Apparatus and method for calculating internal state for artificial emotion

摘要

An apparatus and method for calculating an internal state for artificial emotions are disclosed, of which the method comprises multiplying an input value obtained from a sensor with a first personality set in accordance with at least one low rank element contained in at least one high rank element of a NEO PI-R (Revised NEO Personality Inventory); calculating a personality factor value in a Five-Factor Model of the personality by adding the results of the multiplication; and calculating the internal state by multiplying the personality factor value with a second personality.

主权项

1. A method for calculating a current internal state value for artificial emotions, the method comprising:
obtaining at least one facet value vector by multiplying an input value obtained from a sensor with a first personality set in accordance with at least one low rank element contained in at least one high rank element of a NEO PI-R (Revised NEO Personality Inventory); adding up the at least one facet value vector so as to calculate a personality factor value in a Five-Factor Model of personality; multiplying the personal factor value with a second personality; and calculating the current internal state value by adding one or more values obtained by multiplying the personality factor value with the second personality and one or more values obtained by multiplying the just previous internal state value with a third personality, wherein the at least one facet value vector based on the NEO PI-R is calculated from the following equation:
f(k)=Γu(k) where Γ means a matrix of x×z where the first personality γ expressed in numbers is arranged on a diagonal line, x means the number of low rank elements, z means the number of external input values, u(k) means a matrix of z×1 representing an external input value, k means a sampling time wherein the personality factor value of the personality factor value calculation step is calculated from the following equation:
g(k)=1xTf(k),where 1xT is a transposed matrix of a matrix of x×1 filled with 1, and
wherein the current internal state value of the personality factor value calculation step is calculated from the following equation:
y(k)=βg(k)+αy(k−1), where β means the second personality expressed in numbers, α means the third personality expressed in numbers, and k−1 means the just previous sampling time.