SearchGCN: Powering Embedding Retrieval by Graph Convolution Networks for E-Commerce Search
This work addresses embedding-based candidate retrieval for large-scale e-commerce search, representing an incremental improvement over existing methods.
The paper tackled the problem of applying graph convolution networks to industrial-scale e-commerce search engines, resulting in the deployment of SearchGCN at JD.com, which learns better embedding representations, particularly for long-tail queries and items.
Graph convolution networks (GCN), which recently becomes new state-of-the-art method for graph node classification, recommendation and other applications, has not been successfully applied to industrial-scale search engine yet. In this proposal, we introduce our approach, namely SearchGCN, for embedding-based candidate retrieval in one of the largest e-commerce search engine in the world. Empirical studies demonstrate that SearchGCN learns better embedding representations than existing methods, especially for long tail queries and items. Thus, SearchGCN has been deployed into JD.com's search production since July 2020.