A Concept Knowledge Graph for User Next Intent Prediction at Alipay

TL;DR

This paper presents AlipayKG, a concept knowledge graph and a Transformer-based model for predicting user next intent, deployed at Alipay to serve over 100 million users, improving task performance and explainability.

Contribution

Introduction of AlipayKG, an offline concept knowledge graph, and a Transformer-based model integrating expert rules for accurate, explainable user intent prediction.

Findings

Effective enhancement of downstream task performance

Retains explainability in intent prediction

Deployed at scale for over 100 million users

Abstract

This paper illustrates the technologies of user next intent prediction with a concept knowledge graph. The system has been deployed on the Web at Alipay, serving more than 100 million daily active users. To explicitly characterize user intent, we propose AlipayKG, which is an offline concept knowledge graph in the Life-Service domain modeling the historical behaviors of users, the rich content interacted by users and the relations between them. We further introduce a Transformer-based model which integrates expert rules from the knowledge graph to infer the online user's next intent. Experimental results demonstrate that the proposed system can effectively enhance the performance of the downstream tasks while retaining explainability.