Claim Missing Document
Check
Articles

A Conversion of Signal to Image Method for Two-Dimension Convolutional Neural Networks Implementation in Power Quality Disturbances Identification Berutu, Sunneng Sandino; Chen, Yeong-Chin; Wijayanto, Heri; Budiati, Haeni
JOIV : International Journal on Informatics Visualization Vol 6, No 4 (2022)
Publisher : Society of Visual Informatics

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30630/joiv.6.4.1529

Abstract

The power quality is identified and monitored to prevent the worst effects arise on the electrical devices. These effects can be device failure, performance degradation, and replacement of some device parts. The deep convolutional neural networks (DCNNs) method can extract the complexity of image features. This method is adopted for the power quality disruption identification of the model. However, the power quality signal data is a time series. Therefore, this paper proposes an approach for the conversion of a power quality disturbance signal to an image. This research is conducted in several stages for constructing the approach proposed. Firstly, the size of a matrix is determined based on the sampling frequency values and cycle number of the signal. Secondly, a zero-cross algorithm is adopted to specify the number of signal sample points inserted into rows of the matrix. The matrix is then converted into a grayscale image. Furthermore, the resulting images are fed to the two-dimension (2D) CNNs model for the PQDs feature learning process. When the classification model is fit, then the model is tested for power quality data prediction. Finally, the model performance is evaluated by employing the confusion matrix method. The model testing result exhibits that the parameter values such as accuracy, recall, precision, and f1-score achieve at 99.81%, 98.95%, 98.84, and 98.87 %, respectively. In addition, the proposed method's performance is superior to the previous methods. 
Text Mining dan Klasifikasi Sentimen Berbasis Naïve Bayes Pada Opini Masyarakat terhadap Makanan Tradisional Sunneng Sandino Berutu
Jurnal Sistem Komputer dan Informatika (JSON) Vol. 4 No. 2 (2022): Desember 2022
Publisher : Universitas Budi Darma

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30865/json.v4i2.5138

Abstract

Indonesia has several famous traditional foods and is available in some cities. In addition, several international foods also are interesting to Indonesian. This article analyzes the netizen sentiment for these food categories where the data source is Twitter. The foods are rendang, sate, gudeg, pizza, hamburger, and spaghetti. The text mining approach is adopted to process data. The research steps are data crawling, cleaning, filtering, translating, and splitting. Furthermore, the classifier model based on the Naïve Bayes algorithm is developed. The analysis result shows that the gudeg food reaches a high percentage of positive sentiment with 57,9.  Then, the high rate of negative sentiment is achieved by the rendang food with 21,9 %. Moreover, hamburger food obtains a high percentage of neutral sentiment. Meanwhile, the evaluation of classifier model performance shows that the model with the hamburger dataset achieves a high score for accuracy, precision, and recall parameters with 0.72, 0.72, and 0.68 sequentially. 
Contrastive Learning pada IndoBERT untuk Analisis Sentimen Kebijakan Makan Bergizi Gratis Dwi Dian Sari Nonibenia Hia; Sunneng Sandino Berutu; Jatmika
JSAI (Journal Scientific and Applied Informatics) Vol 9 No 1 (2026): Januari
Publisher : Fakultas Teknik Universitas Muhammadiyah Bengkulu

