Article Per Year (5 Year)
p-Index From 2021 - 2026
10.868
P-Index
This Author published in this journals
All Journal Techno.Com: Jurnal Teknologi Informasi Jurnal Buana Informatika Jurnal Informatika Jurnal Teknologi Informasi dan Ilmu Komputer JUITA : Jurnal Informatika Jurnas Nasional Teknologi dan Sistem Informasi POSITIF Edu Komputika Journal Sistemasi: Jurnal Sistem Informasi Jurnal Pendidikan Informatika dan Sains Jurnal RESTI (Rekayasa Sistem dan Teknologi Informasi) Computatio : Journal of Computer Science and Information Systems RABIT: Jurnal Teknologi dan Sistem Informasi Univrab Jurnal Khatulistiwa Informatika JIKO (Jurnal Informatika dan Komputer) JURNAL MEDIA INFORMATIKA BUDIDARMA Jurnal Pilar Nusa Mandiri JTERA (Jurnal Teknologi Rekayasa) Jurnal Sains dan Informatika INOVTEK Polbeng - Seri Informatika Matrix : Jurnal Manajemen Teknologi dan Informatika JURNAL REKAYASA TEKNOLOGI INFORMASI SINTECH (Science and Information Technology) Journal Jurnal Informatika Universitas Pamulang Jurnal Teknoinfo Jurnal Sisfokom (Sistem Informasi dan Komputer) KACANEGARA Jurnal Pengabdian pada Masyarakat MATRIK : Jurnal Manajemen, Teknik Informatika, dan Rekayasa Komputer Indonesian Journal of Applied Informatics KOMPUTIKA - Jurnal Sistem Komputer KOMPUTA : Jurnal Ilmiah Komputer dan Informatika JIPI (Jurnal Ilmiah Penelitian dan Pembelajaran Informatika) Jurnal Teknologi Terapan Jurnal Teknologi Terpadu EDUMATIC: Jurnal Pendidikan Informatika JISICOM (Journal of Information System, Infomatics and Computing) EVOLUSI : Jurnal Sains dan Manajemen Building of Informatics, Technology and Science JASIEK (Jurnal Aplikasi Sains, Informasi, Elektronika dan Komputer) JISKa (Jurnal Informatika Sunan Kalijaga) Jurnal Teknologi Informasi dan Multimedia Jurnal JTIK (Jurnal Teknologi Informasi dan Komunikasi) JUSTINDO (Jurnal Sistem dan Teknologi Informasi Indonesia) JISA (Jurnal Informatika dan Sains) International Journal of Engineering, Technology and Natural Sciences (IJETS) JOURNAL OF INFORMATION SYSTEM RESEARCH (JOSH) Jurnal Sistem Komputer dan Informatika (JSON) TIN: TERAPAN INFORMATIKA NUSANTARA Jurnal Teknik Informatika (JUTIF) Jurnal Digit : Digital of Information Technology Jurnal FASILKOM (teknologi inFormASi dan ILmu KOMputer) Science in Information Technology Letters Journal of Soft Computing Exploration Jurnal Indonesia : Manajemen Informatika dan Komunikasi Bulletin of Computer Science Research KLIK: Kajian Ilmiah Informatika dan Komputer International Journal Software Engineering and Computer Science (IJSECS) Jurnal Sains dan Teknologi Jurnal Sains dan Teknologi International Journal Science and Technology (IJST) Malcom: Indonesian Journal of Machine Learning and Computer Science Journal of Scientific Research, Education, and Technology Jikom: Jurnal Informatika dan Komputer Journal of Data Science Theory and Application NERO (Networking Engineering Research Operation) SmartComp Jurnal Ilmu Komputer dan Sistem Informasi Jurnal Indonesia : Manajemen Informatika dan Komunikasi Emitor: Jurnal Teknik Elektro IJISCS (International Journal of Information System and Computer Science)
Donny Avianto
Universitas Teknologi Yogyakarta
Aerospace Engineering Agriculture, Biological Sciences & Forestry Arts Automotive Engineering Biochemistry, Genetics & Molecular Biology Chemical Engineering, Chemistry & Bioengineering Chemistry Civil Engineering, Building, Construction & Architecture Computer Science & IT Control & Systems Engineering Decision Sciences, Operations Research & Management Dentistry Earth & Planetary Sciences Economics, Econometrics & Finance Education Electrical & Electronics Engineering Energy Engineering Environmental Science Health Professions Immunology & microbiology Industrial & Manufacturing Engineering Languange, Linguistic, Communication & Media Law, Crime, Criminology & Criminal Justice Library & Information Science Materials Science & Nanotechnology Mathematics Mechanical Engineering Medicine & Pharmacology Neuroscience Nursing Physics Public Health Social Sciences Veterinary Other

