查看文章

The emergence of cloud-based bookkeeping platforms has made it possible to automate tedious tasks, such as bank reconciliation. Bank reconciliation is the process of comparing and matching the recorded transactions of a business with their bank statements. This is an essential part of the bookkeeping process in order to verify the overall accuracy and integrity of the business’s financial records. Traditional approaches to automated bank reconciliation use natural language processing techniques, but these approaches disregard structural relationships between financial records. In this work, we propose an alternative methodology that employs graph representation learning and downstream link prediction to predict matches for bank reconciliation. Our proposed models surpass industry benchmarks and offer a more robust solution, improving the identification of group-based reconciliations; a feature often disregarded by traditional approaches. Additionally, we introduce a novel post-processing technique, Top Boundary Ranking, which enhances the detection of grouped reconciliations.