Netflix Artwork Personalization via LLM Post-training

TL;DR

This paper introduces a method to personalize artwork recommendations on Netflix by post-training large language models, leading to improved user engagement through tailored visual content.

Contribution

It presents a novel approach of post-training LLMs to select personalized artwork for users, addressing user heterogeneity in preferences.

Findings

Achieved 3-5% improvement over Netflix's existing model.

Demonstrated effectiveness of LLM post-training for personalized visual recommendations.

Validated approach on a dataset of 110K data points with 5K user-title pairs.

Abstract

Large language models (LLMs) have demonstrated success in various applications of user recommendation and personalization across e-commerce and entertainment. On many entertainment platforms such as Netflix, users typically interact with a wide range of titles, each represented by an artwork. Since users have diverse preferences, an artwork that appeals to one type of user may not resonate with another with different preferences. Given this user heterogeneity, our work explores the novel problem of personalized artwork recommendations according to diverse user preferences. Similar to the multi-dimensional nature of users' tastes, titles contain different themes and tones that may appeal to different viewers. For example, the same title might feature both heartfelt family drama and intense action scenes. Users who prefer romantic content may like the artwork emphasizing emotional warmth between the characters, while those who prefer action thrillers may find high-intensity action scenes more intriguing. Rather than a one-size-fits-all approach, we conduct post-training of pre-trained LLMs to make personalized artwork recommendations, selecting the most preferred visual representation of a title for each user and thereby improving user satisfaction and engagement. Our experimental results with Llama 3.1 8B models (trained on a dataset of 110K data points and evaluated on 5K held-out user-title pairs) show that the post-trained LLMs achieve 3-5\% improvements over the Netflix production model, suggesting a promising direction for granular personalized recommendations using LLMs.