Method to Teach Model of Extracting Knowledge from Natural Language Texts

A method to teach model of extracting knowledge from natural language texts is given. A feasibility of teaching is provided by the simplicity of extract rules and lattice of lexical contingencies that are key words of the model. The method of teaching creates a set of rules based on teaching examples prepared by man-expert. Some experiments are conducted. The dependence of main quality factors of the taught model on properties of the initial teaching sample is estimated.