Garuda - Garba Rujukan Digital

Bibliometric Analysis: Machine Learning untuk Blended Learning Agus Bahtiar; Mulyawan
Prosiding SISFOTEK Vol 8 No 1 (2024): SISFOTEK VIII 2024
Publisher : Ikatan Ahli Informatika Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar

Blended learning, which combines face-to-face learning methods with digital technology, has grown rapidly thanks to advances in information technology. Along with that, machine learning technology offers great potential to improve personalization and adaptation in blended learning. This research aims to explore the application of machine learning in blended learning systems through bibliometric analysis. By analyzing SCOPUS indexed publications from 2019 to 2024, this study identifies trends, challenges and opportunities in the integration of machine learning with blended learning. The methods used include search keyword definition, initial data collection, refinement of search results, statistical compilation, and data analysis. The main findings show that there is a significant increase in the number of publications on this topic, with the highest peak in 2022. The wide distribution of publications indicates significant international collaboration. Citation analysis indicates that the quality and impact of research is also increasing, with recent publications gaining more citations. This research highlights the importance of applying machine learning in blended learning to improve educational effectiveness and support the development of more adaptive learning methods. The findings provide valuable insights for academics and practitioners to encourage further innovation and improve the quality of education in the digital era.

Improving the School Type Clustering Model on the Foundation Using the K-Means Algorithm (Case Study: Kebon Kelapa Al-Ma'rifah, Cirebon Regency) Hanifah Nur Aulia; Martanto; Arif Rinaldi Dikananda; Mulyawan
Journal of Artificial Intelligence and Engineering Applications (JAIEA) Vol. 4 No. 2 (2025): February 2025
Publisher : Yayasan Kita Menulis

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.59934/jaiea.v4i2.739

This study aims to improve the school type grouping model at the Kebon Kelapa Al-Ma'rifah Foundation, Cirebon Regency, using the K-Means algorithm. Data-based grouping is very important in supporting efficient education management, especially in environments that have various types of schools such as Madrasah Aliyah (MA), Vocational High School (SMK), Madrasah Tsanawiyah (MTs), and Madrasah Ibtidaiyah (MI). The data used comes from the New Student Registration (PPDB) dataset for the 2023–2024 school year, with demographic attributes such as name, place of birth, gender, and time of school entry. The evaluation of clustering quality was carried out using the Davies-Bouldin Index (DBI) to determine the optimal number of clusters. The results show that the optimal number of clusters is K=5 with the lowest DBI value of 0.201, which results in compact and well-separated clusters. The implementation of the K-Means algorithm helps the foundation understand the distribution pattern of students based on attributes such as gender, region, and entry time. This research provides practical benefits, including more targeted resource allocation, improved quality of education, and efficiency in school management. In addition, this research contributes to the development of data mining models in the education sector and opens up opportunities for the exploration of additional attributes such as academic achievement and socioeconomic conditions. Further research is suggested to use alternative algorithms such as K-Medoids or DBSCAN.

The Improvement of Indonesian Film Genre Clustering Model Using the K-Means Algorithm in Film Production Decision-Making Wiratriyana; Martanto; Arif Rinaldi Dikananda; Mulyawan
Journal of Artificial Intelligence and Engineering Applications (JAIEA) Vol. 4 No. 2 (2025): February 2025
Publisher : Yayasan Kita Menulis

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.59934/jaiea.v4i2.765

The Indonesian film industry is expanding rapidly, but understanding audience preferences remains a significant challenge for producers. This study aims to cluster Indonesian films by genre and synopsis using the K-Means algorithm to aid in marketing strategies and content development. The dataset comprises 1,271 Indonesian film entries, including attributes like release year, genre, synopsis, and user ratings. The research follows the Knowledge Discovery in Databases (KDD) framework, which involves data selection, preprocessing, transformation, clustering with K-Means, and evaluation using the Elbow method to identify the optimal number of clusters. The results show that the K-Means algorithm successfully grouped the films into three clusters: drama, horror, and others. The analysis indicates that drama films dominate the high-rating cluster, while horror films are more commonly found in the low-rating category. The use of Principal Component Analysis (PCA) in the visualization aids in interpreting the clustering results, providing a clearer view of the data distribution. These findings highlight the potential for improving film production strategies by aligning content with audience preferences. By understanding genre patterns and ratings, producers can make more informed decisions in marketing and content development.

Development of Educational Game for Introduction Animal Types Using the ADDIE Method Smart Apps Creator In Improving Knowledge Students Artoti, Azzahra Rizky; Martanto; Dikananda, Arif Rinaldi; Mulyawan
Journal of Artificial Intelligence and Engineering Applications (JAIEA) Vol. 4 No. 2 (2025): February 2025
Publisher : Yayasan Kita Menulis

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.59934/jaiea.v4i2.777

The development of technology in education opens up opportunities for innovation to create interactive learning media, especially for early childhood. This research aims to develop educational games based on Smart Apps Creator using the ADDIE method to introduce animal species to Al-Washliyah kindergarten students. The method used is ADDIE, consisting of five stages, namely: Analysis, Design, Development, Implementation, and Evaluation. in this study conducted validity, reliability, normality, homogeneity, and anova tests to measure the effectiveness of this learning media. The results showed that this animal species recognition educational game succeeded in improving student understanding with an average score before the use of learning media of 59.2% increasing to 87.73% after using learning media. Validity and reliability tests show that this learning media meets the criteria of effective, easy-to-use, and interesting learning media.

