Entity resolution using machine learning is critical to the know your customer process. Being number one in the industry means that you know that gaming, cannabis, and financial institutions are targeted by those who would mask the true nature of themselves, or their business. You should have a robust solution for entity resolution in your compliance quiver to avoid fines, negative public relation issues, and other adverse regulatory actions. It is vitally important to understand who you actually have on your books, and it is not necessarily easy to disambiguate persons and entities.

Entity resolution answers the question of, Who? Identity theft, synthetic ID, ultimate beneficial owners and controlling interest parties are critical to uncover in any industry, and arguably more so in the high-risk sector. Performing manual searches through various sources and then attempting to aggregate and analyze that information is a very low-level approach to addressing entity resolution. In addition to the low result yield and false positive issues is the work hour cost associated with it. There are so many avenues open to those who choose to obfuscate their identity, and most can be accomplished with a mobile phone or laptop computer. Online registration is huge from a consumer demand and consumer experience perspective and companies are keen to providing that experience to grow their customer base. There is no shortage of news articles where vast numbers of duplicative or synthetic customers are uncovered assumably by way of entity resolution products. PayPal recently indicated that 4.5 million accounts were fraudulent. That is a BIG number.

Gaming, cannabis, and financial institutions that are being targeted by those who would mask the true nature of themselves, or their business, should have a robust solution for entity resolution in their compliance quiver. It is vitally important to understand who you actually have on your books, and it is not necessarily easy to disambiguate persons and entities. At Clarion Compliance our focus is on solutions that have a high Entity Resolution Rating (ER Rating) with a high F1 score. F1 refers to the evaluation metric used to express performance of a machine learning module. In evaluating solutions, focus some effort on both recall and precision for the solution as well as the false positive concern for efficiencies. Accuracy should be evaluated from a false positive and a false negative perspective. Identifying the connection between seemingly different customers, disparate data and the connection between shared identifiers is a heavy lifting job that machine learning and AI can accomplish far better than a room full of analysts. The use cases for entity resolution solutions are broad and compelling and should be a requirement to maintain the integrity of your business. At Clarion Compliance, we focus on effective control activities, for the purpose of assurance, against the backdrop of reasonable diligence.