Re: OT: Getting into the data modeling field

frand345_at_thonas.com said...
> Sorry about the OT post but I'm not sure where to post.
> I want to get into the data modeling field and would like to know what
> kind of background I would need. I don't have a CS degree, but I have
> taken about a half dozen courses in programming. I don't have any
> practical experience (am currently a computer operator), so what would
> be a good course of action. Would I have to be a DBA first?. I don't
> really want to go through the DBA certification (Oracle eg) process as
> it sounds very expensive and is not really what I want to do.
>
> I think I have quite good analytical and verbal skills. I know that
> book knowledge does not mean much to recruiters, but I feel that even
> though I am self taught, I am reaching the point where I can say I
> know a decent amount about the field. I know my way around ER
> modeling, a good amount of ORM, but not so much UML. I am fairly
> knowledgable about normalization, issues related to business rules,
> etc. I don't know too much about physical implementations, but I can
> bone up on that. Further, I have 'practical' experience modeling
> certain web pages that I go to often, Amazon, eg. In short, there is
> virtually no aspect of data modeling in which I am completely
> clueless. I think that within six months or so I would be able to
> create a simple data model in a job interview if called upon to do so.
>
> Anybody have any thoughts, observations, suggestions? In what other
> newsgroups would this post be more appropriate?
>

As with any field, there are variations on job titles. There are several flavors of "data modelers."

A friend of mine is a very successful data modeler / facilitator. He would tell you that good verbal and inter-personal skills are the most important personality trait for the job. But that's because he stands in front of a group of business subject-area experts (the ones who own, or work with, the raw data), in a conference room with a white board behind him. Then he starts talking, digging information about the data from the heads of the attendees and writing it on the white board as clues and hints are uncovered. He lists and categorizes, erases, and moves stuff around ... until everyone in the room agrees that he has essentially captured the general picture. This is a very people-oriented type of data modeling.

Another friend does it differently. Being somewhat more shy around people, he gathers information about the data through typical research techniques (like analyzing existing reports and forms and such). Sitting at his desk, he organizes the metadata into natural relationships. Once he has it documented, then he presents it to the group of business experts. He spends more time at his desk than in meetings.

Both carry the title Data Modeler ... well, the second is actually called a Data Administrator, but he does data modeling. IMHO, a Data Modeler has to be a global, rather than a detailed, thinker when it's time to gather the data; then shift into a detailed thinker when it comes time to categorize and specify relationships.

I don't think you need to come up from the ranks of developer or DBA to be a successful data modeler ... though I'll be quick to point out that it certainly helps to know something about how your great model will eventually be implemented. :) My first friend, the people-oriented one, was never a developer or DBA; he was an ex-Marine. My second friend, the people-avoidance one, got a promotion from his Director of Engineering job (where developers were "labeled" engineers).

-- 
/Karsten
DBA > retired > DBA