Large and complex datasets provide opportunities for fraudsters

Knowing Thyself: Why accurate data is at the core of good government, and how to achieve it.

‘Know thyself’

These words are the first and most well-known of the maxims carved above the entrance to the Temple of Apollo at Delphi in Ancient Greece. In a building that was both a source of wisdom and a location for political meetings, it’s a concept that would have given pause both to the philosophers and politicians who guided and governed the populace of the time.

It’s an organising idea that can still be applied today, especially in the ways governments utilise the massive data sets that are crucial for delivering services to their constituents.

For government departments, ‘knowing thyself’ is all about having access to accurate data. Accurate data is at the core of the business of government — and any government body is only as good as the data to which it has access. Good data leads to good decision-making, while bad data leads to bad decision-making — so the more accurate the data available, the more governments will be able to understand about the people they serve, and so how best to serve them.

However — as anyone who has worked with large databases knows — errors created by duplications, changes, typos, or merging different databases together over time can drastically affect the quality of the information within a database, rendering it untrustworthy and virtually unusable. Ultimately, the accuracy of an organisation’s data determines how effective it is.

Knowing thyself also means knowing who your main customers are — particularly the suppliers and providers of the services that are required to fulfill the obligations of governments and do so in a transparent and ethical manner. Not having an accurate understanding of these details leaves a government programme open to the potentiality of waste, fraud, or corruption.

So, what’s the solution? How can government organisations achieve the data quality they need to know themselves and so provide effective services?

Achieving accurate data that provides a basis for good government decision-making is a two-step process, which starts with resolving the discrepancies between the individual pieces of data in a database, and then developing an understanding of the relationships between those data points.

Entity Resolution — The Pursuit of the Golden Record

The first step is known as entity resolution. In the language of databases, an ‘entity’ is a unit of information about a person, an organisation, or any other detail that needs to be stored in that database so that when viewed in relationship to the other entities, a meaningful view of reality is established.

As data is collected in a database, the number of entities within it increases, as does the potential for mistakes and inconsistencies that come about by simple human error. Entity resolution, as a process run by intelligent digital tools, overcomes this issue of human error to produce a single, well-defined version of all the data in a database — a state sometimes described as the “single source of truth”, or more poetically, the Golden Record.

Achieving the Golden Record for your data through entity resolution ensures that 1) you know where your data comes from, 2) you know that it is accurate, and 3) you can trust that data as the basis of meaningful interactions that need to happen to deliver government business effectively.

Entity Resolution is a critical element of database management in a wide range of applications that are relevant to government business. Without a trustworthy database containing verified entities, it would be impossible to effectively carry out any of the following kinds of activities, many of which are core government responsibilities:

• Anti-money Laundering (AML)
• Regulatory compliance
• Fraud Detection
• Electronic Health Records (EHR) maintenance
• Customer Data Management (CDM)
• Homeland security (such as visa screening and border control)
• Counter-terrorism (such as monitoring terrorist watch lists)
• Investigations of individuals and organisations
   

Database management with Entity Resolution is, at its heart, the pursuit of truth — and to borrow another phrase, the truth will set you free. Once you have this truth as the cornerstone of your database, you will then be positioned to use it to make the informed recommendations — and have at your disposal a tool for uncovering potential unlawful or suspicious activity which would otherwise be hidden.

This brings us to the second step in the process — building entity relationships.

Entity Relationships — Getting a 360 view of your data

Whether it is in health management, education, regulatory compliance, population data management, or national security, developing a deep understanding of the relationships between the entities in your databases will provide clear insight into the integrity of any programme a government organisation might undertake.

Once you have resolved the entities in a database, analysing the relationships between those entities will assist in the identification of behaviours that may be suspicious, unethical — and potentially criminal or fraudulent. However, if the number of entities you are dealing with runs into the billions, it is effectively impossible to do this without an effective data analytics tool.

Such things can be exposed by deploying data solutions for matching and resolving entities and thus exposing deep and concealed relationships within a dataset. A current example of this is the ANGIE project, initiated by the Australian Tax Office.

Standing for Automated Network & Grouping Identification Engine — ANGIE is powered by a graph database that automatically identifies and groups entities together to detect ‘patterns of interest’ in the multi-layered relationships between individuals and organisations.

Good data = good government — How to achieve it

Assisting government organisations in building clean data sets is a speciality of Intech Solutions, and we have developed a suite of services that deliver the entity resolution needed to create the kind of Golden Records, and identify entity relationships, key amongst which is our solution known as Single Entity View.

The Single Entity View (SEV) framework is a comprehensive data quality and management platform that focuses on extracting, cleansing, and linking all relevant data, potentially across multiple systems, to the correct and uniquely identified entity. This provides a complete and accurate view of information relevant to each entity — that is, the aforementioned Golden Record.

The SEV solution can provide multiple benefits to government department databases by increasing operational efficiencies, enhancing reporting capabilities, increasing insights gained from analytics, enabling effective service delivery, and investigation campaign execution that leverages the benefits of those insights to improve programme delivery and reduced waste.

Additionally, Intech Solutions’ IQ Office software ingests data, no matter the quality, cleanses and standardises it to produce one well structured, high-quality data set, which is optimised for downstream linking, and can be consumed by a variety of data analytics applications.

The philosophers and politicians of Ancient Greece may have found the maxims of the Oracles to be inspirational for the building and governance of their society, but with the new approaches to entity resolution and Graph-based analytics tools available today, government departments have a fresh opportunity to truly know themselves by better understanding their data and delivering on the potentialities within it.

For more detail on Identity Resolution in Graph Data, including techniques for optimising data quality, watch the Data Quality Enhancements presentation delivered at the international Graph+AI Summit. Click here to view presentation and to access your free Data Science Mini-Guide. Get in touch with us today to discuss how our data and Graph solutions can help you do the business of government better — www.intechsolutions.com.au.

Image credit: Intech Solutions