- 1/6/2015 11:06:01 PM
Well, as a political scientist the kind of science I got was the real thing. So it's not the subject matter that's the problem, it's the lack of science education. It can happen in any field and these days it does.
When I was in a PhD program there was serious education. Today there is very little, part. in business and IT, where academia has renounced education and is mainly doing vocational training.
- 1/6/2015 11:02:03 PM
Exactly right: Property rules ARE business rules. As we shall see, there are two categories of business rules: property rules and class rules. They are all components of data interpretation--what the data means. Without the interpretation you are operating with blinders.
You cannot expect people to document and learn the business rules if they are not educated about the existence and function of business modeling and rules. What you can expect is exactly the kind of analytics that you are describing.
- by T Sweeney, Blogger
- 1/6/2015 8:27:55 PM
Ha! Yes, "data science" quite often gets manifested as my guess or desired outcome, buttressed by a few spread sheets and pie charts that mostly support my dubious thesis. More like political data science!
- by geow79, Prospector
- 1/6/2015 8:27:15 PM
Is it valid to assume that your use of the term "Property Rules" is synomous with the term "Business Rules". That is what I learned in database operations & design.
I will have to agree with your on the lack of understanding. It seems like alot of energy goes into "Analytics" but little energy or effort is put into developing Data Dictionary, Meta Data, and Documentation for informed use of a database.
- 1/6/2015 3:43:14 PM
It's not just tempting.
Many if not most analysts are not familiar with data fundamentals, because it is no longer a component of education programs. They are, therefore, unaware that they need to know anything else other than to visually inspect the data in order to ensure that data manipulation and result interpretation makes sense.
Neither are they aware of the distinction between the context of discovery (data mining) and context of validation.
These are some of reasons I frown on the industry concept of "data science" because there is little science in it.