Article Per Year (5 Year)
p-Index From 2021 - 2026
1.116
P-Index
This Author published in this journals
All Journal JURNAL INFORMATIKA Progresif: Jurnal Ilmiah Komputer SKANIKA: Sistem Komputer dan Teknik Informatika SAINTEK Jurnal Riset dan Aplikasi Mahasiswa Informatika (JRAMI)
Karina Imelda
Universitas Pelita Bangsa
Automotive Engineering Civil Engineering, Building, Construction & Architecture Computer Science & IT Control & Systems Engineering Electrical & Electronics Engineering Engineering Industrial & Manufacturing Engineering Other

Published : 6 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 6 Documents
Search

 1 
Optimalisasi Pengolahan Air Limbah Dengan Menggunakan Metode Ozon Microbubble Untuk Menurunkan Kadar COD (Chemical Oxygen Demand) Dan TSS (Total Suspended Solid) Di PT Industri Kimia Aji Susanto; Karina Imelda; Dodit Ardiatma; Nur Ilman Ilyas
Prosiding Sains dan Teknologi Vol. 1 No. 1 (2022): Seminar Nasional Sains dan Teknologi (SAINTEK) ke 1 - Juli 2022
Publisher : DPPM Universitas Pelita Bangsa

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

Abstract

This study aims to determine the effect of variations in contact time on decreasing levels of Chemical Oxygen Demand (COD) and Total Suspended Solid (TSS) using Ozone Microbubble and determine the efficiency level of Ozone Microbubble. This research uses liquid waste from WWTP at PT. Chemical Industry Cikarang with variations in processing time of 20, 40, 60 minutes. The test results before the Ozone Microbubble method was applied did not meet the quality standards and after being applied with a time variation of 20, 40, 60 minutes, they met the Hyundai Industrial Estate wastewater quality standards. The decrease for COD levels with a time of 60 minutes was 614.6 mg/L. Meanwhile, the TSS level is 170.8 mg/L. Both still meet the quality standards of the Hyundai Industrial Estate. The efficiency level for COD is 33% and TSS is 32%. From the results of reducing COD and TSS, it can be concluded that the use of Ozone Microbubble can be used in chemical waste. In addition to the relatively high percentage of reduction, in terms of costs in this research, it is also relatively cheap and can be a proposed idea if it is to be implemented. Keywords: Ozon, Microbubble, Chemical Industry Liquid Waste, COD, TSS
Model Rekomendasi Konten Edukasi Diabetes pada Instagram dengan Integrasi TF-IDF dan Cosine Similarity berbasis Natural Language Processing Anggita Risqi Nur Clarita; Muhamad Fatchan; Karina Imelda
JURNAL INFORMATIKA Vol 15, No 1 (2026): Jurnal Informatika
Publisher : Informatics Engineering Department, Dayanu Ikhsanuddin University

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.55340/jiu.v15i1.2719

Abstract

Meningkatnya prevalensi Diabetes Melitus menuntut penguatan literasi kesehatan digital mandiri. Namun, mayoritas edukasi di Instagram bersifat generalis dan belum tersegmentasi sesuai karakteristik medis pasien. Penelitian ini bertujuan mengembangkan model rekomendasi konten edukasi diabetes personal menggunakan pendekatan Content-Based Filtering. Berbeda dari model pencocokan kata konvensional terdahulu yang gagal menangani tingginya tumpang tindih istilah medis pada teks non-formal, kebaruan penelitian ini terletak pada integrasi knowledge-base terminologi medis hierarkis (Tipe 1, Tipe 2, Umum) yang diselaraskan bersama praktisi kesehatan untuk memandu akurasi pembobotan. Metode yang digunakan meliputi Term Frequency-Inverse Document Frequency (TF-IDF) untuk representasi fitur dan Cosine Similarity untuk mengukur kemiripan leksikal antar-vektor teks. Hasil evaluasi terhadap 75 data uji menunjukkan capaian akurasi klasifikasi back-end sebesar 84% (Interval Kepercayaan 95%: [75,70%, 92,30%]), dengan kualitas urutan pemeringkatan rekomendasi terarah pada nilai rata-rata Precision at 3 (P@3) sebesar 88%. Analisis data empiris mengonfirmasi kesenjangan di lapangan di mana 72,45% konten didominasi materi Umum. Kesimpulannya, model ini menunjukkan potensi performa baik dalam meminimalkan bias prediksi pada dataset terkait, meski masih memiliki keterbatasan overlap kata kunci pada kategori Tipe 2 (Recall 0,76). Pengembangan ke depan memerlukan arsitektur multimodal untuk memproses informasi dari media visual.
Analisis Opini Konsumen terhadap Testimoni Produk Pelapak pada Online Marketplace dengan Pendekatan Sentimen Analisis Handala Simetris Harahap; Suprapto; Asep Suprianto; Karina Imelda
Prosiding Sains dan Teknologi Vol. 4 No. 1 (2025): Seminar Nasional Sains dan Teknologi (SAINTEK) ke 4 - Februari 2025
Publisher : DPPM Universitas Pelita Bangsa

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

Abstract

Testimoni dan produk review, dimana setiap calon pembeli dapat melihat testimoni atau review yang diberikanoleh pembeli lain terhadap produk yang telah dibeli sehingga calon pembeli baru akan mengetahui baikburuk nya produk tersebut sebelum memutuskan untuk membeli atau tidak. Selain itu testimoni userberguna untuk menilai reputasi sebuah online marketplace, dengan online marketplace yang begitu beragamdiperlukan sebuah model untuk membantu pelanggan dalam memilih toko dan online marketplace yang tepat.Pada tulisan ini penulis menggukan sentiment analisis dalam menganalisa product review dari tiap pelanggan.Sehingga mengetahui kelebihan dan kekurangan dari tiap online marketplace yang sedang diteliti.
Komparasi Algoritma Klasifikasi Machine Learning Dengan Penerapan Metode Ensemble Stacking untuk Menganalisa Sentimen terhadap Kesehatan Mental Annisa Maulana Majid; Karina Imelda; Ismasari Nawangsih
SKANIKA: Sistem Komputer dan Teknik Informatika Vol 8 No 2 (2025): Jurnal SKANIKA Juli 2025
Publisher : Universitas Budi Luhur

