Article Per Year (5 Year)
p-Index From 2020 - 2025
0.444
P-Index
This Author published in this journals
All Journal Jurnal Teknologi Informasi dan Ilmu Komputer Journal of Information Systems Engineering and Business Intelligence
Larasati, Sza Sza Amulya
Unknown Affiliation
Computer Science & IT Engineering

Published : 2 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 2 Documents
Search

 1 
Optimizing IndoBERT for Revised Bloom's Taxonomy Question Classification Using Neural Network Classifier Darfiansa, Lazuardy Syahrul; Fitriyani; Larasati, Sza Sza Amulya
Journal of Information Systems Engineering and Business Intelligence Vol. 11 No. 2 (2025): June
Publisher : Universitas Airlangga

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.20473/jisebi.11.2.226-237

Abstract

Background: A major challenge in Indonesian education system is the continued dominance of exam questions that primarily assess basic thinking skills, such as remembering and understanding. In order to effectively nurture students with critical, analytical, and creative thinking skills, the integration of higher-order thinking questions has become increasingly urgent. An effective conceptual framework that can be utilized in this regard is Revised Bloom's Taxonomy (BT). This framework classifies cognitive skills into 6 levels, namely remember, understand, apply, analyze, evaluate, and create. Furthermore, the framework is particularly important as it promotes the development of exam questions that transcend lower-level thinking skills, fostering a deeper and higher level of understanding among students. In this context, automated systems powered by deep learning (DL) have shown promising accuracy in classifying questions based on BT levels, thereby offering practical support for educators aiming to design more meaningful and intellectually stimulating assessments.  Objective: This research aims to develop a classification system that can effectively classify Indonesian exam questions based on BT using IndoBERT pretrained models. These models were combined with Convolutional Neural Network (CNN) and Long Short-Term Memory (LSTM) classifiers (referred to as IndoBERT-CNN and IndoBERT-LSTM) to determine the model with the highest performance.   Methods: The dataset utilized was self-collected and underwent several stages of preparation, including expert labeling and splitting. Furthermore, preprocessing was conducted to ensure the dataset was consistent and free from irrelevant features related to case folding, tokenization, stopword removal, and stemming. Hyperparameter fine-tuning was subsequently carried out on IndoBERT, IndoBERT-CNN, and IndoBERT-LSTM. Model performance was evaluated using Accuracy, F-Measure, Precision, and Recall.  Results: The fine-tuned IndoBERT model results showed that IndoBERT-LSTM outperformed IndoBERT-CNN. The optimal hyperparameter configuration, batch size of 64 and learning rate of 5e-5, showed the highest performance, achieving Accuracy of 88.75%, Precision of 85%, Recall of 88%, and F-Measure of 86%.  Conclusion: IndoBERT, IndoBERT-CNN, and IndoBERT-LSTM reflected promising results, although the performance of the models was significantly affected by respective architectures and hyperparameter settings. Among the three observed models, IndoBERT was found to perform best with smaller batch sizes and moderate learning rates. IndoBERT-CNN achieved stronger results with a higher learning rate and similar batch sizes. IndoBERT-LSTM recorded the highest accuracy with larger batch sizes for gradient stability. However, IndoBERT was constrained by its focus on Indonesian language, and the interpretability of the predictions made, specifically in relation to expert-labeled data, remained unclear.  Keywords: Bloom’s Taxonomy, CNN, Hyperparameter Fine-Tuning, IndoBERT, LSTM, Question Classification
Penerapan Decision Tree dan Random Forest dalam Deteksi Tingkat Stres Manusia Berdasarkan Kondisi Tidur Larasati, Sza Sza Amulya; Dewi, Elok Nuraida Kusuma; Farhansyah, Brahma Hanif; Bachtiar, Fitra Abdurrachman; Pradana, Fajar
Jurnal Teknologi Informasi dan Ilmu Komputer Vol 11 No 5: Oktober 2024
Publisher : Fakultas Ilmu Komputer, Universitas Brawijaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.25126/jtiik.2024117993

