Realities of electronic information management

Picture the scene. It is your first day in a brand new job. Your role is that of information architect in a company with lots and lots of information that needs managing. You have a sizeable budget, enough people and (bliss!) board-level backing to take a green field approach to implementing an EIM infrastructure.

At the start of the week you are understandably pumped, psyched and ready to blaze a trail in your new role. You take a look at the information assets that need to be managed. You build a conceptual model of the information universe. You harden that model into a collection of information structures. You further harden them by selecting products/tools that support your chosen informationstructures. You get all that done by Friday and head home for the weekend, breezing through reception leaving an odor of supreme confidence in your wake. You put you best sunglasses on and climb into the company car. You tap the steering wheel to the beat of your favorite music. Life is so good.

It's important to enjoy such experience when they come your way as they are few and far between and, in my experience, they last one week on average. From there, things can get real messy, real fast. Welcome to the world of information management.

In the full flush of irrational exuberance that typically accompanies the early stage of information management projects, effort centers around picking the perfect structure for the information assets and the workflows they traverse. Roughly speaking, there are three choices:

1) Put the information into a linked set of tabular structures - the database approach.

2) Put the information into a linked set of hierarchical structure - the XML approach.

3) Put the information into a linked set of freeform documents - the web pages, word processor approach.

Unless you are very lucky, you will not be long into this exercise before you find that none of the above are perfect for your needs. In fact, if your experiences mirror mine, you will find that the closer you look, the more exceptions you come across. Exceptions that add more warts to your information model. Over time, the pristine edifice you had in your head at the end of that first Friday, has metamorphosed into a more carbuncular structure. A structure adorned with compromises and "temporary" fixes. A structure held aloft by constant minding and the collective telekinetic wills of the minds involved in its creation.

Here is a tip for calming your soul if this scenario rings a bell with you:

There is no such thing as the perfect information model.

While I'm at it, here is another tip:

No information model survives the first encounter with real information in the real world.

The real trick of information management is to recognize that all models are compromises and that all information structures must constantly change to reflect the constantly changing real world.

A tall order and not a problem I have a solution for (sorry). I do however have some suggestions.

1. Do not knock yourself out looking for the perfect model. It doesn't exist. Spend an appropriate amount of time, then stop, make a decision and move on. The world will move on, whether you like it or not.

2. Do not castigate yourself when it turns out imperfect. That perfection is unattainable.

3. Think of whatever structure you choose as being simply the physical manifestation of the information from which you can derive any number oflogical views. At the end of the first Friday, when you think your model is perfect, your logical view matches your physical view. As the weeks go by, the physical view stays the same, but you can add and rework as many logical views as you need to in order keep pace with what the real world foists upon you.

4. If a particular logical view ends up predominating, you may want to transform your information assets to reflect that. It then becomes your new, improved physical view from which a new generation of logical views will be derived.

There are a number of technologies/approaches that help in the practicalapplication of this idea of constantly evolving information structures.

There are various technologies that facilitate the creation of an "information facade" - a way of treating a logical view of an information asset as the physical view thus insulating applications from the carbuncular structures that develop at the physical view level. These are particularly powerful in the case of tabular structures from databases as they allow you to move significantly beyond the concept of "view" that is typically provided in database systems. Not surprisingly, in the times we live in, these views are often exposed as XML[2].

Also of note is XFML[1] - a notation for faceted classification of information assets. This meta-data driven approach allows you to create a variety of tabular and hierarchical views from the meta-data attached to your information assets.

The next time you are having one of those blissful first weeks in an information management role, you might like to factor some faceting technologies into your thinking on day one. It will repay itself in weeks two, three, four . . .


Join the newsletter!

Error: Please check your email address.

More about Logical

Show Comments