An Application of Object Detection in Contract Analysis

In this paper, we discuss the application of visual object-detection analysis on an information governance business use case. Specifically, we discuss how handwriting detection can serve as the centerpiece of a workflow where signatures are detected in PDF documents, and the number of detected signatures is utilized as a method of singling out countersigned and fully executed contracts. We outline the business case and the recommended workflow involving final validation using an eDiscovery platform and provide statistics from a real-world project to illustrate the potential effectiveness of the solution.