Utilizing machine and deep learning algorithms to identify learning-related features in electroencephalography data during second language acquisition

Neurolinguistics explores how the brain processes and acquires language, particularly the neural mechanisms involved in second language acquisition (SLA). This study investigates EEG-based neurolinguistic patterns during the process of learning Arabic and Hindi as second languages. EEG data were collected from 10 participants (6 males and 4 females; M = 34.1, SD = 4.7 years) engaged in language acquisition tasks, and the signals underwent preprocessing steps, including noise removal, independent component analysis (ICA), and spectral feature extraction. Various machine learning and deep learning models were applied to classify brain activity associated with SLA, achieving high accuracy: Random Forest (93.90%), XGBoost (94.95%), CatBoost (94.69%), CNN (92.17%), SVM (83.33%), and K-NN (86.61%). The results highlight significant neural activation in the left frontal and temporal lobes, particularly in Broca’s and Wernicke’s areas, when processing second-language stimuli. This study contributes to neurolinguistics by demonstrating how machine learning can effectively classify EEG patterns related to language learning, paving the way for future research on cognitive and computational models of SLA.