Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Enhancing Content-based Recommendation via Large Language Model

Wentao Xu, Qianqian Xie, Shuo Yang, Jiangxia Cao, Shuchao Pang

TL;DR

This paper tackles content-based recommendation by leveraging explicit textual reviews and ID signals. It introduces LoID, a two-part approach: (1) a plug-in semantic knowledge transfer using LoRA adapters on large language models to capture multi-domain semantics, and (2) an ID-based contrastive objective to align content-derived representations with ID signals, enabling cross-domain transfer. The method merges source-domain LoRAs with DARE, then applies target-domain Re-LoRA and an ID-aware attention-contrastive module to predict ratings, achieving significant gains over SOTA baselines on 11 Amazon-domain datasets. The findings demonstrate the effectiveness of plugin-based semantic transfer and content/ID alignment for scalable, multi-domain recommender systems with practical deployment potential. Future work includes incorporating additional modalities to further enrich user/item representations.

Abstract

In real-world applications, users express different behaviors when they interact with different items, including implicit click/like interactions, and explicit comments/reviews interactions. Nevertheless, almost all recommender works are focused on how to describe user preferences by the implicit click/like interactions, to find the synergy of people. For the content-based explicit comments/reviews interactions, some works attempt to utilize them to mine the semantic knowledge to enhance recommender models. However, they still neglect the following two points: (1) The content semantic is a universal world knowledge; how do we extract the multi-aspect semantic information to empower different domains? (2) The user/item ID feature is a fundamental element for recommender models; how do we align the ID and content semantic feature space? In this paper, we propose a `plugin' semantic knowledge transferring method \textbf{LoID}, which includes two major components: (1) LoRA-based large language model pretraining to extract multi-aspect semantic information; (2) ID-based contrastive objective to align their feature spaces. We conduct extensive experiments with SOTA baselines on real-world datasets, the detailed results demonstrating significant improvements of our method LoID.

Enhancing Content-based Recommendation via Large Language Model

TL;DR

This paper tackles content-based recommendation by leveraging explicit textual reviews and ID signals. It introduces LoID, a two-part approach: (1) a plug-in semantic knowledge transfer using LoRA adapters on large language models to capture multi-domain semantics, and (2) an ID-based contrastive objective to align content-derived representations with ID signals, enabling cross-domain transfer. The method merges source-domain LoRAs with DARE, then applies target-domain Re-LoRA and an ID-aware attention-contrastive module to predict ratings, achieving significant gains over SOTA baselines on 11 Amazon-domain datasets. The findings demonstrate the effectiveness of plugin-based semantic transfer and content/ID alignment for scalable, multi-domain recommender systems with practical deployment potential. Future work includes incorporating additional modalities to further enrich user/item representations.

Abstract

In real-world applications, users express different behaviors when they interact with different items, including implicit click/like interactions, and explicit comments/reviews interactions. Nevertheless, almost all recommender works are focused on how to describe user preferences by the implicit click/like interactions, to find the synergy of people. For the content-based explicit comments/reviews interactions, some works attempt to utilize them to mine the semantic knowledge to enhance recommender models. However, they still neglect the following two points: (1) The content semantic is a universal world knowledge; how do we extract the multi-aspect semantic information to empower different domains? (2) The user/item ID feature is a fundamental element for recommender models; how do we align the ID and content semantic feature space? In this paper, we propose a `plugin' semantic knowledge transferring method \textbf{LoID}, which includes two major components: (1) LoRA-based large language model pretraining to extract multi-aspect semantic information; (2) ID-based contrastive objective to align their feature spaces. We conduct extensive experiments with SOTA baselines on real-world datasets, the detailed results demonstrating significant improvements of our method LoID.
Paper Structure (20 sections, 14 equations, 1 figure, 3 tables)

This paper contains 20 sections, 14 equations, 1 figure, 3 tables.

Table of Contents

  1. Introduction
  2. Preliminary
  3. Problem Statement
  4. Low-Rank Adaptation (LoRA)
  5. Drop and Rescale (DARE)
  6. Methodology
  7. Overview
  8. Source Domain LoRA Pretraining
  9. Target Domain Re-LoRA
  10. ID-based Contrastive Learning
  11. Model Optimization
  12. Experiments
  13. Experimental Setup
  14. Datasets
  15. Baselines
  16. ...and 5 more sections

Figures (1)

  • Figure 1: Model Architecture of the proposed LoID.