Friday, 2nd December 2011
There are four primary core functions performed by those who do analytics. The first of these is the matching function.
Matching can be used for three purposes including identity matching to confirm the identity of each citizen, discrepancy matching to identify those who have for example failed to declare interest earned on a bank account for tax purposes and data fusion to link and match data to remove duplicate records.
Matching is moving to the next level of sophistication where data is being matched on the basis of knowledge. If two unrelated data items do not match on the basis of content they can be made to connect (ie made to snap together) if it is ascertained that at a knowledge level there is a relationship between them. For example, the connection can be made if person X that lives at address Y is a cousin of person A who lives at address B.
Matching is the most powerful and pervasive weapon that governments have to prevent identity crime, protect a country from terrorists, to identify those who overstay their travel visas, to find those who have not paid their debts and similar. It can be done a massive scale provided those doing it have access to a wide variety of data sources and have the computing power and algorithms to do comparisons between data items at lightning speed.
The second analytics function in importance is mapping because it is a graphical approach that involves identifying links between people, places and events such as the links between suspects involved in a crime in terms of their names, addresses, contact details and relationships with other subjects of interest.
Mapping is used in social network analysis to understand relationships between entities and to identify causes and effects. An example is in epidemiology to help understand how diseases, such as influenza, spread in a population. When mapping is combined with matching it gives organizations two powerful tools to discover entities and events of interest and the relationships between them.?The third analytics function is mining. This is about the discovery of new knowledge such as finding new configurations, correlations, patterns and trends in data.
An example is identifying which patients are the most expensive to service by health practitioners. ?The fourth analytics function is modelling. This includes making classifications, predictions or other calculations such as estimates of future population growth. ?Whereas matching and mapping functions are suited to high-volume searches, the mining and modelling functions are usually much more focused activities that concentrate on specific issues.
If an organization wants to gain the most from analytics it needs to develop all four capabilities of matching, mapping, mining and modelling. They are, when used, together potent weapons for finding important insights and understandings in data.