Marzieh Mehdizadeh

Title: Marzieh Mehdizadeh: Inventor in Identity Fraud Detection and Knowledge Graphs.

Introduction

Marzieh Mehdizadeh is an inventor based in Sunnyvale, California, whose work focuses on applying knowledge graphs to identity verification, transaction analysis, and fraud detection. She is associated with Jumio Corporation and holds 2 patents.

Latest Patents

Mehdizadeh’s patent work includes “System and Method of Knowledge Graph Based ID Anomaly Detection in Order to Determine if the Anomaly Is Associated with ID Fraud.” The invention describes processing image fragments from an identity document and evaluating them through a workflow informed by a knowledge graph schema. Comparing evaluation results with a gold standard can help identify image anomalies and potential fraud, with human-curated knowledge serving as an additional information source.

Her other patent is “Knowledge Graph Generation for Contextual Decision Making.” It describes encoding transaction-related data fragments as nodes in transactional and knowledge graphs, then connecting and clustering those nodes to identify patterns. Contextual information from the patterns can support decisions, including reversing decisions to identify possible fraud.

Career Highlights

At Jumio Corporation, Mehdizadeh’s disclosed inventions address the use of structured data relationships and contextual analysis in identity and transaction workflows. Her patents illustrate how knowledge graphs can help organize information and surface patterns relevant to fraud detection.

Collaborations

Mehdizadeh’s listed coworkers include Ashwin Pingali and Arman Mohseni-Kabir.

Conclusion

Marzieh Mehdizadeh’s patent portfolio highlights knowledge graph applications for detecting identity anomalies and supporting contextual fraud decisions.