High-quality decision making by aim-oriented modeling

One of the propositions of the paper is to represent classification, evaluation, judgement, decision making and choice as an evaluation task by defining different purposes of evaluation. The main result of this approach is the knowledge based evaluation model. The model is able to produce the aggregated characteristic for object in terms of goal contribution. The aggregated characteristic takes into account all the object´s features (both quantitative and qualitative), all demands to the object and some weight factors. Analysis of the aggregated characteristics allow one to evaluate a single object (e.g. its time changing), to compare all estimated objects, to make forecasting for single object dynamics and to do a clear explication for decision making. The main principles and general stages of vector model forming are described. The vector model is a base for the universal knowledge based system EXPEK PRO, which was successfully used for various AI tasks