Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

KGUF: Simple Knowledge-aware Graph-based Recommender with User-based Semantic Features Filtering

Salvatore Bufi, Alberto Carlo Maria Mancino, Antonio Ferrara, Daniele Malitesta, Tommaso Di Noia, Eugenio Di Sciascio

TL;DR

KGUF addresses the complexity of knowledge-graph based recommender systems by using user-driven feature filtering to prune semantic KG features and a simple, linear aggregation to fuse collaborative and content signals. The method models items via latent KG features and learns user/item embeddings through multi-hop propagation with a knowledge-aware correction term, controlled by a balancing parameter α. It uses decision trees to select user-relevant semantic features, reducing noise and computation. Experiments on MovieLens 1M, Yahoo! Movies, and Facebook Books show KGUF achieving comparable or superior accuracy to state-of-the-art KGCF methods, while maintaining a simpler formulation and providing code and data for reproducibility.

Abstract

The recent integration of Graph Neural Networks (GNNs) into recommendation has led to a novel family of Collaborative Filtering (CF) approaches, namely Graph Collaborative Filtering (GCF). Following the same GNNs wave, recommender systems exploiting Knowledge Graphs (KGs) have also been successfully empowered by the GCF rationale to combine the representational power of GNNs with the semantics conveyed by KGs, giving rise to Knowledge-aware Graph Collaborative Filtering (KGCF), which use KGs to mine hidden user intent. Nevertheless, empirical evidence suggests that computing and combining user-level intent might not always be necessary, as simpler approaches can yield comparable or superior results while keeping explicit semantic features. Under this perspective, user historical preferences become essential to refine the KG and retain the most discriminating features, thus leading to concise item representation. Driven by the assumptions above, we propose KGUF, a KGCF model that learns latent representations of semantic features in the KG to better define the item profile. By leveraging user profiles through decision trees, KGUF effectively retains only those features relevant to users. Results on three datasets justify KGUF's rationale, as our approach is able to reach performance comparable or superior to SOTA methods while maintaining a simpler formalization. Link to the repository: https://github.com/sisinflab/KGUF.

KGUF: Simple Knowledge-aware Graph-based Recommender with User-based Semantic Features Filtering

TL;DR

KGUF addresses the complexity of knowledge-graph based recommender systems by using user-driven feature filtering to prune semantic KG features and a simple, linear aggregation to fuse collaborative and content signals. The method models items via latent KG features and learns user/item embeddings through multi-hop propagation with a knowledge-aware correction term, controlled by a balancing parameter α. It uses decision trees to select user-relevant semantic features, reducing noise and computation. Experiments on MovieLens 1M, Yahoo! Movies, and Facebook Books show KGUF achieving comparable or superior accuracy to state-of-the-art KGCF methods, while maintaining a simpler formulation and providing code and data for reproducibility.

Abstract

The recent integration of Graph Neural Networks (GNNs) into recommendation has led to a novel family of Collaborative Filtering (CF) approaches, namely Graph Collaborative Filtering (GCF). Following the same GNNs wave, recommender systems exploiting Knowledge Graphs (KGs) have also been successfully empowered by the GCF rationale to combine the representational power of GNNs with the semantics conveyed by KGs, giving rise to Knowledge-aware Graph Collaborative Filtering (KGCF), which use KGs to mine hidden user intent. Nevertheless, empirical evidence suggests that computing and combining user-level intent might not always be necessary, as simpler approaches can yield comparable or superior results while keeping explicit semantic features. Under this perspective, user historical preferences become essential to refine the KG and retain the most discriminating features, thus leading to concise item representation. Driven by the assumptions above, we propose KGUF, a KGCF model that learns latent representations of semantic features in the KG to better define the item profile. By leveraging user profiles through decision trees, KGUF effectively retains only those features relevant to users. Results on three datasets justify KGUF's rationale, as our approach is able to reach performance comparable or superior to SOTA methods while maintaining a simpler formalization. Link to the repository: https://github.com/sisinflab/KGUF.
Paper Structure (18 sections, 10 equations, 2 figures, 2 tables)

This paper contains 18 sections, 10 equations, 2 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Related work
  3. Background
  4. Knowledge graphs for recommendation
  5. Decision trees
  6. Methodology
  7. Modeling items through knowledge graph
  8. Refining item features with user decision
  9. Knowledge-aware aggregation
  10. Model learning
  11. Experimental settings
  12. Datasets and Knowledge Graph
  13. Baselines
  14. Evaluation protocol and reproducibility
  15. Results and discussion
  16. ...and 3 more sections

Figures (2)

  • Figure 1: An example of how KGUF models the item representation. Figure \ref{['fig:sub1']} shows the entities in the knowledge graph linked to the items. In figure \ref{['fig:sub2']}, the most relevant semantic features selected by KGUF are reported.
  • Figure 2: An example of how KGUF models the item representation. Figure \ref{['fig:sub1']} shows the entities in the knowledge graph linked to the items. In figure \ref{['fig:sub2']}, the most relevant semantic features selected by KGUF are reported.