Mastering Data Management

More in Government

• More Mobile, More Competitive Government
• We Are Our Own Columbos
• The End of the Social Security Number?
• Five Rules of Successful Identity and Social Network Analysis

Share this Post

Recent Comments

• Lorraine Fernandes on Response: A Universal Identifier Isn’t a Silver Bullet
• Henrik Liliendahl Sørensen on Response: A Universal Identifier Isn’t a Silver Bullet
• Response: A Universal Identifier Isn’t a Silver Bullet | Mastering Data Management on 3 Best Practices in Identity Management

Chatting to the CIO of a major regional police force recently, he mentioned ‘what I really need is a computer system that will tap me on the shoulder with the knowledge of 6000 police officers’. This got me thinking about what would actually be needed and I came up with the following:

A trusted system of record
Law enforcement data generally refers to Persons, Objects, Locations and Events (POLE) and is held in multiple sources. Therefore the first step has to be to bring data together from each source, identify and remove duplicates, and then associate records from disparate sources. Once this is achieved, it is necessary to derive and store the relationships between entities. This must happen in real time - nobody said it would be easy!

Quality of Data
We know that data quality in many systems is questionable. If you bring data together from multiple sources, how do you know which bits are accurate? How do you present a consistent view of disparate data about the same entity without losing information? How do you derive the elusive ‘golden record’?

We need to provide metrics and measures for data quality, provide mechanisms for data enrichment and provide stewardship tools to allow data to be corrected – to ultimately offer a real time trusted, or composite, view of each entity and relationship.

Once we have a trusted system of record of quality, how do we actually make use of it in the pro-active manner suggested by my contact? To ask the system for what it knows is one thing, to trawl for interesting patterns in the data is another but actually get the computer to ‘tap you on the shoulder’ when it finds something of interest, in real time, is yet another. The technologies that help in this area are:

Network Analysis
There is a need to overlay the trusted system of record derived above with a mathematical network analysis layer that allows the system to search for patterns that may be of interest.

This leads to a number of interesting problems: How do you tell the system which patterns are of interest? How do you get the officer on the beat to explain to the system what he wants to find? How do you limit the search paths such that you do not need infinite hardware in order to run the system? How much data do you have to process to find a result? This step makes the previous ones seem easy!

Triggers, events and workflow
Once you have found something of interest there is a need to appropriately communicate it to the right person in a relevant timescale. The question is, who needs to know what and when do they need to know it?

You want to avoid swamping an individual with irrelevant information but ensure that he does get valuable information at the right time. The investigative officer may need a lot of information, whereas an officer on the street may need a traffic light type alert to determine whether to arrest somebody immediately.

It is clear that to achieve the aim is not straight forward; however, MDM software is beginning to break down the barriers. It is also apparent that solutions that meet at least initial requirements are beginning to come to market. Watch this space!