Forum Moderators: open
(I know what Jeremy's going to be doing tonight :))
However for the next two weeks, my evenings will be spent lazily perusing those, to see if there are any new developments in the academic realm of searching I need to know about.
I have this theory about how indexing works. For every set of variables, there is a limit of the number of permutations of those variables. And for searching, I believe the engines have already pegged the variable list, in as far as the limit of those variables which lend themselves to indexing on such large scales as today's search engines.
With this variable limit in mind, consider the continuum of variable differentiation. This is where we find ourselves, as SEO's, today. We each are looking at the continuum of variable differentiation, and attempting to reach the section which coincides with an SE's representation of the variable matrix. For those of us who participate here, I believe we have a collective good inclination as to where the ideal mix of varialbes lies.
However, the field is continuing to evolve, and there are always those varialbes today which might lend themselves to mass computation tomorrow. And thus, might become part of the search engine algorithm, where as few years back, things like massive link analysis just didn't occur. I don't know whether it was a design decision, lack of bandwidth and computational power, or what.
The evolution of algorithmic design is fascinating, and I look forward to each time an SE ramps up it's mathematical model of knowledge. Considering the vast breadth of subjects, materials, and information already available, how might different variable sets be combined to produce new and interesting ways of information retrieval? It seems to me, the combinations are wonderously varied.
I keep thinking about how last year IBM created a new computer, I believe called 8,192. It was the biggest computer ever built, a *nix based cluster of 8,192 computers working together. Google recently doubled it's size of computers, now totaling 8,000 in two data centers.
How much might the accessibility of such large scale computational power alter the current variable set? There are always more varialbes to be considered, and with this limit increasing, the variable landscape might increase in tandem.