Published : 91 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 91 Documents
Search

 5   6   7   8   9 
PENERAPAN ALGORITMA K-NEAREST NEIGHBOR UNTUK DETEKSI DINI STATUS GIZI PASIEN DEWASA Wijayanti, Dian; Hermawan, Arief; Avianto, Donny
POSITIF : Jurnal Sistem dan Teknologi Informasi Vol 10 No 2 (2024): Positif : Jurnal Sistem dan Teknologi Informasi
Publisher : P3M Politeknik Negeri Banjarmasin

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.31961/positif.v10i2.2255

Abstract

Assessing the nutritional status of adult patients is essential to gain a comprehensive understanding of their condition and assist healthcare workers in planning appropriate treatment. However, manual assessment is time-consuming and labor-intensive, especially when the number of patients exceeds the number of available healthcare workers. This can hinder the timely and accurate delivery of nutritional care. The K-Nearest Neighbor (KNN) algorithm is a commonly used method for nutritional status classification, particularly in toddlers, pregnant women, or for obesity classification in adults. The use of KNN for early detection of adult nutritional status remains rarely explored. This study applies the KNN algorithm to classify the nutritional status of adult patients using data from the Alamanda 1 ward and the ICU ward at Sleman Regional General Hospital, collected from January 2 to October 18, 2023. The dataset includes patient height, weight, and nutritional status. The algorithm was implemented using RapidMiner with odd k-values less than 20, and data splits of 90:10, 70:30, and 50:50 for training and testing. Results show that the optimal k-values for the highest accuracy were k = 1 and k = 3 using the 70:30 data split, both achieving an accuracy of 96.77%. The highest sensitivity, 97.61%, was also achieved at k = 3 with the same data split. The KNN algorithm demonstrates strong potential to be developed into an early detection system for assessing the nutritional status of adult patients in hospitals, supporting faster and more accurate nutritional care services
The Impact of Extreme Data Imbalance on Evaluation Metrics of Deep Learning Models for Loan Default Prediction Budiyanto, Irfan; Hermawan, Arief; Avianto, Donny; Kusban, Muhammad
Emitor: Jurnal Teknik Elektro Vol 25, No 2: July 2025
Publisher : Universitas Muhammadiyah Surakarta

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.23917/emitor.v25i2.10719

Abstract

The growth of financial technology has made online loans more accessible, but it has also increased the risk of borrowers failing to repay. Developing a reliable system to predict loan defaults is therefore very important. A common problem in these predictions is an imbalance in the data – there are far fewer cases of loan defaults (the minority class) than loans that are paid back on time (the majority class). This imbalance can cause the prediction models to be biased. This research specifically investigates the effect of an extremely increased data imbalance ratio (from 1:170 to 1:33,612) on the evaluation metrics of a Deep Neural Network (DNN) model, particularly when using the Adaptive Synthetic Sampling (ADASYN) oversampling technique. The method used involves adopting a previous research approach that combines ADASYN to handle data imbalance and DNN for prediction, applied to an updated Lending Club dataset with a more severe level of imbalance. The results demonstrate a critical breakdown in key evaluation metrics. Compared to previous research, Accuracy remains high (0.9515) and Specificity is strong (0.9516). However, there is a catastrophic decrease in Precision to almost zero (0.0001), a very low Recall (0.1667), and a resulting F1-Score that is also nearly zero (0.0002). A visual analysis using Principal Component Analysis (PCA) reveals that this decline in Precision is caused by synthetic minority samples generated by ADASYN completely overlapping with the original majority cluster, leading to a massive number of false positives. In conclusion, ADASYN fails to maintain a usable performance level under extreme imbalance conditions, rendering the model ineffective for its intended purpose and highlighting the critical need for alternative strategies when dealing with severe minority class scarcity.
Klasifikasi Citra Produk Chiffon Cake Dengan Metode K-Nearest Neighbors Dan Grey Level Co-Occurrence Matrix Untuk Quality Control Novaldy, Olwin Kirab; Hermawan, Arief; Avianto, Donny
Jurnal Teknologi Informasi dan Ilmu Komputer Vol 12 No 4: Agustus 2025
Publisher : Fakultas Ilmu Komputer, Universitas Brawijaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.25126/jtiik.124