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.36085/jsai.v9i1.9963

Abstract

Transformer-based language models such as IndoBERT still face limitations in topic and sentiment analysis of short social media texts, particularly due to embedding anisotropy, semantic overlap between topics, and limited sensitivity to implicit sentiment intensity. This study aims to evaluate the effectiveness of integrating SimCSE-based contrastive learning to optimize IndoBERT vector representations for sentiment analysis of the “Free Nutritious Meals” public policy. A comparative experimental approach was employed using an equal number of topics (three topics) and evaluated through BERTopic and Aspect-Based Sentiment Analysis (ABSA). The results demonstrate that the contrastive learning–based model substantially improves cluster separability, indicated by an increase of more than 1000% in the Silhouette Score compared to the baseline model, along with a reduction in topic overlap of approximately 40–50%. In addition, topic keyword diversity increased by more than 75%, yielding more informative and interpretable topic representations. In aspect-based sentiment analysis, the contrastive model exhibited approximately a 50% improvement in sensitivity to sentiment intensity and achieved perfect classification of implicit high-confidence sentiments that were previously misclassified as neutral by the baseline model. These findings confirm that contrastive learning–based embedding optimization effectively addresses the limitations of conventional embeddings and enhances the quality of topic modeling and aspect-based sentiment analysis for Indonesian social media texts.
Penerapan Quantum Machine Learning Untuk Klasifikasi Ulasan Asli Dan Palsu Pada Amazon Krisna Putri Telaumbanua; Sunneng Sandino Berutu; Aninda Astuti
Jutisi : Jurnal Ilmiah Teknik Informatika dan Sistem Informasi Vol 15, No 2 (2026): April 2026
Publisher : STMIK Banjarbaru

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35889/jutisi.v15i2.3498

Abstract

Classifying product reviews on e-commerce platforms, both real and fake, requires a model that can effectively represent text data patterns. This study aims to compare the performance of several Quantum Machine Learning methods, namely QNN, QSVC, and Hybrid Quantum kernel and Classical SVM, in classifying Amazon product reviews. The study uses a quantitative approach with a computational experimental design. Review data is represented using TF-IDF, standardized, and reduced in dimension with Principal Component Analysis (PCA) before being transformed into quantum feature space. Performance evaluation is carried out using accuracy, precision, recall, F1-Score, and MCC metrics. The experimental results show that QNN achieved the best performance with an accuracy value of 85,63%, an F1-Score of 0.9130, and an MCC of 0.5608, while QSVC and the hybrid approach achieved an accuracy of 83.23% with an MCC of 0.4331. These results indicate that QNN has more balanced classification performance.Keywords: Quantum Neural Network; Fake Review Detection; Amazon Reviews; Natural Language Processing; Quantum Machine Learning. AbstrakKlasifikasi ulasan produk pada platform e-commerce, baik ulasan asli maupun palsu, memerlukan model yang mampu merepresentasikan pola data teks secara efektif. Penelitian ini bertujuan untuk membandingkan kinerja beberapa metode Quantum Machine Learning (QML), yaitu QNN, Quantum Support Vector (QSVC), dan Hybrid Quantum kernel and Classical SVM, dalam mengklasifikasikan ulasan produk Amazon. Penelitian menggunakan pendekatan kuantitatif dengan desain eksperimen komputasional. Data ulasan direpresentasikan menggunakan TF-IDF, distandardisasi, dan direduksi dimensinya dengan Principal Component Analysis (PCA) sebelum ditransformasikan ke ruang fitur kuantum. Evaluasi kinerja dilakukan menggunakan metrik accuracy, precision, recall, F1-Score, dan MCC. Hasil eksperimen menunjukkan bahwa QNN memperoleh kinerja terbaik dengan nilai accuracy sebesar 85,63%, F1-Score 0.9130, dan MCC 0.5043, sedangkan QSVC dan pendekatan hybrid mencapai accuracy 83,23% dengan MCC 0,4331. Hasil ini menunjukkan bahwa QNN memiliki performa klasifikasi yang lebih seimbang. 
Analisis Perbandingan Quantum Machine Learning Dalam Klasifikasi Berita Politik Fakta Dan Hoaks Linda Kristiani Zebua; Sunneng Sandino Berutu; Aninda Astuti
Jutisi : Jurnal Ilmiah Teknik Informatika dan Sistem Informasi Vol 15, No 2 (2026): April 2026
Publisher : STMIK Banjarbaru

