Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

A Recommender System for NFT Collectibles with Item Feature

Minjoo Choi, Seonmi Kim, Yejin Kim, Youngbin Lee, Joohwan Hong, Yongjae Lee

TL;DR

The paper tackles the challenge of recommending NFT collectibles in a market with sparse feedback and user anonymity. It introduces a data-efficient graph-based recommender that builds a bipartite user-item graph from transaction data and enriches item representations with multimodal features (image, text, price) using an NGCF-inspired architecture trained with a pairwise BPR objective. Through experiments on five NFT collections, the approach demonstrates substantial accuracy gains over baselines, with text and price features proving particularly informative and high-order graph connectivity contributing to improved recommendations. The findings show that graph-based methods can effectively leverage NFT-specific features and transaction signals to deliver more precise NFT recommendations, offering practical benefits for users and NFT marketplaces.

Abstract

Recommender systems have been actively studied and applied in various domains to deal with information overload. Although there are numerous studies on recommender systems for movies, music, and e-commerce, comparatively less attention has been paid to the recommender system for NFTs despite the continuous growth of the NFT market. This paper presents a recommender system for NFTs that utilizes a variety of data sources, from NFT transaction records to external item features, to generate precise recommendations that cater to individual preferences. We develop a data-efficient graph-based recommender system to efficiently capture the complex relationship between each item and users and generate node(item) embeddings which incorporate both node feature information and graph structure. Furthermore, we exploit inputs beyond user-item interactions, such as image feature, text feature, and price feature. Numerical experiments verify the performance of the graph-based recommender system improves significantly after utilizing all types of item features as side information, thereby outperforming all other baselines.

A Recommender System for NFT Collectibles with Item Feature

TL;DR

The paper tackles the challenge of recommending NFT collectibles in a market with sparse feedback and user anonymity. It introduces a data-efficient graph-based recommender that builds a bipartite user-item graph from transaction data and enriches item representations with multimodal features (image, text, price) using an NGCF-inspired architecture trained with a pairwise BPR objective. Through experiments on five NFT collections, the approach demonstrates substantial accuracy gains over baselines, with text and price features proving particularly informative and high-order graph connectivity contributing to improved recommendations. The findings show that graph-based methods can effectively leverage NFT-specific features and transaction signals to deliver more precise NFT recommendations, offering practical benefits for users and NFT marketplaces.

Abstract

Recommender systems have been actively studied and applied in various domains to deal with information overload. Although there are numerous studies on recommender systems for movies, music, and e-commerce, comparatively less attention has been paid to the recommender system for NFTs despite the continuous growth of the NFT market. This paper presents a recommender system for NFTs that utilizes a variety of data sources, from NFT transaction records to external item features, to generate precise recommendations that cater to individual preferences. We develop a data-efficient graph-based recommender system to efficiently capture the complex relationship between each item and users and generate node(item) embeddings which incorporate both node feature information and graph structure. Furthermore, we exploit inputs beyond user-item interactions, such as image feature, text feature, and price feature. Numerical experiments verify the performance of the graph-based recommender system improves significantly after utilizing all types of item features as side information, thereby outperforming all other baselines.
Paper Structure (28 sections, 1 figure, 2 tables)

This paper contains 28 sections, 1 figure, 2 tables.

Table of Contents

  1. 1. Introduction
  2. 2. Related Works
  3. Recommender Systems
  4. NFT Recommender Systems
  5. 3. Data
  6. Data Preparation
  7. Data Description and Sparsity
  8. Item Features
  9. Image
  10. Text
  11. Price
  12. 4. Method
  13. Problem Definition
  14. Item Feature Embedding
  15. Model Architecture
  16. ...and 13 more sections

Figures (1)

  • Figure 1: Performance comparison between NGCF and NGCF variant models over the top K recommendations on different datasets.