History-Augmented Collaborative Filtering for Financial Recommendations
This addresses the need for adaptive recommendations in finance, where client behavior is non-stationary, but it is incremental as it builds on existing collaborative filtering methods with a temporal focus.
The paper tackled the problem of adapting recommender systems to changing client behavior in finance by proposing a novel collaborative filtering algorithm that uses users' and items' recent interaction histories for dynamic recommendations, with performance evaluated on a proprietary G10 bond database from BNP Paribas.
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.