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35889/jutisi.v15i2.3524

Abstract

This study analyzes the comparative performance of Quantum Machine Learning in classifying factual and hoax political news using three approaches, namely Quantum Neural Network (QNN), Quantum Support Vector Classifier (QSVC), and Hybrid Quantum Kernel with Classical SVM. News data is represented using TF-IDF and its dimensionality is reduced using Principal Component Analysis, then balanced using SMOTE. Feature transformation is carried out to the quantum domain through angle encoding, then applied to the QML model. Performance evaluation is carried out using accuracy, precision, recall, and F1-Score. The experimental results show that QSVC has the best performance with an accuracy of 0.629 and an F1-Score of 0.735, followed by QNN and Hybrid Quantum Kernel Classical SVM. This study proves that the quantum kernel-based approach is effective in classifying medium-dimensional text, while also demonstrating the potential of Quantum Machine Learning as an alternative method for classifying factual and hoax political news.Keywords: Quantum Machine Learning; Quantum Neural Network; Quantum Support Vector Classifier; Hybrid Quantum Kernel; News Classification AbstrakPenelitian ini menganalisis perbandingan kinerja Quantum Machine Learning dalam klasifikasi berita politik fakta dan hoaks dengan menggunakan tiga pendekatan, yaitu Quantum Neural Network (QNN), Quantum Support Vector Classifier (QSVC), dan Hybrid Quantum Kernel dengan Classical SVM. Data berita direpresentasikan menggunakan TF-IDF dan direduksi dimensinya dengan Principal Component Analysis, kemudian diseimbangkan menggunakan SMOTE. Transformasi fitur dilakukan ke domain kuantum melalui angle encoding, kemudian diterapkan pada model QML. Evaluasi kinerja dilakukan menggunakan accuracy, precision, recall, dan F1-Score. Hasil eksperimen menunjukkan QSVC memiliki performa terbaik dengan accuracy 0,629 dan F1-Score 0,735, diikuti QNN dan Hybrid Quantum Kernel Classical SVM. Penelitian ini membuktikan bahwa pendekatan berbasis quantum kernel efektif dalam klasifikasi teks berdimensi sedang, sekaligus menunjukkan potensi Quantum Machine Learning sebagai alternatif metode klasifikasi berita politik fakta dan hoaks. 
Analisis Perbandingan ANN dan Hybrid QNN pada Klasifikasi Multikarakteristik Data Kharisteas Josan Sedi; Sunneng Sandino Berutu; Aninda Astuti
Jutisi : Jurnal Ilmiah Teknik Informatika dan Sistem Informasi Vol 15, No 2 (2026): April 2026
Publisher : STMIK Banjarbaru

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35889/jutisi.v15i2.3485

Abstract

The rapid advancement of machine learning has driven the exploration of various computational models to address classification problems across datasets with diverse characteristics. This study aimed to compare the performance of Artificial Neural Network (ANN) and Hybrid Quantum Neural Network (Hybrid QNN) under controlled experimental conditions. Three benchmark datasets, namely Red Wine Quality, Banknote Authentication, and Leukemia Gene Expression, were used with a training and test data split of 80:20. The evaluation was conducted using accuracy, precision, recall, F1-score, and computational time. The results showed that on the Gene Expression Leukemia dataset, Hybrid QNN achieved an accuracy of 0.9333, significantly outperforming ANN, which obtained 0.6111. Conversely, ANN demonstrated competitive performance and higher computational efficiency on low- and medium-dimensional datasets. These findings indicate that the advantages of Hybrid QNN are contextual and strongly dependent on data characteristics.Keywords: Artificial Neural Network; Hybrid Quantum Neural Network; classification; machine learningAbstrakPerkembangan pesat pembelajaran mesin telah mendorong eksplorasi berbagai model komputasi untuk menyelesaikan permasalahan klasifikasi pada data dengan karakteristik yang beragam. Penelitian ini bertujuan untuk membandingkan kinerja Artificial Neural Network (ANN) dan Hybrid Quantum Neural Network (Hybrid QNN) dalam kondisi eksperimen yang terkontrol. Tiga dataset acuan digunakan, yaitu Red Wine Quality, Banknote Authentication, dan Gene Expression Leukemia, dengan pembagian data latih dan uji sebesar 80:20. Evaluasi dilakukan menggunakan metrik akurasi, precision, recall, F1-score, serta waktu komputasi. Hasil eksperimen menunjukkan bahwa pada dataset Gene Expression Leukemia, Hybrid QNN mencapai akurasi 0,9333, jauh lebih tinggi dibandingkan ANN sebesar 0,6111. Sebaliknya, ANN menunjukkan performa yang kompetitif dan lebih efisien pada dataset berdimensi rendah hingga menengah. Temuan ini menunjukkan bahwa keunggulan Hybrid QNN bersifat kontekstual dan bergantung pada karakteristik data. 
PEMANFAATAN ANALISIS SENTIMEN ULASAN GOOGLE UNTUK MENINGKATKAN KUALITAS LAYANAN PADA UMKM KEDAI TEH KALASAN Sunneng Sandino Berutu; Meisyes Christie Bawuna; Jatmika Jatmika; Riko Gesmani; Jeniwanti Carolina Kotte
ABDI WINA JURNAL PENGABDIAN KEPADA MASYARAKAT Vol. 6 No. 1 (2026): Abdi Wina Edisi Juni 2026
Publisher : Lembaga Penelitian dan Pengabdian Kepada Masyarakat Universitas Kristen Wira Wacana Sumba

