Currently popular data modeling methods are entrenched in finding nouns and then later to map them to classes or their attributes. This has always caused common problems of ad hocness to decide which nouns contribute to classes/entities or their attributes.
Because of this ad-hoc affair, data modeling always looked as an art than a science.
In fact, the industry and academia have no clear idea of what these models are representing? No doubt, in an explicit way.
We shall say, the data model pertains to information, which we can state as so and so facts or meaningful sentences. That does not really take place at all.
We build data models without a clear idea of what is information to us for our purposes.
To come out of these problems, there is a clever way of modifying existing fact oriented analysis methods and adopting them to data modeling.
These methods have a clear understanding that information is meaningful sentences, also called facts.
I would appreciate peoples comments on the approach taken in my new fact oriented analysis method that you can find at www.EnglishToUML.com.
With the best regards,
Venkat