Abstract

Masalah kesehatan mental menjadi isu global yang sangat umum terjadi, termasuk perubahan suasana hati, perbedaan kepribadian, ketidakmampuan mengatasi masalah, serta mengisolasi diri dari keramaian. Berdasarkan data dari World Health Organization (WHO), gangguan kecemasan dan stres menjadi gangguan mental yang paling sering terjadi dari 970 juta kasus yang dilaporkan sepanjang tahun 2019. Stres telah banyak dikaitkan dengan tidur. Penelitian ini akan mengungkap hubungan kondisi tidur pada manusia dengan tingkat stres yang sedang diderita dengan 5 tingkatan: normal, stres ringan, stres sedang, stres tinggi, stres sangat tinggi. Data yang digunakan merupakan data kontinyu dengan 8 fitur: ‘sr’ (snoring rate), ‘rr’ (respiration rate), ‘t’ (body temperature), ‘lm’ (limb movement), ‘bo’ (blood oxygen), ‘rem’ (rapid eye movement), ‘sh’ (sleeping hours), dan ‘hr’ (heart rate). Setiap fitur memiliki rentang nilai yang tidak sama, sehingga dilakukan normalisasi untuk menyeragamkan rentang tersebut. Hyperparameter tuning dilakukan dengan teknik k-fold cross validation dan model dirancang dengan algoritma klasifikasi Decision Tree serta Random Forest. Hasilnya, 5 fitur: tingkat mendengkur, laju respirasi, pergerakan anggota tubuh termasuk bola mata, serta detak jantung saat tidur berbanding lurus dengan tingkat stres. Semakin tinggi nilai kelima fitur tersebut mengindikasikan tingkat stres yang lebih tinggi. Sedangkan dengan 3 fitur lainnya: suhu tubuh, kadar oksigen, dan waktu tidur memberikan hasil sebaliknya. Dengan kata lain, ketiga nilai tersebut berbanding terbalik dengan tingkat stres yang diderita. Model Decision Tree memiliki akurasi 0,99 dan Random Forest memiliki akurasi 1,0. Hasil penelitian ini diharapkan dapat memberikan insight bagi peneliti lain pada bidang yang sama dan dapat menjadi acuan dalam mendeteksi stres yang sedang diderita.       Abstract Stress is often associated with sleep. This research aims to uncover the relationship between human sleep conditions and the level of stress experienced, categorized into five levels: not stressed, very mildly stressed, mildly stressed, highly stressed, and very highly stressed. The data used consists of continuous data with eight features: 'snoring rate' (snoring rate), 'respiration rate' (respiration rate), 'body temperature' (body temperature), 'limb movement' (limb movement), 'blood oxygen' (blood oxygen), 'rapid eye movement' (rapid eye movement), 'sleep hours' (sleep hours), and 'heart rate' (heart rate). Each feature has a different value range, so normalization is performed to standardize these ranges. Hyperparameter tuning is done using k-fold cross-validation, and the model is designed using the Decision Tree and Random Forest classification algorithms. The results show that five features: snoring rate, respiration rate, limb movement including eye movement, and heart rate during sleep are directly proportional to the level of stress. Higher values for these five features indicate higher levels of stress. On the other hand, the other three features: body temperature, blood oxygen level, and sleep hours yield the opposite results. In other words, the values of these three features are inversely proportional to the level of stress experienced. The Decision Tree model has an accuracy of 0.99, and the Random Forest model has an accuracy of 1.0. The results of this research are expected to provide insights for other researchers in the same field and can serve as a reference for detecting ongoing stress.
Co-Authors Dewi, Elok Nuraida Kusuma Fajar Pradana Farhansyah, Brahma Hanif Fitra Abdurrachman Bachtiar Fitriyani Lazuardy Syahrul Darfiansa