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.58300/b426k896

Abstract

Perkembangan teknologi digital mendorong perubahan dalam cara konsumen menyampaikan pengalaman terhadap produk dan layanan, salah satunya melalui ulasan daring pada platform Google Review. Namun, pemanfaatan data ulasan tersebut sebagai dasar evaluasi layanan masih belum optimal pada sebagian pelaku Usaha Mikro, Kecil, dan Menengah (UMKM). Kegiatan pengabdian kepada masyarakat ini bertujuan untuk meningkatkan literasi digital dan kemampuan mitra UMKM Kedai Teh Kalasan dalam memanfaatkan analisis sentimen untuk memahami persepsi pelanggan serta meningkatkan kualitas layanan. Metode pelaksanaan dilakukan melalui lima tahapan, yaitu identifikasi masalah, perencanaan, pelaksanaan, observasi, dan evaluasi dengan pendekatan partisipatif. Data yang digunakan berupa 575 ulasan pelanggan yang diperoleh dari Google Review, kemudian diolah melalui tahapan preprocessing, anotasi aspek, dan analisis sentimen untuk mengklasifikasikan opini pelanggan ke dalam kategori positif, negatif, dan netral. Hasil analisis menunjukkan bahwa aspek rasa dan kenyamanan tempat didominasi oleh sentimen positif, sehingga menjadi faktor utama kepuasan pelanggan. Sementara itu, aspek pelayanan dan harga menunjukkan variasi persepsi, serta aspek penyajian masih memiliki peluang untuk ditingkatkan. Kegiatan ini memberikan dampak positif berupa peningkatan pemahaman mitra dalam mengolah dan memanfaatkan data ulasan pelanggan secara sistematis. Dengan demikian, analisis sentimen dapat menjadi alat yang efektif dalam mendukung pengambilan keputusan berbasis data untuk peningkatan kualitas layanan UMKM secara berkelanjutan.
Identifikasi Tingkat Intensitas Opini dalam Analisis Sentimen Berbasis Aspek Menggunakan Enhanced Triplet Extraction Gabriel Jimmy Richardo Chastelo B; Sunneng Sandino Berutu; Haeni Budiati
Bulletin of Computer Science Research Vol. 6 No. 3 (2026): April 2026
Publisher : Forum Kerjasama Pendidikan Tinggi (FKPT)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47065/bulletincsr.v6i3.1074

