History-Augmented Collaborative Filtering for Financial Recommendations

TL;DR

This paper introduces a new neural network-based collaborative filtering algorithm that dynamically adapts to changing user behaviors over time, specifically tailored for financial recommendation systems, demonstrated on proprietary bond trading data.

Contribution

It presents a novel history-augmented collaborative filtering method that captures temporal context to improve recommendations in non-stationary financial environments.

Findings

Effective in modeling non-stationary user behaviors

Improves recommendation accuracy in financial data

Demonstrated on proprietary bond trading dataset

Abstract

In many businesses, and particularly in finance, the behavior of a client might drastically change over time. It is consequently crucial for recommender systems used in such environments to be able to adapt to these changes. In this study, we propose a novel collaborative filtering algorithm that captures the temporal context of a user-item interaction through the users' and items' recent interaction histories to provide dynamic recommendations. The algorithm, designed with issues specific to the financial world in mind, uses a custom neural network architecture that tackles the non-stationarity of users' and items' behaviors. The performance and properties of the algorithm are monitored in a series of experiments on a G10 bond request for quotation proprietary database from BNP Paribas Corporate and Institutional Banking.