Scholarlink

This chapter examines how multi-modal data like text, image, audio, and videos can enhance recommendation systems. It introduces core integration strategies (early, late, and hybrid fusion) and contrasts them with emerging multi-modal large language models (LLMs) in terms of architecture, training, and use cases.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Leveraging Multi-modal Data

  • 章节

  • 发表日期：

    2025年

查看全文
  • Jianqiang Jay Wang

摘要

This chapter examines how multi-modal data like text, image, audio, and videos can enhance recommendation systems. It introduces core integration strategies (early, late, and hybrid fusion) and contrasts them with emerging multi-modal large language models (LLMs) in terms of architecture, training, and use cases.