Abstract

Chiffon cake adalah salah satu kue yang populer, dan kepuasan pelanggan sangat dipengaruhi oleh kualitas produk chiffon cake. Oleh karena itu, diperlukan sistem pengendalian kualitas yang efektif untuk mendeteksi chiffon cake yang cacat. Dalam penelitian ini, digunakan metode K-Nearest Neighbors (KNN) sebagai alat klasifikasi untuk mendeteksi produk chiffon cake yang cacat. Tujuan utama penelitian ini adalah membuat model sistem pengendalian kualitas yang dapat mengklasifikasikan chiffon cake secara otomatis ke dalam dua kategori: "lolos" dan "cacat." Sistem ini diharapkan dapat meningkatkan efisiensi proses produksi dan mengurangi kemungkinan produk cacat sampai ke tangan konsumen. Studi ini menggunakan KNN dan Gray Level Co-Occurrence Matrix (GLCM). Metode klasifikasi KNN bergantung pada pemilihan tetangga terdekat data untuk menentukan kategori kelasnya, sedangkan GLCM adalah teknik ekstraksi fitur yang digunakan untuk mengukur tekstur gambar dengan menganalisis hubungan antara dua piksel dalam orde kedua. Untuk melatih model KNN, studi ini menggunakan dataset yang diambil sendiri melalui pemotretan produk chiffon, kemudian diberi label "lolos" dan "cacat." Setelah melatih model, penulis melakukan pengujian dengan data uji untuk mengevaluasi kinerjanya. Hasil penelitian menunjukkan bahwa penerapan KNN memungkinkan pengklasifikasian chiffon cake dengan akurasi 90,4%. Validasi lebih lanjut dengan dataset yang lebih besar dan beragam diperlukan untuk memastikan bahwa model tetap robust dan dapat diandalkan dalam berbagai kondisi produksi.   Abstract Chiffon cake is one of the popular types of cake, and customer satisfaction is greatly influenced by the quality of the chiffon cake. Therefore, an effective quality control system is necessary to detect defective chiffon cakes. In this research, the K-Nearest Neighbors (KNN) method is used as a classification tool to detect defective chiffon cakes. The main objective of this study is to create a quality control system model that can automatically classify chiffon cakes into two categories: "Pass" and "Not Pass." This model is expected to increase production efficiency and reduce the risk of defective products reaching consumers. This study uses KNN and the Gray Level Co-Occurrence Matrix (GLCM). The KNN classification method determines the class category by selecting the data's nearest neighbors, while GLCM is a feature extraction method that measures image texture by analyzing the correlation between two pixels in the second order. To train the KNN model, this study used a dataset of manually photographed products, labeled as "lolos" and "cacat" After training the model, this study evaluates its performance using test data. The research results showed that the KNN application can classify chiffon cakes with an accuracy of 90.4%. Further validation with larger and more diverse datasets is recommended to enhance the model's robustness and applicability.
A HYBRID ARIMA-MLP ALGORITHM USING ARIMA AND MLP TO IMPROVE ESTIMATION MODEL PERFORMANCE IN SOLAR RADIATION SENSOR DATA Syahab, Alfin Syarifuddin; Hermawan, Arief; Avianto, Donny
IJISCS (International Journal of Information System and Computer Science) Vol 7, No 3 (2023): IJISCS (International Journal of Information System and Computer Science)
Publisher : Bakti Nusantara Institute

