Development of algorithm for fuzzy art appraisal model

Abstract

This paper presents an algorithm for a fuzzy art appraisal model, which is a hierarchical model based on a base price and the following adjustment. In the first step of the model, we determine the list of linguistic variables, their number, types of terms and types of membership functions for each term. Then, we analyze the subject area, process expert information and build a knowledge base containing 50 predicate rules of inference. The analysis shows that the model reflects a 4.37% error in a porcelain figurine appraisal. The paper also outlines recommendations on the implementation of the developed algorithm for fuzzy art appraisal model using Fuzzy Logic Toolbox for Matlab package and explains package limitations such as the need for strong authentication of art pieces, identification of mass artworks and a limited range of artwork that can be appraised.