Concentration level classification from social media distraction based on random forest method

In today’s digital era, the use of social media has become an inseparable part of daily life, often interfering with an individual’s ability to focus due to distractions from social media. Numerous studies have been conducted to measure the level of concentration of individuals in the face of such distractions. This study attempts to measure the impact of social media notifications on the level of individual concentration by using the Concentration Grid Test (CGT) while also recording Electroencephalogram (EEG) and Electrocardiogram (ECG) data. The CGT is used as a method to assess concentration by sequencing numbers from 1 to 100 in a 10x10 grid for one minute. EEG and ECG data are recorded simultaneously to identify neurophysiological correlations to concentration levels. The collected data is then processed with features of brain waves Alpha, Beta, Theta, Gamma, the Beta/Theta ratio, and heart rate (BPM) values. The Random Forest (RF) classification method is utilized to classify concentration levels based on the extraction of these features, including maximum Gamma value, Beta/Theta ratio, average Beta, Gamma values, and BPM. The research findings demonstrate that the RF model successfully classified concentration levels into two categories, namely good and poor concentration, with a cross-validation accuracy reaching 71.53% to measure the influence of social media distractions on an individual’s concentration.