So I’ve been giving this multi-dimesional data store a bit of thought, or well, it has to be something beyond that. I have set up some criteria as to what it must accomplish:
So first of, it is used for storing observed knowledge and thought. Secondly this knowledge has to constantly be sorted in such a way, that the most important(to the current and overall context of the executing character) thoughts are always at the top(R-tree sorting?). And last but not least it must be able to generalize observations into groups, so that not every detail is visible at one single time, and perhaps even forgotten by being too mixed and too generalized with the rest of the thoughts.
The last point might sound a bit confusing, but let’s look at a simple examle. Two weeks ago I spent a week in New York, being there I was always up and running and excited. Now, two weeks later, it seems like it all happened in one day, I have forgotten most of my basic observations, and overgeneralized the trip into a series of highend events. We all know that it is only the highpoints that really stick in memory, however attention, this is related to point 2, generalizing, is knowing that I was in New York 2 weeks ago. Eventually I will only remember that I was in New York, and perhaps name the big places I visited.
Basically a good combination of prioritization and generalization has to be met by the data-handle, of the multi-dimensional data-structure(the semantic net). So far I’ve given it the name of a StackGeneral, and no idea how I will implement it, but I guess it will come.