Effective Dynamic Network Layout for Transaction Analysis

Given an incomplete dataset, FishEye employs various tools, including artificial intelligence, to propose links that enhance the dataset. This paper addresses the challenge of identifying the most reliable tools for completing a knowledge graph. Additionally, the paper tackles the issue of identifying companies potentially involved in illegal, unreported, and unregulated (IUU) fishing. To aid in this endeavor, we introduce a web application designed to detect temporal patterns associated with companies that halt operations and reemerge with new identities. By aggregating data monthly and strategically displaying AI-generated bundle nodes using dropdown controls, our analysis uncovers suspicious patterns, particularly within the Chub Mackerel context. Notably, the Shark and Lichen bundles emerge as dependable graph-completion tools. The overlap between generated data and the knowledge graph in these two bundles underscores the reliability of predictions.