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.56327/ijiscs.v7i3.1617

Abstract

Ground-based solar radiation measurements help solar energy projects and applications. Various models have been developed to estimate solar radiation. Then, several additional models were created using improved machine learning. Currently, estimating solar radiation with the help of hybrid models is more efficient. In this research, the basic concepts of modeling procedures for hybrid between the Autoregressive Integrated Moving Average (ARIMA) and the Multilayer Perceptron (MLP) are used to improve the performance of the ARIMA and MLP models in estimating solar radiation data from a pyranometer sensor installed on the automatic weather station (AWS) at Stasiun Klimatologi Daerah Istimewa Yogyakarta.  The test results of the estimation model based on the coefficient of determination (R2) value and root mean square error (RMSE) show that the ARIMA model can provide a high coefficient of determination value in each data splitting scenario. The MLP estimation model shows a coefficient of determination value that is lower than the ARIMA model. On the other hand, MLP is able to improve the RMSE value in the ARIMA model in 70:30 and 90:10 splitting data. Furthermore, the ARIMA-MLP hybrid estimation model is able to improve the RMSE value of the ARIMA and MLP models even though the coefficient of determination value is not as good as the ARIMA model. This research shows that the ARIMA-MLP hybrid model is able to contribute to increasing the accuracy value in RMSE compared to the ARIMA and MLP models in estimating solar radiation sensor data.
Media Pembelajaran Aksara Jawa untuk Anak Sekolah Dasar Menggunakan Augmented Reality Hanif, Rifqi Fadhlurrahman; Avianto, Donny
Jurnal Informatika Universitas Pamulang Vol 9 No 1 (2024): JURNAL INFORMATIKA UNIVERSITAS PAMULANG
Publisher : Teknik Informatika Universitas Pamulang

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32493/informatika.v9i1.39598

Abstract

Learning the Javanese script used in the Javanese language is challenging, especially for those who are just starting to learn the language. Students and the general public rarely use or are interested in Javanese script because of its difficulty. In order to protect the local language, researchers came up with the idea of teaching Javanese script to students, especially students who are just starting to learn in elementary school. This is done by creating a learning media application that is interesting, easy to understand, and entertaining. This Javanese script learning application is made using Augmented Reality (AR) technology and penanda-based tracking mechanism. This program is a learning tool that can turn 2D objects into 3D objects and bring the virtual world into the real world because it uses Augmented Reality technology. When used with the app's quiz feature, this educational tool can help elementary school students to improve their memory and skills, as well as inspire them to learn more through engaging instructions. The app was tested through blackbox testing to demonstrate its feasibility, with 100% successful results for all intended buttons and functions. AR camera distance testing resulted in a 90% success rate in detecting penandas at a certain distance.
RECOGNITION OF REAL-TIME HANDWRITTEN CHARACTERS USING CONVOLUTIONAL NEURAL NETWORK ARCHITECTURE Gumilang, Muhammad Satrio; Avianto, Donny
Jurnal Teknik Informatika (Jutif) Vol. 4 No. 5 (2023): JUTIF Volume 4, Number 5, October 2023
Publisher : Informatika, Universitas Jenderal Soedirman

