I’m not sure what to call this. It’s just an idea I just had sort of following on from yesterday’s post. What if the reports and data mining results we produce were ‘active’. In other words, what if they were set up so that programs could process them and use them. There would have to be
- an information backing to the reports
- a set of data in the reports and
- a presentation layout design for the reports.
The second two already tend to exist. The first one is the interesting one. It would tell you
- what was in the report
- what its context was
- what its relationships were, and
- what its meaning was, and
- (what its query was).
The Information Backing For a Report
The information representation would include
- Knowledge Representation ontologies
- Endemes, and
- Endeme powered fields.
Telling you what the report was would require the first two. Telling you the context is would require all three. including the relationships would require all three. including the meaning would require all three.
I haven’t actually figured out how to do this. My plan is:
- build a bunch of lists of primitives – level 1.
- notice that some of the lists parallel each other and switch multiple lists of primitives to a single list of a class that contains the primitives as members – level 2.
- Activate the primitive members of the class so that they are activated fields powered by endemes. i.i make a single list of activated fields in the class with each member now characterized by an endeme in a single endene set – level 3.
- Move the mechanics related to the endeme set out of the class into its own class. This one I have never done, we’ll see if this works and what the results are when I do this – level 4.
- Build a report inside the program – level 4.
- Figure out an ontology system to store the report layout, data, and information – level 4.
- store the system in a file – level 4.
- Build a report generator to generate the report based on the file – level 4.