AUTONOMINĖ EMOCIJŲ ATPAŽINIMO SISITEMA: DUOMENŲ GAVYBOS METODAI
AUTONOMOUS EMOTION RECOGNITION SYSTEM: APPROACHES FOR DATA MINING

Summary/Abstract: The detection of emotion is becoming an increasingly important field for human-computer interaction as the advantages emotion recognition offer become more apparent and realisable. However there are still many issues (data filtering, parameter‘s extraction, data preprocessing, interpreting, adaptive control) in developing adaptive systems proeviding user-friendly e-health and e-social care for people with movement disabilities services based on physiological parameter’s recognition. Such systems include different intellectual components for control and monitoring of sensors by supporting multi-agent activities and, in accordance to the recognition of certain situations, integrate the possibilities to affect and control the devices of disable persons. So this paper presents principle of modelling of an autonomous emotion recognition system to creating of an intelligent e-health care environment. The model is based on remote research of human emotional states and remote bio robots intelligent control with ATmega8/16/32 microcontrollers. The proposed model uses skin conductance signal to recognize human emotional state i.e. the main process of this system is based on exploratory’s analog signal transformation to one of discreete emotional state (surprise, happy, calmness, sleepiness, sad, disgust, anger and fear). Using Firebird database to store physiological parameters makes proposed model more universal and extended in possibilities. There are described signal transformations, filtering, data recording methods using Atmel AVR microcontrollers, digital oscilloscope and R statistical environment. There are proposed self organizing maps (SOM) and multilayer perceptron (MLP) combinations for emotional state recognition and improved MLP training approach, which increases the learning rate and classification accuracy, in this paper as well.

• Details
• Contents