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.52436/1.jutif.2023.4.5.993

Abstract

Pattern recognition, including handwriting recognition, has become increasingly common in everyday life, as is recognizing important files, agreements or contracts that use handwriting. In handwriting recognition, there are two types of methods commonly used, namely online and offline recognition. In online recognition, handwriting patterns are associated with pattern recognition to generate and select distinctive patterns. In handwritten letter patterns, machine learning (deep learning) is used to classify patterns in a data set. One of the popular and accurate deep learning models in image classification is the convolutional neural network (CNN). In this study, CNN will be implemented together with the OpenCV library to detect and recognize handwritten letters in real-time. Data on handwritten alphabet letters were obtained from the handwriting of 20 students with a total of 1,040 images, consisting of 520 uppercase (A-Z) images and 520 lowercase (a-z) images. The data is divided into 90% for training and 10% for testing. Through experimentation, it was found that the best CNN architecture has 5 layers with features (32, 32, 64, 64, 128), uses the Adam optimizer, and conducts training with a batch size of 20 and 100 epochs. The evaluation results show that the training accuracy is between 85, 90% to 89.83% and testing accuracy between 84.00% to 87.00%, with training and testing losses ranging from 0.322 to 0.499. This research produces the best CNN architecture with training and testing accuracy obtained from testing. The developed CNN model can be used as a reference or basis for the development of more complex handwriting pattern recognition models or for pattern recognition in other domains, such as object recognition in computer vision, facial recognition, and other object detection.
Fuzzy Mamdani for Equality of Employee Salary Fakharudin, Panji Rangga Adzan Fajar; Avianto, Donny
Sistemasi: Jurnal Sistem Informasi Vol 13, No 1 (2024): Sistemasi: Jurnal Sistem Informasi
Publisher : Program Studi Sistem Informasi Fakultas Teknik dan Ilmu Komputer

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32520/stmsi.v13i1.3621

Abstract

Every job in a coffee shop has a salary. Salary is a form of recognition or imbalance for the results achieved. Salary is often also called wages, which is an imbalance in the services provided regularly to employees. Fuzzy has several methods, one of which is Fuzzy Mamdani which is used to make inferences or take the best decision in a problem that has subtle values. In 1975, Ebrahim Mamdani proposed the Fuzzy Mamdani method. Fuzzy Mamdani is a method that uses linguistic rules and has a fuzzy algorithm so that it can be explained mathematically and is easy to understand. The input values for the criteria for length of work are 8, experience 8, and dependents 4. The output is based on Fuzzy Mamdani's calculations, the employee gets a salary of IDR 2.21 million. The results of this study research that the wage income offered by Warkop IN`DA to its employees is good and not far from the minimum wage of the coffee shop. This third variable has a big influence on the final calculation results. The system created is quite good because the MAE result is 0.567 and the MAPE result is 36.720%.
PENGENALAN CITRA RAMBU LALU LINTAS MENGGUNAKAN EKSTRAKSI FITUR MOMENWARNA DAN K-NEAREST NEIGHBOR Rizarta, Rusma Eko Fiddy; Avianto, Donny
Computatio : Journal of Computer Science and Information Systems Vol. 3 No. 1 (2019): COMPUTATIO : JOURNAL OF COMPUTER SCIENCE AND INFORMATION SYSTEMS
Publisher : Faculty of Information Technology, Universitas Tarumanagara

