An Unified Search and Recommendation Foundation Model for Cold-Start Scenario

TL;DR

This paper introduces a unified multi-domain foundation model leveraging large language models and novel fusion techniques to improve search and recommendation tasks, especially in cold-start scenarios, with successful deployment in Alipay.

Contribution

It proposes a novel framework combining LLM-based domain-invariant features and aspect gating fusion, enabling joint training across multiple search and recommendation domains for cold-start scenarios.

Findings

Outperforms state-of-the-art transfer learning methods in cold-start tasks.

Successfully deployed in Alipay's online services.

Enhances multi-domain search and recommendation performance.

Abstract

In modern commercial search engines and recommendation systems, data from multiple domains is available to jointly train the multi-domain model. Traditional methods train multi-domain models in the multi-task setting, with shared parameters to learn the similarity of multiple tasks, and task-specific parameters to learn the divergence of features, labels, and sample distributions of individual tasks. With the development of large language models, LLM can extract global domain-invariant text features that serve both search and recommendation tasks. We propose a novel framework called S\&R Multi-Domain Foundation, which uses LLM to extract domain invariant features, and Aspect Gating Fusion to merge the ID feature, domain invariant text features and task-specific heterogeneous sparse features to obtain the representations of query and item. Additionally, samples from multiple search and recommendation scenarios are trained jointly with Domain Adaptive Multi-Task module to obtain the multi-domain foundation model. We apply the S\&R Multi-Domain foundation model to cold start scenarios in the pretrain-finetune manner, which achieves better performance than other SOTA transfer learning methods. The S\&R Multi-Domain Foundation model has been successfully deployed in Alipay Mobile Application's online services, such as content query recommendation and service card recommendation, etc.