Formulating a learning factor using ERP signals evoked by a known and unknown language

Abstract

Hearing enriches our lives and gives us the ability to communicate with others. The hearing provides us with enormous sources of information. Moreover, knowledge acquisition in humans is based basically on seeing and listening. Humans have a distinct ability to learn languages and our brains get easily acquitted to visual information. The present study is an attempt to analyze the event-related potential (ERP) signals evoked by known and unknown language words. In which, the study was described the various listening effects of sentences on brain activity. The effect is studied through the analyses of recorded EEG signals along with activities. The study compared and analyzed the brain oscillations responses at all frequency bands (delta, theta, alpha, beta, gamma). The PSD (power spectral density) in the (Cz, Fz, Pz, T3, T4, T5, T6, P3, P4, C3, C4, F3, F4) electrodes positions according to (10-20 EEG electrodes system) at different brain lobes for subjects listening to two different languages (known and unknown languages) were analyzed. The study came out with a learning language meter, which is, the correlation coefficient of the brain responses signals between the languages. Support Vector Machines (SVM) and Artificial neural network (ANN) classifier is used to classify the responses between the two language processing. The results of the study confirmed in a different way that known language and unknown language will cause a different reaction in the left frontal brain region. The author recommend the proposed learning meter can be used to measure the responses between different languages.