Energy Credit Best Practices – Chapter: Information Technology http://ccro.org © Copyright 2022, CCRO. All rights reserved. 29 • Data Mining – Think of Data Mining similarly to commodities mining, such as coal ore, oil, etc. This is the processing of large amounts of randomly stored data to extract valuable components. Typically, the output from Data Mining is input into other advanced data techniques. Therefore, the better the Data Mining tasks are performed, the better subsequent business decisions can be made. Data Mining sources include emails, transactional Systems (for trends, etc.), social sentiment (such as Twitter, Facebook, Glassdoor, etc.). When done appropriately, Data Mining will also associate other 3rd party data to the data being mined. Examples of this include: Pulling in weather patterns when certain business-level events occur, storing traffic patterns and transportation issues when deliveries are delayed, etc. • Deep Learning – An approach where decisions are formulated based on a series of layers. Each layer performs a specific task and no more. This allows that layer to learn from a minor decision point and thus becomes much more proficient at its task than from a more significant, more complex task. The concept of “deep” implies that hundreds or thousands of layers can be automatically generated as your Systems continue to learn. Over time, they can successfully predict the appropriate paths instead of having to transverse each layer. 3.5.3 Shared Ledger (Block Chain) Within the last four to six years, Blockchain technology has been at the forefront of a technology set attempting to solve a reasonably universal issue: secure information or transactions in a manner that cannot be modified outside of secure, authorized channels. Blockchain’s largest market currently is in the FinTech space, it also has an important place in the commodity and energy markets and throughout the energy supply chain. • Example - Recently, a gas and fuel distribution chain has teamed up to create a broad Blockchain implementation strategy to help prevent fraud and theft. As fuel theft has a worldwide impact of over $1b per year, having a way to securely trace and verify these types of transactions has become very important. This works because Blockchain Shared Ledger entries are used from the refinery to the delivery vehicle and from the delivery vehicle to the station tank. At each handoff, a new “block” was added to the Blockchain with amounts transferred and the product's quality. This allowed both the supplier and customer to understand if/when theft or a lower-quality product was introduced. This initiative was seen as a quality improvement service to the supplier’s offering that differentiated them from other suppliers and provided confidence to the customer that they are receiving the quantity and quality of the product they purchased. One can expand this application into the Natural Gas pipeline world, as an example, where dealing with tariffs, injection sites, and varying pressurized pipelines introduces different levels of quality.