Resumen

GONCALES, L. J.; FARIAS, K.; KUPSSINSKU, L.  y  SEGALOTTO, M.. The effects of applying filters on EEG signals for classifying developers’ code comprehension. J. appl. res. technol [online]. 2021, vol.19, n.6, pp.584-602.  Epub 22-Mar-2022. ISSN 2448-6736.  https://doi.org/10.22201/icat.24486736e.2021.19.6.1299.

EEG signals are a relevant indicator for measuring aspects related to human factors in software engineering. EEG is used in software engineering to train machine learning techniques for various applications, including classifying task difficulty and developers’ experience. The EEG signal contains noise such as abnormal readings, electrical interference, and eye movements, which are usually not of interest to the analysis, and therefore contribute to the lack of precision of the machine learning techniques. However, research on software engineering has not evidenced how effective the filtering of EEG signals is, even with evident benefits of filtering the EEG signals in signal processing and clinical image studies. In this paper, we analyzed the effects of using filtered EEG signals for classifying developers’ code comprehension. This filter consists of high and low pass filtering designed with an FIR filter using a Hamming window. This filtering process also removes abnormal signals and executes the Independent Component Analysis (ICA) using the fast ICA method for removing EOG components. We applied the filtered EEG signals to train a random forest (RF) machine learning technique to classify the developers' code comprehension. This study also trained another random forest classifier with unfiltered EEG data. We evaluated both models using 10-fold cross-validation. This work measures the classifiers' effectiveness using the f-measure metric. This work used the t-test, Wilcoxon, and U Mann Whitney to analyze the difference in the effectiveness measures (f-measure) between the classifier trained with filtered EEG and the classifier trained with unfiltered EEG. The tests pointed out a significant difference after applying EEG filters to classify developers' code comprehension with the random forest classifier. The conclusion is that the EEG filters significantly improve the effectiveness of classifying code comprehension using the random forest technique.

Palabras llave : EEG; high pass filter; low pass filter; ICA; software engineering; program comprehension; machine learning.

        · texto en Inglés     · Inglés ( pdf )