Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

COMET: NFT Price Prediction with Wallet Profiling

Tianfu Wang, Liwei Deng, Chao Wang, Jianxun Lian, Yue Yan, Nicholas Jing Yuan, Qi Zhang, Hui Xiong

TL;DR

COMET tackles NFT price prediction by fusing collection-level trends with token-level sale forecasting through a wallet-aware, graph-based framework. It introduces a practical problem definition, a wallet profiling framework, and a temporal transaction graph model with community-aware fusion, plus a Transformer-based token sale encoder, all trained in a hierarchical manner. Extensive experiments show COMET outperforms strong baselines and ablations across collection- and token-level tasks, and deployment in a real-world Microsoft pipeline demonstrates practical impact. The work highlights the value of public wallet behavior and community structure as signals for NFT price dynamics, offering a scalable tool for investors.

Abstract

As the non-fungible token (NFT) market flourishes, price prediction emerges as a pivotal direction for investors gaining valuable insight to maximize returns. However, existing works suffer from a lack of practical definitions and standardized evaluations, limiting their practical application. Moreover, the influence of users' multi-behaviour transactions that are publicly accessible on NFT price is still not explored and exhibits challenges. In this paper, we address these gaps by presenting a practical and hierarchical problem definition. This approach unifies both collection-level and token-level task and evaluation methods, which cater to varied practical requirements of investors. To further understand the impact of user behaviours on the variation of NFT price, we propose a general wallet profiling framework and develop a COmmunity enhanced Multi-bEhavior Transaction graph model, named COMET. COMET profiles wallets with a comprehensive view and considers the impact of diverse relations and interactions within the NFT ecosystem on NFT price variations, thereby improving prediction performance. Extensive experiments conducted in our deployed system demonstrate the superiority of COMET, underscoring its potential in the insight toolkit for NFT investors.

COMET: NFT Price Prediction with Wallet Profiling

TL;DR

COMET tackles NFT price prediction by fusing collection-level trends with token-level sale forecasting through a wallet-aware, graph-based framework. It introduces a practical problem definition, a wallet profiling framework, and a temporal transaction graph model with community-aware fusion, plus a Transformer-based token sale encoder, all trained in a hierarchical manner. Extensive experiments show COMET outperforms strong baselines and ablations across collection- and token-level tasks, and deployment in a real-world Microsoft pipeline demonstrates practical impact. The work highlights the value of public wallet behavior and community structure as signals for NFT price dynamics, offering a scalable tool for investors.

Abstract

As the non-fungible token (NFT) market flourishes, price prediction emerges as a pivotal direction for investors gaining valuable insight to maximize returns. However, existing works suffer from a lack of practical definitions and standardized evaluations, limiting their practical application. Moreover, the influence of users' multi-behaviour transactions that are publicly accessible on NFT price is still not explored and exhibits challenges. In this paper, we address these gaps by presenting a practical and hierarchical problem definition. This approach unifies both collection-level and token-level task and evaluation methods, which cater to varied practical requirements of investors. To further understand the impact of user behaviours on the variation of NFT price, we propose a general wallet profiling framework and develop a COmmunity enhanced Multi-bEhavior Transaction graph model, named COMET. COMET profiles wallets with a comprehensive view and considers the impact of diverse relations and interactions within the NFT ecosystem on NFT price variations, thereby improving prediction performance. Extensive experiments conducted in our deployed system demonstrate the superiority of COMET, underscoring its potential in the insight toolkit for NFT investors.
Paper Structure (31 sections, 12 equations, 9 figures, 4 tables)

This paper contains 31 sections, 12 equations, 9 figures, 4 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Problem Definition
  4. NFT Price Prediction System
  5. Data Engineering
  6. Data Collection
  7. Data Preprocessing.
  8. Wallet Profiling Framework
  9. Temporal Transaction Graph Construction
  10. Global Community Detection
  11. Transaction Graph Network
  12. Temporal Attention Representation
  13. Hierarchical Prediction and Optimization
  14. Collection-level Price Trend Predictor
  15. Token-level Sale Price Predictor
  16. ...and 16 more sections

Figures (9)

  • Figure 1: An illustration of NFT ecosystem. (a) Multimodal contents: collection-level name and description, and token-level file and properties. (b) Multi-behaviour transactions: mint, sale, transfer, and burn.
  • Figure 2: An overview of NFT price prediction system with wallet profiling.
  • Figure 3: Temporal transaction graph construction.
  • Figure 4: Results of edge ablation.
  • Figure 5: Results of module ablation.
  • ...and 4 more figures