Saved in:
Bibliographic Details
Main Authors: Gebre, Binyam, Ranta, Karoliina, Elzen, Stef van den, Kuiper, Ernst, Baars, Thijs, Heskes, Tom
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2402.16073
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866929265488953344
author Gebre, Binyam
Ranta, Karoliina
Elzen, Stef van den
Kuiper, Ernst
Baars, Thijs
Heskes, Tom
author_facet Gebre, Binyam
Ranta, Karoliina
Elzen, Stef van den
Kuiper, Ernst
Baars, Thijs
Heskes, Tom
contents 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 paper, we propose 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. We tested and deployed this method to personalise promotional items at Bol, one of the largest e-commerce platforms of the Netherlands and Belgium. The method enhanced customer engagement and experience, leading to a significant 4.9% uplift in conversions.
format Preprint
id arxiv_https___arxiv_org_abs_2402_16073
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Pfeed: Generating near real-time personalized feeds using precomputed embedding similarities
Gebre, Binyam
Ranta, Karoliina
Elzen, Stef van den
Kuiper, Ernst
Baars, Thijs
Heskes, Tom
Information Retrieval
Artificial Intelligence
Machine Learning
H.3.3
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 paper, we propose 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. We tested and deployed this method to personalise promotional items at Bol, one of the largest e-commerce platforms of the Netherlands and Belgium. The method enhanced customer engagement and experience, leading to a significant 4.9% uplift in conversions.
title Pfeed: Generating near real-time personalized feeds using precomputed embedding similarities
topic Information Retrieval
Artificial Intelligence
Machine Learning
H.3.3
url https://arxiv.org/abs/2402.16073