Clustering Data Penjualan Produk Makanan pada Toko Toserba Yogya Siliwangi dengan Menggunakan Metode K-Means Noviati; Mulyawan; Kurnia, Dian Ade; Rinaldi, Ade Rizki
MEANS (Media Informasi Analisa dan Sistem) Volume 7 Nomor 1
Publisher : LPPM UNIKA Santo Thomas Medan

Product availability is one of the important factors to increase sales and maintain customer satisfaction in meeting their needs. With this, the company needs to analyze sales data, both for the best-selling products or those that are not selling well from sales reports every month, especially for food products. Of course, this is not easy, especially for a large enough retailer such as the Yogya Siliwangi Toserba which has thousands of product items and thousands of sales data every month. The above problems can be solved by grouping the data using the k-means clustering algorithm on rapidminer with variables taken by the name of goods, incoming goods, outgoing goods and stock. The goal is to maximize sales and maintain product stock availability to meet the diverse needs of consumers. From the calculation of the k-means algorithm using the rapidminer application, the results obtained are in the form of three clusters, cluster_1 3 items, cluster_2 13 items and cluster_0 454 items with Devies Bouldin results being 0.478.

Pengelolaan Aplikasi Layanan Administrasi pada Kelurahan Argasunya Garsandi, Akmal Maulana; Mulyawan; Dana, Raditya Danar; Kaslani; Tohidi, Edi
MEANS (Media Informasi Analisa dan Sistem) Volume 7 Nomor 1
Publisher : LPPM UNIKA Santo Thomas Medan

Argasunya Village is a combined area of several Rukun Warga (RW). Government at the village and sub-district levels is an element of government that is directly related to the community. Kelurahan is also an administrative division under the subdistrict. The village office has not used the system in making a Business Certificate. Therefore, the author tries to make a service application for making a business certificate based on a web application with the hope of simplifying the procedures for making a business certificate, the author uses data sources (Observation and Interviews). The programming language used is PHP language with MySql database. For system design, the author uses the Prototype method. From the results of the tests carried out, the system functionality can run well and the making of online and onsite business certificates is appropriate and can run well..

Potensi Ekstrak Bunga Kecombrang (Etlingera elatior) dan Tanaman Mimosa pudica L. sebagai Edible Coating untuk Memperpanjang Masa Simpan pada Buah Apel: Potential of Kecombrang Flower Extract (Etlingera elatior) and Plants Mimosa pudica L. as an Edible Coating to Extend the Shelf Life of Apples Mulyawan; Dian Indratmi; Erfan Dani Septia; Yusufa Alif Hidayat; Rovi Amallia Malikah
Jurnal Hortikultura Indonesia (JHI) Vol. 16 No. 1 (2025): Jurnal Hortikultura Indonesia (JHI)
Publisher : Indonesian Society for Horticulture / Department of Agronomy and Horticulture

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.29244/jhi.16.1.58-69

The apple industry in Indonesia, particularly in East Java Province, faces significant challenges related to fruit spoilage, which can result in substantial economic losses. One potential solution to address this issue is the application of natural-based edible coatings. This study aims to examine the effects of combining torch ginger (Etlingera elatior) flower extract and mimosa (Mimosa pudica) plant extract as the main ingredients in edible coatings on the quality and shelf life of apples. A Completely Randomized Design (CRD) was employed with four treatments: no coating (P0), coating with 2% extract (P1), 4% extract (P2), and 6% extract (P3). The results revealed that the combination of these extracts contained antimicrobial compounds such as dodecanal, octane, 1,1-diethoxy-, squalene, and methyl stearate, which effectively inhibited spoilage. The P3 treatment (6% extract) proved the most effective in maintaining apple quality, as indicated by stable weight, firmness, and sustained sugar and vitamin C content during storage. Keywords: post-harvest, storage capacity, secondary metabolites

Analisis Dan Prediksi Risiko Kelahiran Bayi Menggunakan K-Means Dan Deep Neural Network (DNN) Mukhlisin Ilahudin; Nana Suarna; Agus Bahtiar; Mulyawan; Irfan Ali
Jurnal Sistem Informasi dan Teknologi Vol 6 No 1 (2026): Jurnal Sistem Informasi dan Teknologi (SINTEK)
Publisher : LPPM STMIK KUWERA

