RecExplainer: Aligning Large Language Models for Explaining Recommendation Models

📄 arXiv: 2311.10947v2 📥 PDF

作者: Yuxuan Lei, Jianxun Lian, Jing Yao, Xu Huang, Defu Lian, Xing Xie

分类: cs.IR, cs.AI

发布日期: 2023-11-18 (更新: 2024-06-22)

备注: 12 pages, 9 figures, 5 tables

DOI: 10.1145/3637528.3671802

🔗 代码/项目: GITHUB


💡 一句话要点

提出RecExplainer以解决推荐模型黑箱解释问题

🎯 匹配领域: 支柱九:具身大模型 (Embodied Foundation Models)

关键词: 推荐系统 黑箱模型 大型语言模型 可解释性 个性化推荐 行为对齐 意图对齐

📋 核心要点

  1. 现有推荐模型通常作为黑箱运作,缺乏透明性,导致用户和开发者难以理解其决策过程。
  2. 本文提出使用大型语言模型作为替代模型,通过行为对齐、意图对齐和混合对齐三种方法来解释推荐模型。
  3. 在三个公共数据集上的实验结果显示,所提方法在理解和模拟目标模型方面表现优异,生成的解释质量高且具有独特性。

📝 摘要(中文)

推荐系统在在线服务中被广泛应用,嵌入式模型因其在表示复杂信号方面的表现力而受到青睐。然而,这些模型通常作为黑箱运作,缺乏透明性,给用户和开发者带来不便。本文首次探索使用大型语言模型(LLMs)作为替代模型来解释黑箱推荐模型。通过训练LLMs理解和模拟目标推荐模型的行为,利用其丰富的世界知识和多步推理能力,Aligned LLMs能够提供个性化的解释。为实现有效对齐,本文提出了行为对齐、意图对齐和混合对齐三种方法。综合实验表明,该方法在理解和模拟目标模型方面取得了良好效果,生成高质量且独特的解释。

🔬 方法详解

问题定义:本文旨在解决推荐模型的黑箱特性,现有方法缺乏透明性,难以为用户提供可解释的推荐理由。

核心思路:通过训练大型语言模型(LLMs)来理解和模拟目标推荐模型的行为,利用其推理能力生成个性化的解释。

技术框架:整体架构包括三个主要模块:行为对齐(在语言空间中模拟用户偏好和物品信息)、意图对齐(在潜在空间中理解模型行为)和混合对齐(结合语言和潜在空间)。

关键创新:最重要的创新在于将LLMs作为推荐模型的替代解释工具,利用其强大的推理能力和自然语言处理能力,显著提高了解释的质量和个性化程度。

关键设计:在设计中,采用了多种损失函数来优化对齐效果,并在网络结构上结合了语言模型和推荐模型的特征表示,以增强模型的理解能力。

🖼️ 关键图片

fig_0
fig_1
fig_2

📊 实验亮点

实验结果表明,所提方法在三个公共数据集上均表现优异,相较于基线模型,解释的质量和个性化程度显著提升,具体性能数据未明确提供,但整体表现令人满意。

🎯 应用场景

该研究的潜在应用领域包括电子商务、社交媒体和内容推荐等,能够为用户提供更透明和个性化的推荐解释,提升用户体验和信任度。未来,该方法可扩展到其他类型的黑箱模型解释中,具有广泛的实际价值。

📄 摘要(原文)

Recommender systems are widely used in online services, with embedding-based models being particularly popular due to their expressiveness in representing complex signals. However, these models often function as a black box, making them less transparent and reliable for both users and developers. Recently, large language models (LLMs) have demonstrated remarkable intelligence in understanding, reasoning, and instruction following. This paper presents the initial exploration of using LLMs as surrogate models to explaining black-box recommender models. The primary concept involves training LLMs to comprehend and emulate the behavior of target recommender models. By leveraging LLMs' own extensive world knowledge and multi-step reasoning abilities, these aligned LLMs can serve as advanced surrogates, capable of reasoning about observations. Moreover, employing natural language as an interface allows for the creation of customizable explanations that can be adapted to individual user preferences. To facilitate an effective alignment, we introduce three methods: behavior alignment, intention alignment, and hybrid alignment. Behavior alignment operates in the language space, representing user preferences and item information as text to mimic the target model's behavior; intention alignment works in the latent space of the recommendation model, using user and item representations to understand the model's behavior; hybrid alignment combines both language and latent spaces. Comprehensive experiments conducted on three public datasets show that our approach yields promising results in understanding and mimicking target models, producing high-quality, high-fidelity, and distinct explanations. Our code is available at https://github.com/microsoft/RecAI.