Fuente: WIPO "tomato" A method for detecting outlier evaluators in an evaluation process includes Collecting a data matrix Sij to a computer system using an automatic input interface, wherein the elements of the matrix Sij are real numbers in a predefined range, each of the elements of the matrix Sij represent a numerical evaluation provided by an evaluator j for an evaluated entity i. For each entity i calculating by the computer system an Initial Consensus in the form of a vector Mi using a robust measure of location. Calculating by the computer system an Adjusted Distances EDj between each of the data matrix Sij provided by the evaluator j for entity i and the Initial Consensus Mi. Applying by the computer system a nonlinear transformation function ƒ to the Adjusted Distances EDj. Identifying by the computer system the outlier evaluators by detecting transformed Adjusted Distances




E
⁢

D
j
*





exceeding a robust upper-bound threshold.