Martin Kragh gives us an insight into his experiences with customer data models and how flexibility is key to ensuring they're not working against you.
I enjoy talking to both i2 users as well as non-i2 users who are using other platforms, and one aspect that really struck me was how important flexibility and adaptability are when it comes to data models.
So, I wanted to share my perspective on why this is. Do reach out to me with your viewpoints and experiences too - I'd love to hear your opinions.
What is a data model?
Wikipedia's definition is:
A data model is an abstract model that organizes elements of data and standardizes how they relate to one another and to the properties of real-world entities. 
In i2, we refer to these organizational elements as Entities, Links and Properties (ELP).
Why is a data model important?
A data model allows us to import data and transform it into a standard format in order to derive insights. They are typically customised for each organization as the relevance of the data will vary. For example, data about "snow" can have very differing importance and meaning depending on whether you work for the Drug Enforcement Administration (DEA) or the National Weather Service.
However, not all data models are created equally. What works perfectly in one situation, can be next to useless in a different investigation.
The importance of flexibility
Back when I worked in the police, our intelligence requirements kept changing depending on the crime we were analysing. For each investigation, the criminal networks, modus operandi, and the questions we needed answers to could be very different. I needed an analysis tool that was flexible as I am creative.
So, what's with the photo of the child's toy? For me, the plastic box with the holes represents the data model; the square peg is the intelligence data; and the hammer is being used to align the data to the data model.
Speaking to users using other analysis platforms, one of their biggest challenges is that the “box with the holes”:
- Is too restrictive, meaning that not all the data can be exploited; or
- Artificially drives the investigation due to the constraints of the data model; or
- Has been designed by the wider organization, meaning that a "one size fits all" model is too complex or becomes too complicated to use.
To give some context to this, I've been told in conversations with analysts:
I am prevented from importing this spreadsheet of social media connections into our current solutions since the data does not fit the current data model.
Do you ever find yourself in this situation? Fighting with data entry rather than getting on with the task of analysing the data?
Augmenting your data model
Every intelligence analysis platform that I know of has a fixed data model.
Such environments are necessary for you to display, analyze and eventually present the data after analysis. However, being able to augment your data model with additional entities or links without needing to change the master data model needs to be a key feature to help the analyst.
Having the ability to augment your data model on an ad-hoc basis allows you to benefit from the data using the master data model, while giving you the ability to adapt it to answer specific questions. It will give you greater flexibility and make your investigations easier to conduct and more comprehensive.
One way i2 provides this flexibility is with the Flexible Importer in Analysis Studio and Analysis Hub. Watch the video below to see it in action.
If you'd like to find out more, or simply to discuss data models and your experiences, then please contact me, my details are below.
Martin worked for almost 16 years in the Danish Police, where he spent his last five years as an analyst in the Danish National Center of Investigation. His expertise in i2 led him to conduct internal i2 enablement sessions for various police districts across Denmark.
He joined i2 in 2015 and is our technical sales leader for EMEA.