Penerapan Natural Language Processing dan Machine Learning untuk Prediksi Stres Siswa SMA Berdasarkan Analisis Teks

This research explores the application of Natural Language Processing (NLP) and Machine Learning in predicting stress among high school students. As stress in students often goes unnoticed, there is a need for effective methods to identify it early. To address this issue, this research develops a text-based stress prediction model using NLP for feature extraction and Machine Learning for classification. Core NLP techniques include data cleaning, stopword removal, tokenization, and lemmatization to process text data, while feature extraction is achieved through methods such as Bag Of Words (BOW), Term Frequency-Inverse Document Frequency (TF-IDF), and N-grams (Unigram, Bigram, Trigram). The Machine Learning models tested include Logistic Regression, Naive Bayes, Random Forest, and Support Vector Machine (SVM). Results from the experiments showed that the Naive Bayes model using Bigram features achieved the highest accuracy of 95.6%, with the other models achieving around 93%. Despite the strong performance of the models, errors such as False Positive and False Negative were still found, indicating room for improvement. This research shows that NLP combined with Machine Learning provides an effective approach to identifying student stress, with promising potential for mental health interventions in educational settings.