Take your standard Collins dictionary and price all the words along eight different categories, then price the categories. Choose a different set of categories depending upon your reading choices. If your pricing is accurate, the bots will find graph structures in the text, and report the text to you if it matches your search graph. Teach it to read with child stories, then price the various graphs from a representative set of stories. Run the bots to recurse on the graphs.
It should do math proofs if the various conditions were priced. Square bound, maximum entropy, finite, Integer, etc; all the favorite math words. Then price them according the theory you think need proving. Let it run through all the math text. Researchers would develop many many columns for pricing in their Collins dictionary. Then communicate among papers with graph matching.
My semantic pattern matcher would work in reverse, give it a set of graphs and let it run through the semantic network. It will match by price and re-order the network according to its search graph.
No comments:
Post a Comment