AutoFraudNet: A Multimodal Network to Detect Fraud in the Auto Insurance Industry

TL;DR

AutoFraudNet is a multimodal neural network designed to detect auto insurance fraud by effectively fusing diverse data types, overcoming training challenges, and reducing overfitting, leading to improved detection performance.

Contribution

This work introduces AutoFraudNet, a novel multimodal framework with a cascaded slow fusion and lightweight design to enhance fraud detection in auto insurance claims.

Findings

Multimodal approaches outperform unimodal and bimodal methods.

AutoFraudNet improves PR AUC by over 3%.

Effective fusion of multiple data modalities enhances detection accuracy.

Abstract

In the insurance industry detecting fraudulent claims is a critical task with a significant financial impact. A common strategy to identify fraudulent claims is looking for inconsistencies in the supporting evidence. However, this is a laborious and cognitively heavy task for human experts as insurance claims typically come with a plethora of data from different modalities (e.g. images, text and metadata). To overcome this challenge, the research community has focused on multimodal machine learning frameworks that can efficiently reason through multiple data sources. Despite recent advances in multimodal learning, these frameworks still suffer from (i) challenges of joint-training caused by the different characteristics of different modalities and (ii) overfitting tendencies due to high model complexity. In this work, we address these challenges by introducing a multimodal reasoning framework, AutoFraudNet (Automobile Insurance Fraud Detection Network), for detecting fraudulent auto-insurance claims. AutoFraudNet utilizes a cascaded slow fusion framework and state-of-the-art fusion block, BLOCK Tucker, to alleviate the challenges of joint-training. Furthermore, it incorporates a light-weight architectural design along with additional losses to prevent overfitting. Through extensive experiments conducted on a real-world dataset, we demonstrate: (i) the merits of multimodal approaches, when compared to unimodal and bimodal methods, and (ii) the effectiveness of AutoFraudNet in fusing various modalities to boost performance (over 3\% in PR AUC).