Stripe has patented a method using identity graphs for fraud detection during transactions. The method involves processing document images to detect fraudulent identity documents through machine learning models. The final score determines if the document is fraudulent, enhancing security measures. GlobalData’s report on Stripe gives a 360-degree view of the company including its patenting strategy. Buy the report here.

According to GlobalData’s company profile on Stripe, Rebate eligibility processing was a key innovation area identified from patents. Stripe's grant share as of May 2024 was 41%. Grant share is based on the ratio of number of grants to total number of patents.

Fraud detection during transactions using identity graphs

Source: United States Patent and Trademark Office (USPTO). Credit: Stripe Inc

A newly granted patent (Publication Number: US12002054B1) outlines a method for detecting fraudulent identity documents using document images. The process involves receiving a document image, extracting data from the image, and processing it through a single machine learning model with multiple backbones, each associated with a specific fraud determination aspect. These backbones generate intermediate signals that are combined to produce a final fraud score, indicating whether the document is fraudulent. When this final score meets a predetermined threshold, the system identifies the document as fraudulent. The method includes various machine learning model backbones, such as a convolutional neural network, transformer models, and identity feature models, to analyze different aspects of the document for fraud detection.

Furthermore, the patent describes a system integrated into a commerce platform for verifying identity documents provided by users. The system can transmit positive or negative fraud determinations to the service system based on the final fraud score generated. Additionally, intermediate fraud scores are stored for further analysis, and explanation data is generated to explain the fraud determination process to the service system. The system utilizes a neural network model with different machine learning backbones and heads, including multilayer perceptron heads, to enhance the accuracy of fraudulent identity document detection. Feedback data on fraud determinations is used to annotate and retrain the system periodically, ensuring continuous improvement in fraud detection capabilities.

To know more about GlobalData’s detailed insights on Stripe, buy the report here.

Data Insights

From

The gold standard of business intelligence.

Blending expert knowledge with cutting-edge technology, GlobalData’s unrivalled proprietary data will enable you to decode what’s happening in your market. You can make better informed decisions and gain a future-proof advantage over your competitors.

GlobalData

GlobalData, the leading provider of industry intelligence, provided the underlying data, research, and analysis used to produce this article.

GlobalData Patent Analytics tracks bibliographic data, legal events data, point in time patent ownerships, and backward and forward citations from global patenting offices. Textual analysis and official patent classifications are used to group patents into key thematic areas and link them to specific companies across the world’s largest industries.