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.36080/skanika.v8i2.3561

Abstract

Mental health often goes undetected due to the absence of physical symptoms, which hinders timely and appropriate intervention. Many individuals choose to express their emotions on social media rather than access professional services. However, the use of social media can potentially worsen mental health conditions and even impact physical well-being. Therefore, early detection through the analysis of digital data, particularly social media posts, using machine learning approaches is essential. Previous research on mental health sentiment analysis has utilized classification algorithms, but accuracy improvement remains necessary. This study compares single classification algorithms and applies an ensemble stacking method that combines multiple classifiers as base learners and a meta-learner. The results show that the stacking method achieves a higher accuracy of 88.13%.
Kesehatan mental telah menjadi isu penting yang banyak dibahas melalui media sosial sehingga diperlukan metode otomatis untuk mengidentifikasi kategori kondisi kesehatan mental berdasarkan teks. Penelitian ini bertujuan membangun model klasifikasi teks me Annisa Maulana Majid; Ismasari Nawangsih; Karina Imelda
Progresif: Jurnal Ilmiah Komputer Vol. 22 No. 3 (2026): Juli
Publisher : STMIK Banjarbaru

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35889/progresif.v22i3.4070

Abstract

Mental health had become an important issue widely discussed on social media, creating the need for an automated method to identify mental health conditions based on textual data. This study aimed to develop a text classification model using Bidirectional Encoder Representations from Transformers and improve the transparency of prediction results through Shapley Additive Explanations and Local Interpretable Model-agnostic Explanations. The dataset consisted of 51,073 text records categorized into seven mental health classes. The research stages included data and text cleaning, label encoding, data splitting, tokenization, model training, evaluation, and result interpretation. The testing results showed that the model achieved an accuracy of 82% and a weighted average F1-score of 0.82. The interpretation results indicated that specific words and phrases contributed to class predictions. The findings demonstrated that the model performed text classification effectively, while both interpretation methods improved the transparency of the model’s decision-making process. Keywords: Sentiment Analysis; BERT; XAI; SHAP; LIME   Abstrak Kesehatan mental telah menjadi isu penting yang banyak dibahas melalui media sosial sehingga diperlukan metode otomatis untuk mengidentifikasi kategori kondisi kesehatan mental berdasarkan teks. Penelitian ini bertujuan membangun model klasifikasi teks menggunakan Bidirectional Encoder Representations from Transformers (BERT) serta meningkatkan transparansi hasil prediksi melalui Shapley Additive Explanations (SHAP) dan Local Interpretable Model-agnostic Explanations (LIME). Dataset yang digunakan terdiri atas 51.073 teks dalam tujuh kategori kesehatan mental. Tahapan penelitian meliputi pembersihan data dan teks, pengodean label, pembagian data, tokenisasi, pelatihan model, evaluasi, serta interpretasi hasil. Hasil pengujian menunjukkan bahwa model memperoleh akurasi sebesar 82% dan nilai F1-score rata-rata tertimbang sebesar 0,82. Interpretasi menunjukkan bahwa frasa dan kata tertentu memberikan kontribusi terhadap prediksi kelas. Hasil penelitian membuktikan bahwa model mampu melakukan klasifikasi dengan kinerja yang baik, sedangkan kedua metode interpretasi meningkatkan transparansi keputusan model.
Penerapan Algortima Random Forest Classifier untuk Rekomendasi Produk Skincare Berdasarkan Kondisi Kulit Pengguna Zahra Nurhaliza; Abdul Halim Anshor; Karina Imelda
Jurnal Riset dan Aplikasi Mahasiswa Informatika (JRAMI) Vol. 7 No. 03 (2026): Jurnal Riset dan Aplikasi Mahasiswa Informatika (JRAMI)
Publisher : Program Studi Teknik Informatika, FTIK, Universitas Indraprasta PGRI

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30998/jrami.v7i03.1383

Abstract

The selection of skincare products that do not match skin characteristics may cause problems such as irritation, acne, and dry skin. However, many users still experience difficulties in identifying their skin condition, resulting in less accurate product selection. This study aims to develop a skin-type classification and skincare product recommendation system based on the Random forest classifier algorithm by utilizing textual product data. The study employed a quantitative approach with an experimental method using secondary data from Kaggle, with ingredients and afterUse as the main attributes. The research stages included preprocessing, rule-based label construction, feature weighting using Term Frequency–Inverse Document Frequency (TF-IDF), an 80:20 data split, model training, evaluation, and implementation of a web-based system using Streamlit. The evaluation results showed that the model achieved an accuracy of 0.8229, weighted precision of 0.8215, weighted recall of 0.8229, and weighted F1-score of 0.8130. Therefore, the Random forest classifier is capable of supporting text-based skin-type classification with good performance and producing a web-based skincare product recommendation system.
Co-Authors Abdul Halim Anshor Aji Susanto Anggita Risqi Nur Clarita Annisa Maulana Majid Asep Suprianto Dodit Ardiatma Handala Simetris Harahap Muhamad Fatchan Nawangsih, Ismasari Nur Ilman Ilyas Suprapto Zahra Nurhaliza