Global-Local Item Embedding for Temporal Set Prediction

TL;DR

This paper introduces GLOIE, a novel method combining global and local temporal information for set prediction, utilizing VAE and dynamic graphs, and demonstrates superior performance on benchmark datasets.

Contribution

GLOIE is the first to integrate global and local temporal patterns in set prediction using VAE and attention mechanisms, with a Tweedie decoder for better modeling of real-world data distributions.

Findings

Outperforms previous state-of-the-art methods on benchmark datasets.

Effectively models zero-inflated and long-tailed data distributions.

Utilizes attention to combine global and local item embeddings.

Abstract

Temporal set prediction is becoming increasingly important as many companies employ recommender systems in their online businesses, e.g., personalized purchase prediction of shopping baskets. While most previous techniques have focused on leveraging a user's history, the study of combining it with others' histories remains untapped potential. This paper proposes Global-Local Item Embedding (GLOIE) that learns to utilize the temporal properties of sets across whole users as well as within a user by coining the names as global and local information to distinguish the two temporal patterns. GLOIE uses Variational Autoencoder (VAE) and dynamic graph-based model to capture global and local information and then applies attention to integrate resulting item embeddings. Additionally, we propose to use Tweedie output for the decoder of VAE as it can easily model zero-inflated and long-tailed distribution, which is more suitable for several real-world data distributions than Gaussian or multinomial counterparts. When evaluated on three public benchmarks, our algorithm consistently outperforms previous state-of-the-art methods in most ranking metrics.