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.24912/computatio.v3i1.4272

Abstract

The traffic signs are signs with specific shape and symbols, letters, numbers, or words which have the aim to warn or inform the road users. However, there are many road users who are not aware of the meaning of each signs. In this research, we develop an application which can classify a road sign image into three classes, priority four-way crossroad, do-not-park sign, and follow-this-road sign. Initially, the system will do preprocessing step such as grays calling, histogram equalization, and input image segmentation. Next, the feature extraction step will be conducted, namely the spatial moment feature extraction, normalized centering, and color statistics. Lastly, the feature representation from both extraction methods will be used to classify the image using K-nearest neighbor. Experiment result shows that the combination of both feature extraction methods gives promising result. From 21 training images and 15 testing images, the system can recognize the traffic signs with 100% accuracy with K=3, 86.6% with K=5, and 86.6% with K=7. Rambu lalu lintas merupakan salah satu alat perlengkapan jalan dalam bentuk tertentu yang memuat lambang, huruf, angka, kalimat yang digunakan untuk memberikan perintah, larangan, peringatan dan petunjuk bagi pengguna jalan agar tertib berlalu lintas. Namun, banyak di antara pengguna jalan yang belum mengetahui arti dari setiap rambu lalu lintas yang terpasang.Pada penelitian ini, dibuatlah suatu aplikasi yang mampu melakukan klasifikasi citra rambu ke dalam 3 kelas yaitu: peringatan simpang empat prioritas, larangan parkir dan perintah memasuki jalur atau lajur yang ditunjuk. Mula-mula sistem akan melakukan prapemrosesan seperti seperti: grayscalling, histogram equalization, dan segmentasi pada citra input. Selanjutnya, tahap ekstraksi ciri akan dilakukan pada citra hasil pra-pemrosesan. Adapun metode ekstraksi ciri yang digunakan pada penelitian kali ini adalah ekstraksi fitur momen spasial dan pusat ternormalisai (momen) dan ekstraksi fitur statistika warna (warna). Terakhir, nilai fitur yang dihasilkan oleh kedua metode tersebut akan diklasifikasi mengguakan K-Nearest Neighbor. Hasil uji coba menunjukkan bahwa metode ekstraksi fitur gabungan momen-warna memberikan hasil yang menjanjikan. Dari 21 citra latih dan 15 citra uji yang digunakan, sistem mampu mengenali rambu dengan tepat 100% pada K=3 , 86,6% pada K=5, dan 86,6% pada K=7. 
Analisis Komparasi Algoritma K-Means Dan K-Medoids Dalam Segmentasi Data Untuk Strategi Promosi Mahasiswa Baru Di Universitas X Syafrudin, Teguh; Teguh Syafrudin; Arief Hermawan; Donny Avianto; Indra Maulana
Komputika : Jurnal Sistem Komputer Vol. 14 No. 2 (2025): Komputika: Jurnal Sistem Komputer
Publisher : Computer Engineering Departement, Universitas Komputer Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.34010/komputika.v14i2.16698

Abstract

Persaingan dalam merekrut mahasiswa baru semakin ketat, sehingga perguruan tinggi memerlukan strategi promosi yang efektif dan tepat sasaran. Salah satu cara untuk meningkatkan efektivitas promosi adalah dengan melakukan segmentasi calon mahasiswa berdasarkan data penerimaan. Penelitian ini menawarkan solusi dengan membandingkan performa algoritma K-Means dan K-Medoids dalam segmentasi data Penerimaan Mahasiswa Baru (PMB) Universitas X tahun 2024. Metode yang digunakan meliputi tahapan pengumpulan data, preprocessing (pembersihan, normalisasi, dan transformasi), implementasi algoritma K-Means dan K-Medoids, serta evaluasi kualitas klaster menggunakan Davies-Bouldin Index (DBI). Hasil penelitian menunjukkan bahwa konfigurasi tiga klaster (K=3) memberikan nilai DBI terendah, dengan K-Medoids mencapai 1,038, lebih baik dibandingkan K-Means sebesar 1,059. Klaster dominan menunjukkan bahwa lulusan SMK mendominasi sebesar 40,45% dan cenderung memilih program studi Pendidikan TIK. Kontribusi penelitian ini adalah memberikan panduan bagi institusi pendidikan dalam memilih algoritma klasterisasi yang paling sesuai untuk mendukung strategi promosi yang lebih akurat, efisien, dan terarah.
ANALISIS KINERJA SUPPORT VECTOR MACHINE DAN NAÏVE BAYES CLASSIFIER DALAM KLASIFIKASI SENTIMEN DAN EMOSI PADA UMPAN BALIK MAHASISWA TERHADAP KINERJA DOSEN Syafrudin, Teguh; Hermawan, Arief; Avianto, Donny
Jurnal Digit : Digital of Information Technology Vol 15, No 2 (2025)
Publisher : Universitas Catur Insan Cendekia (CIC) Cirebon