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.56995/sintek.v6i1.206

Risiko kelahiran bayi merupakan indikator penting dalam evaluasi kesehatan ibu dan anak sehingga diperlukan pendekatan analitis yang mampu mengidentifikasi pola risiko secara akurat. Penelitian ini bertujuan menganalisis dan memprediksi risiko kelahiran bayi dengan mengintegrasikan metode K-Means dan Deep Neural Network (DNN). Dataset yang digunakan terdiri dari 983 data rekam medis ibu hamil yang telah melalui tahap pengumpulan data, pembersihan, dan preprocessing meliputi normalisasi, encoding variabel kategorikal, penanganan outlier, serta seleksi fitur. Metode K-Means digunakan untuk mengelompokkan data berdasarkan kemiripan karakteristik klinis guna membentuk representasi pola risiko awal, yang selanjutnya digunakan sebagai fitur tambahan pada model DNN. Model DNN dirancang menggunakan beberapa hidden layer dengan fungsi aktivasi ReLU dan regularisasi dropout. Hasil pengujian menunjukkan bahwa model menghasilkan akurasi sebesar 61,93% dan nilai ROC AUC sebesar 0,6402, yang mengindikasikan performa moderat dalam memprediksi risiko kelahiran bayi. Stabilitas kurva loss dan akurasi menunjukkan proses pelatihan yang berjalan dengan baik tanpa overfitting signifikan. Secara praktis, model ini berpotensi digunakan sebagai alat bantu awal bagi tenaga kesehatan dalam mengidentifikasi ibu hamil dengan risiko kelahiran lebih tinggi sehingga dapat dilakukan pemantauan dan intervensi lebih dini.

Analisis Dan Prediksi Risiko Kelahiran Bayi Menggunakan K-Means Dan Deep Neural Network (DNN) Mukhlisin Ilahudin; Nana Suarna; Agus Bahtiar; Mulyawan; Irfan Ali
Jurnal Sistem Informasi dan Teknologi Vol 6 No 1 (2026): Jurnal Sistem Informasi dan Teknologi (SINTEK)
Publisher : LPPM STMIK KUWERA

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.56995/sintek.v6i1.206

Risiko kelahiran bayi merupakan indikator penting dalam evaluasi kesehatan ibu dan anak sehingga diperlukan pendekatan analitis yang mampu mengidentifikasi pola risiko secara akurat. Penelitian ini bertujuan menganalisis dan memprediksi risiko kelahiran bayi dengan mengintegrasikan metode K-Means dan Deep Neural Network (DNN). Dataset yang digunakan terdiri dari 983 data rekam medis ibu hamil yang telah melalui tahap pengumpulan data, pembersihan, dan preprocessing meliputi normalisasi, encoding variabel kategorikal, penanganan outlier, serta seleksi fitur. Metode K-Means digunakan untuk mengelompokkan data berdasarkan kemiripan karakteristik klinis guna membentuk representasi pola risiko awal, yang selanjutnya digunakan sebagai fitur tambahan pada model DNN. Model DNN dirancang menggunakan beberapa hidden layer dengan fungsi aktivasi ReLU dan regularisasi dropout. Hasil pengujian menunjukkan bahwa model menghasilkan akurasi sebesar 61,93% dan nilai ROC AUC sebesar 0,6402, yang mengindikasikan performa moderat dalam memprediksi risiko kelahiran bayi. Stabilitas kurva loss dan akurasi menunjukkan proses pelatihan yang berjalan dengan baik tanpa overfitting signifikan. Secara praktis, model ini berpotensi digunakan sebagai alat bantu awal bagi tenaga kesehatan dalam mengidentifikasi ibu hamil dengan risiko kelahiran lebih tinggi sehingga dapat dilakukan pemantauan dan intervensi lebih dini.

Analisis Sentimen Ulasan Pengguna Aplikasi E-Commerce Toco Menggunakan Algoritma Naive Bayes Purnamasari, Adinda; Astuti, Rini; Anam, Khaerul; Gifthera Dwilestari; Mulyawan
Jurnal Ilmiah Sistem Informasi (JISI) Vol. 5 No. 1 (2026): MARET
Publisher : Universitas Muhammadiyah Metro

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.24127/jisi.v5i1.10627

This study aims to analyze the sentiment of user reviews of the Toco e-commerce application using the Naïve Bayes Multinomial algorithm. The dataset consists of 1425 reviews with an unbalanced distribution between positive and negative classes. Review data was collected from the Google Play Store platform, then processed automatically through the stages of case folding, normalization, stopword removal, and stemming. Modeling was carried out by dividing the data into training and test data, and classifying sentiment using the Naïve Bayes approach. From the evaluation results, the model's accuracy in sentiment classification reached 88%, with higher performance achieved in the majority class (positive) compared to the minority class (negative), as reflected in the low precision and recall values. This study emphasizes the need to handle unbalanced data so that the analysis results reflect the diverse perceptions of users. This research provides a baseline for sentiment analysis for local e-commerce applications and contributes to the development of automated analytics systems to support decision-making in the Indonesian e-commerce industry.

Title

Found 10 Documents
Search

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Title Search

Found 10 Documents Search

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Title

Found 10 Documents
Search