and Biological Feedback Loops
Department of Psychophysiology,
Moscow State University, Moscow, Russia
One of the forms of human-computer interaction (HCI) is computer-assisted education (CAE). To conduct CAE in a corridor of optimal functional states avoiding frustration and drowsiness two contours of feedback are suggested: a cognitive and biological one.
A cognitive feedback has to correct the process of concept formation. The terms to be acquired are represented in a geometrical model - a semantic space. The semantic space is constructed from a matrix of subjective differences between respective terms using multidimensional scaling (MDS). Each term in the ndimensional semantic space is characterized by specific location being represented by an excitation vector composed from excitations of long-term memory units. The semantic subjective differences are equal to distances between the ends of such excitation vectors. Two terms are synonyms if their excitation vectors are equal. It means that their representations in the long-term memory coincide. In the process of learning the student semantic space is matched against expert semantic space. The discrepancies between the spaces are used to generate instruction to pay attention to a content of respective terms to eliminate a mismatch between student and expert semantic spaces.
The vector encoding of semantics with participation of long-term memory is based on vector code at perceptual level. The study of color vision in normal trichromatic subjects has shown that colors of different hue, lightness and saturation are located on a hypersphere in the four-dimensional Euclidean space. Each color stimulus is encoded by an excitation vector having four componentsexcitations of four color-coding neurons: red-green, blue-yellow, bright and dark ( Izmailov and Sokolov, 1991). Color working memory and color long-term memory are replica of the perceptual color space. Color names of the native language are labels of specific long-term memory units ( Sokolov, 1998).