Tuesday, October 16, 2007

What is essential? Why do we make data models?

We need to make data models to make it clear what data are essential to the system.  We need to understand how information flows.  But does this mean that we have to detail every single process all the time? I don't think so.

For example, if you have a department that you would like to study for their systems integration and stuff, you could say that you have particular processes which use certain data and if those have elementary processes (the really detailed, step by step thing, as in those you see in the manual of procedures, etc.) it's possible that you don't need to dwell on the long details of the process.  But you need to understand what information is there, if it's the same information within that process and how it's transformed or used as you go through that process.

Sometimes dwelling too much on the details of the process keeps us away from what information is really important to the system, to the users.

We need to make our data model clear. Though the processes make it easier for us to understand how data interact with each other, processes tend to change.  But the data tend to stay there.  We might add some bits of data along the way, maybe later on, but first, we need to see what's really crucial.

If we keep on just talking about processes, things will never get done especially if your processes seem to change a lot.  Like which one goes first and stuff like that.  I think that's a reason why systems don't get done.  People complain too much about their processes but look for a solution that's geared towards data.  If they want to re-org, then they have to do it on their own and not foist it down the throats of people studying how they use their data.

Or maybe I should write a different post on that matter altogether.

No comments:


Blog Archive

The Learner

My photo

a crafty cat who loves food, linux, anime and crafting.

business analyst by day, blogger by night. also a frustrated artist.