteit Jheronimus Academy of Data Science

Delen

Recommender Systems Netherlands - RecSysNL

18th Recommender Systems Netherlands meetup at JADS

Recommender Systems Netherlands - RecSysNL

Details

Leden zijn ook geïnteresseerd in