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.51920/jd.v15i2.442

Abstract

Evaluasi kinerja dosen melalui umpan balik mahasiswa merupakan komponen penting dalam meningkatkan mutu pembelajaran di perguruan tinggi. Penelitian ini bertujuan untuk mengembangkan model klasifikasi sentimen dan emosi menggunakan pendekatan hybrid machine learning dengan mengombinasikan algoritma Naïve Bayes dan Support Vector Machine (SVM). Dataset berasal dari umpan balik mahasiswa yang telah dianotasi secara manual ke dalam tiga kategori sentimen positif, netral, negatif dan delapan kategori emosi. Proses preprocessing dilakukan melalui tokenisasi, stemming, dan transformasi data ke dalam bentuk TF-IDF. Hasil klasifikasi menunjukkan bahwa SVM memiliki performa terbaik untuk klasifikasi sentimen dengan akurasi mencapai 90%, mengungguli Naïve Bayes yang hanya memperoleh akurasi 80%. Sebaliknya, performa klasifikasi emosi jauh lebih rendah, dengan akurasi maksimum 35% pada model SVM dan 20% pada Naïve Bayes. Beberapa emosi seperti “marah”, “termotivasi”, dan “senang” tidak dapat dikenali oleh model karena ketidakseimbangan distribusi data dan konteks emosi yang sulit ditangkap dari teks pendek. Temuan ini menunjukkan bahwa pendekatan hybrid efektif untuk klasifikasi sentimen dalam kondisi data terbatas, namun klasifikasi emosi memerlukan pendekatan lanjutan seperti reduksi label atau penggunaan model berbasis konteks untuk mencapai hasil yang lebih baik.Kata kunci: Kinerja Dosen, Naïve Bayes Classifier (NBC), Support Vector Machine (SVM), Umpan Balik Mahasiswa.
Co-Authors Adicahya, Bina Sukma Adityo Permana Wibowo Alwani, Adie G. Amalia Rizki Wulandari Andri Yudha Pratama Apriansyah, Ferryma Arba Ardiansyah, Diky Aribowo Aribowo Arief Hermawan Arieska Restu Harpian Dwika Ashari, Nadia Asrul Gunawan Aziz Perdana Baiq Nurul Azmi Bayu Tri Nugroho, Bayu Tri Bimantoro, Nazar Iqbal Bowo Hirwono Budiyanto, Irfan Cahaya Muzaddidah Dewi, Amelia Citra Dian Wijayanti Dimas Dwi Kurniawan Dimas Rizqi Kurniawan Dwi Ratnawati, Dwi Edi Priyanto Enggar Novianto Enggar Novianto Erfin Nur Rohma Khakim Fadhila, Arifa Farras Fadilah, Faiz Fahri Putra Herlambang Fakharudin, Panji Rangga Adzan Fajar Faqih, Allan Bil Febiansyah Annaufal Ahnaf Fauzi Ferdinandus Edwin Penalun Gumilang, Muhammad Satrio Gunawan, Asrul Hanif, Rifqi Fadhlurrahman Hardiyantari, Oktavia Ida Kumala Sari Iin Rohmatika Aulia Ilmy Eka Handayani Imantoko Imantoko Indra Maulana Iqbal, Muhammad Izza Irfan Budiyanto Jagad Raya Ramadhan Khalifatur Rauf Kusumastuti, Asriana Dyah Laode Izat Trianto Haradin Lidya Nurmala Eva Maulana, Adha Muh Arifandi Muhammad Irsyad Indra Fata Muhammad Kusban Muhammad Rizki Muhammad Rizki Nasmah Nur Amiroh Novaldy, Olwin Kirab Nur Widiastuti Nurazila, Siti Octavianus, Yonathan Perdana, Aziz Purba, Yurjaa Ghoniyyan Putra, Kristianto Pratama Dessan Rahma Nur Azizah Reski Noviana Rian Oktafiani Rian Oktafiani Rianto Rianto Risnanto, Ari Rizarta, Rusma Eko Fiddy Rizki Purnomo Pratama Rizky Samudra Falasyfa Roselilie Simbulan Roy Fasti Rubangi Rubangi Rudi, Rudiono Rusma Eko Fiddy Rizarta Saputra, Candra Heru Satriya Adhitama Setiawan, Muhhamad Ajun Siti Rokhanah Soraya Fatmawati Sri Wulandari SRI WULANDARI Sutarman Sutarman Syafrudin, Teguh Syahab, Alfin Syarifuddin Teguh Syafrudin Tri Untoro, Iwan Hartadi Tri Widodo Vivianti Wahid, Ach. Nur Aqil Wayan Praka