Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Cross-Domain Sequential Recommendation via Neural Process

Haipeng Li, Jiangxia Cao, Yiwen Gao, Yunhuai Liu, Shuchao Pang

TL;DR

This work raises a challenging and unexplored question: How to unleash the potential of non-overlapped users' behaviors to empower CDSR.

Abstract

Cross-Domain Sequential Recommendation (CDSR) is a hot topic in sequence-based user interest modeling, which aims at utilizing a single model to predict the next items for different domains. To tackle the CDSR, many methods are focused on domain overlapped users' behaviors fitting, which heavily relies on the same user's different-domain item sequences collaborating signals to capture the synergy of cross-domain item-item correlation. Indeed, these overlapped users occupy a small fraction of the entire user set only, which introduces a strong assumption that the small group of domain overlapped users is enough to represent all domain user behavior characteristics. However, intuitively, such a suggestion is biased, and the insufficient learning paradigm in non-overlapped users will inevitably limit model performance. Further, it is not trivial to model non-overlapped user behaviors in CDSR because there are no other domain behaviors to collaborate with, which causes the observed single-domain users' behavior sequences to be hard to contribute to cross-domain knowledge mining. Considering such a phenomenon, we raise a challenging and unexplored question: How to unleash the potential of non-overlapped users' behaviors to empower CDSR?

Cross-Domain Sequential Recommendation via Neural Process

TL;DR

This work raises a challenging and unexplored question: How to unleash the potential of non-overlapped users' behaviors to empower CDSR.

Abstract

Cross-Domain Sequential Recommendation (CDSR) is a hot topic in sequence-based user interest modeling, which aims at utilizing a single model to predict the next items for different domains. To tackle the CDSR, many methods are focused on domain overlapped users' behaviors fitting, which heavily relies on the same user's different-domain item sequences collaborating signals to capture the synergy of cross-domain item-item correlation. Indeed, these overlapped users occupy a small fraction of the entire user set only, which introduces a strong assumption that the small group of domain overlapped users is enough to represent all domain user behavior characteristics. However, intuitively, such a suggestion is biased, and the insufficient learning paradigm in non-overlapped users will inevitably limit model performance. Further, it is not trivial to model non-overlapped user behaviors in CDSR because there are no other domain behaviors to collaborate with, which causes the observed single-domain users' behavior sequences to be hard to contribute to cross-domain knowledge mining. Considering such a phenomenon, we raise a challenging and unexplored question: How to unleash the potential of non-overlapped users' behaviors to empower CDSR?

Paper Structure

This paper contains 23 sections, 18 equations, 4 figures, 5 tables.

Table of Contents

  1. Introduction
  2. PRELIMINARY
  3. Neural Processes
  4. Problem Setting
  5. Methodology
  6. Embedding Layer
  7. Sequence Encoder
  8. Prior and Posterior Distributions Alignment
  9. Interest Adaptive Layer
  10. Training and Inference
  11. Model Optimization
  12. EXPERIMENTS
  13. Datasets
  14. Experimental Setting
  15. Evaluation Metrics
  16. ...and 8 more sections

Figures (4)

  • Figure 1: An illustration of overlapped/non-overlapped users, where previous methods focus on the overlapped users' behavior learning.
  • Figure 2: Neural processes in the training/testing phases
  • Figure 3: The framework of CDSRNP in the training phase. $\mathbf{z}_s$ will be used instead of $\mathbf{z}_q$ in the testing phase
  • Figure 4: Performance of different hyper-parameters