Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Enhancing Conversational Recommender Systems with Tree-Structured Knowledge and Pretrained Language Models

Yongwen Ren, Chao Wang, Peng Du, Chuan Qin, Dazhong Shen, Hui Xiong

TL;DR

PCRS-TKA addresses hallucination and context underutilization in pretrained language models for conversational recommender systems by integrating knowledge graphs through a retrieval-augmented, prompt-based framework. It constructs dialogue-specific knowledge trees from KGs, selectively filters information, and explicitly models collaborative user preferences via a specialized supervision signal with semantic alignment to fuse inputs. The system combines a static KG encoder with dynamic, serialized knowledge trees fed into a frozen PLM, enabling end-to-end reasoning over structured graph information. Experiments on INSPIRED and ReDial show consistent improvements in both item recommendations and conversational quality, surpassing strong baselines and demonstrating robustness across PLMs and settings.

Abstract

Recent advances in pretrained language models (PLMs) have significantly improved conversational recommender systems (CRS), enabling more fluent and context-aware interactions. To further enhance accuracy and mitigate hallucination, many methods integrate PLMs with knowledge graphs (KGs), but face key challenges: failing to fully exploit PLM reasoning over graph relationships, indiscriminately incorporating retrieved knowledge without context filtering, and neglecting collaborative preferences in multi-turn dialogues. To this end, we propose PCRS-TKA, a prompt-based framework employing retrieval-augmented generation to integrate PLMs with KGs. PCRS-TKA constructs dialogue-specific knowledge trees from KGs and serializes them into texts, enabling structure-aware reasoning while capturing rich entity semantics. Our approach selectively filters context-relevant knowledge and explicitly models collaborative preferences using specialized supervision signals. A semantic alignment module harmonizes heterogeneous inputs, reducing noise and enhancing accuracy. Extensive experiments demonstrate that PCRS-TKA consistently outperforms all baselines in both recommendation and conversational quality.

Enhancing Conversational Recommender Systems with Tree-Structured Knowledge and Pretrained Language Models

TL;DR

PCRS-TKA addresses hallucination and context underutilization in pretrained language models for conversational recommender systems by integrating knowledge graphs through a retrieval-augmented, prompt-based framework. It constructs dialogue-specific knowledge trees from KGs, selectively filters information, and explicitly models collaborative user preferences via a specialized supervision signal with semantic alignment to fuse inputs. The system combines a static KG encoder with dynamic, serialized knowledge trees fed into a frozen PLM, enabling end-to-end reasoning over structured graph information. Experiments on INSPIRED and ReDial show consistent improvements in both item recommendations and conversational quality, surpassing strong baselines and demonstrating robustness across PLMs and settings.

Abstract

Recent advances in pretrained language models (PLMs) have significantly improved conversational recommender systems (CRS), enabling more fluent and context-aware interactions. To further enhance accuracy and mitigate hallucination, many methods integrate PLMs with knowledge graphs (KGs), but face key challenges: failing to fully exploit PLM reasoning over graph relationships, indiscriminately incorporating retrieved knowledge without context filtering, and neglecting collaborative preferences in multi-turn dialogues. To this end, we propose PCRS-TKA, a prompt-based framework employing retrieval-augmented generation to integrate PLMs with KGs. PCRS-TKA constructs dialogue-specific knowledge trees from KGs and serializes them into texts, enabling structure-aware reasoning while capturing rich entity semantics. Our approach selectively filters context-relevant knowledge and explicitly models collaborative preferences using specialized supervision signals. A semantic alignment module harmonizes heterogeneous inputs, reducing noise and enhancing accuracy. Extensive experiments demonstrate that PCRS-TKA consistently outperforms all baselines in both recommendation and conversational quality.

Paper Structure

This paper contains 20 sections, 15 equations, 3 figures, 4 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Conversational Recommendation Systems
  4. Unifying PLMs and KGs in Recommendations
  5. Methodology
  6. Methodology
  7. Task Formulation
  8. Overview of the Approach
  9. User Preference Extraction Module
  10. Knowledge Tree Enhanced Module
  11. Multimodal Information Integration Module
  12. Prompt Learning Module
  13. Experiments
  14. Experimental Setup
  15. Overall Performance
  16. ...and 5 more sections

Figures (3)

  • Figure 1: The network architecture of the PCRS-TKA framework.
  • Figure 2: Ablation study on INSPIRED dataset for recommendation task.
  • Figure 3: Model performance comparison with varing degree and depth of knowledge tree on INSPIRED dataset.