Abstract

Conventional sentiment analysis often overlooks variations in the intensity of opinions within text reviews. This is due to the limitations of the Aspect-Based Sentiment Analysis (ABSA) approach, which is restricted to three main triplet components. This study aims to develop and expand the Aspect-Sentiment-Opinion Triplet Extraction (ASOTE) framework to extract entity relationships and sentiment polarity by integrating opinion intensity detection. This study implements the ABSA approach by expanding the triplet structure into four components: aspect, opinion, intensifier, and sentiment (Enhanced Triplet). Data was collected via web scraping of Twitter (X) comments related to the Free Nutritious Meals program, which served as a case study to test the model’s ability to analyze public sentiment. The data then undergoes pre-processing and BIO Tagging, and is classified using a fine-grained sentiment approach to capture the nuances of emotional intensity in greater detail. A Transformer-based model, namely IndoBERT, was used to understand the context and intensity of meaning in the Indonesian language. Evaluation results on the test data show that the model achieved an accuracy of 88% and an average F1-score of 0.88 in sentiment polarity classification between entities, indicating strong model performance. These results demonstrate that providing a framework that is more sensitive to the intensity of opinions when classifying the nuances of public sentiment is a highly effective solution. 
Implementasi dan Pelatihan SIWAREK serta Solar Home System bagi Pokdarwis Opak Grembyangan Emerita Setyowati; Sunneng Sandino Berutu; Susi Siswati
Jurnal Atma Inovasia Vol. 6 No. 1 (2026)
Publisher : Lembaga Penelitian dan Pengabdian pada Masyarakat

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.24002/jai.v6i1.12870

Abstract

 The Tourism Awareness Group or Kelompok Sadar Wisata (Pokdarwis) of Mutihan, Madurejo Village, Prambanan, is pioneering the development of a tourism area along the Opak River. The main challenges faced include the lack of lighting facilities and difficulties in operating food stalls, although the buildings are already available. The community service team, in collaboration with local partners, initiated two main programs: the installation of a Solar Home System (SHS) to provide access to clean energy and reduce dependency on the state electricity grid (PLN), and training on financial management and digitalization for stall operations. The implementation methods included program socialization, financial management training, development of a stall information system, SHS installation, SHS operation training, as well as program monitoring and evaluation. As a result, a stall financial system named Sistem Informasi Keuangan Warung Opak (SIWAREK) was developed, and two SHS units were installed at the tourism site to supply renewable energy. The financial management and information system training provide a strong foundation for Pokdarwis to operate the food stalls. The SHS can be utilized for night lighting, charging visitors’ mobile phones and laptops, and also serve as a tourism icon showcasing the use of renewable energy
Implementation of Latent Dirichlet Allocation Topic Modeling and VADER on Aspect-Based Sentiment Analysis Kevin Miracle Satoko Kevin; Sunneng Sandino Berutu; Jatmika; Retno Palupi
Infact: International Journal of Computers Vol. 10 No. 01 (2026): Journal of Science and Computers
Publisher : Universitas Kristen Immanuel

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.61179/infact.v10i01.708

Abstract

Aspect-Based Sentiment Analysis on a Product or Service is Crucial for Enhancing Customer Satisfaction. This Study Applies Latent Dirichlet Allocation (LDA) Topic Modeling to Identify Aspects. Then, the Valence Aware Dictionary and Sentiment Reasoner (VADER) Lexicon Method is Adopted to Determine Sentiment on These Aspects. The Data Source Comes from Customer Reviews of a Gelato Ice Cream Shop at Taman Siswa. Data was collected from Google Maps Using the Web Scraping Method via the Instant Data Scrapper Application. The Experimental Results Show that the LDA Method Identified 3 Aspects: Flavor, Place, and Service. Meanwhile, Sentiment Measurement Using VADER on the Flavor Aspect Revealed a Positive Sentiment of 213%, Negative Sentiment of 60%, and Neutral Sentiment of 218%. The Next Aspect, Place, Had a Positive Sentiment of 32%, Negative Sentiment of 4%, and Neutral Sentiment of 4%, while the Service Aspect Composed of 32% Positive Sentiment, 3% Negative Sentiment, and 3% Neutral Sentiment. Overall, the Positive Sentiment for the Flavor Aspect (Taste) Outweighed the Negative and Neutral Sentiments for the Place (Location) and Service (Service) Aspects.