Conv4Rec: A 1-by-1 Convolutional AutoEncoder for User Profiling through Joint Analysis of Implicit and Explicit Feedbacks

Antoine Ledent, Petr Kasalický, Rodrigo Alves, Hady W. Lauw · 2025

Read Conv4Rec: A 1-by-1 Convolutional AutoEncoder for User Profiling through Joint Analysis of Implicit and Explicit Feedbacks by Antoine Ledent, Petr Kasalický, Rodrigo Alves, Hady W. Lauw (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2509-07499.txt.

Introduction

Antoine Ledent and 3 co-authors introduce Conv4Rec in 2025, a convolutional autoencoder for profiling users in recommendation systems. Recommenders must learn from explicit ratings and from implicit feedback, the pattern of what users chose to view at all. Most models blur the two signals or use only one, losing information about items users never saw but might have loved.

The architecture learns associations between interaction types that carry over to every user and item. It predicts separately the chance a user consumes an item and the chance they rate it highly if they do. That split lets it surface items a user would not naturally encounter yet would enjoy. The authors also prove generalization bounds, among the first for autoencoders in recommendation, and show their loss recovers the exact sampling distribution under its assumptions.

Catalog record

CREATORAntoine Ledent, Petr Kasalický, Rodrigo Alves, Hady W. Lauw
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE167 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2509-07499

Browse the full library · About the Tower of Babel