Transaction Categorization with Relational Deep Learning in QuickBooks

TL;DR

This paper introduces Rel-Cat, a graph-based deep learning model for automatic transaction categorization in QuickBooks, improving accuracy, scalability, and cold start adaptation by formulating the task as link prediction within a relational database.

Contribution

The paper presents a novel graph-based model, Rel-Cat, that leverages relational data and NLP techniques to enhance transaction categorization in QuickBooks, outperforming existing models.

Findings

Rel-Cat outperforms existing QuickBooks models in accuracy.

The model scales effectively with increasing data.

It addresses cold start challenges with minimal data adaptation.

Abstract

Automatic transaction categorization is crucial for enhancing the customer experience in QuickBooks by providing accurate accounting and bookkeeping. The distinct challenges in this domain stem from the unique formatting of transaction descriptions, the wide variety of transaction categories, and the vast scale of the data involved. Furthermore, organizing transaction data in a relational database creates difficulties in developing a unified model that covers the entire database. In this work, we develop a novel graph-based model, named Rel-Cat, which is built directly over the relational database. We introduce a new formulation of transaction categorization as a link prediction task within this graph structure. By integrating techniques from natural language processing and graph machine learning, our model not only outperforms the existing production model in QuickBooks but also scales effectively to a growing customer base with a simpler, more effective architecture without compromising on accuracy. This design also helps tackle a key challenge of the cold start problem by adapting to minimal data.