Let It Go? Not Quite: Addressing Item Cold Start in Sequential Recommendations with Content-Based Initialization

TL;DR

This paper introduces a novel method for cold-start in sequential recommenders by adding a small trainable adjustment to frozen content embeddings, balancing adaptation and semantic integrity, leading to improved performance.

Contribution

The paper proposes a new approach that combines frozen content embeddings with a trainable delta, enhancing cold-start item recommendations without losing semantic meaning.

Findings

Consistent improvements across multiple datasets.

Effective in e-commerce and music recommendation scenarios.

Balances adaptation and semantic preservation.

Abstract

Many sequential recommender systems suffer from the cold start problem, where items with few or no interactions cannot be effectively used by the model due to the absence of a trained embedding. Content-based approaches, which leverage item metadata, are commonly used in such scenarios. One possible way is to use embeddings derived from content features such as textual descriptions as initialization for the model embeddings. However, directly using frozen content embeddings often results in suboptimal performance, as they may not fully adapt to the recommendation task. On the other hand, fine-tuning these embeddings can degrade performance for cold-start items, as item representations may drift far from their original structure after training. We propose a novel approach to address this limitation. Instead of entirely freezing the content embeddings or fine-tuning them extensively, we introduce a small trainable delta to frozen embeddings that enables the model to adapt item representations without letting them go too far from their original semantic structure. This approach demonstrates consistent improvements across multiple datasets and modalities, including e-commerce datasets with textual descriptions and a music dataset with audio-based representation.