Effective Diversification of Multi-Carousel Book Recommendation
Daniël Wilten, Gideon Maillette de Buy Wenniger, Arjen Hommersom, Paul Lucassen, Emiel Poortman
TL;DR
The paper addresses diversifying book recommendations presented as multiple carousels on public library platforms, where implicit feedback yields sparse exact-ground-truth signals. It combines a LightGCN-based one-dimensional recommender with four two-dimensional carousel strategies to improve beyond-accuracy metrics such as diversity, serendipity, and novelty, while maintaining relevance via ABIS-based evaluation. The contributions include a practical two-dimensional prototype for libraries, a suite of item-selection strategies (diversity, serendipity, novelty, and their combination), and evaluation metrics tailored to sparse, implicit data. The findings demonstrate that the combined strategy offers a favorable trade-off, enhancing diversity, serendipity, and novelty concurrently, suggesting a viable path for deploying more engaging library catalogs in real-world settings.
Abstract
Using multiple carousels, lists that wrap around and can be scrolled, is the basis for offering content in most contemporary movie streaming platforms. Carousels allow for highlighting different aspects of users' taste, that fall in categories such as genres and authors. However, while carousels offer structure and greater ease of navigation, they alone do not increase diversity in recommendations, while this is essential to keep users engaged. In this work we propose several approaches to effectively increase item diversity within the domain of book recommendations, on top of a collaborative filtering algorithm. These approaches are intended to improve book recommendations in the web catalogs of public libraries. Furthermore, we introduce metrics to evaluate the resulting strategies, and show that the proposed system finds a suitable balance between accuracy and beyond-accuracy aspects.