cover
Article Per Year (5 Year)
All
Home Page
OAI Link
Editorial Team
Contact
Reviewer
Google Scholar
Contact Name
FIRMAN TEMPOLA
Contact Email
firma.tempola@unkhair.ac.id
Phone
-
Journal Mail Official
if_jiko@unkhair.ac.id
Editorial Address
-
Location
Kota ternate,
Maluku utara
INDONESIA
Jiko (Jurnal Informatika dan komputer)
Published by Universitas Khairun
ISSN : 26148897     EISSN : 26561948     DOI : -
Core Subject : Science,
Computer Science & IT
Jiko (Jurnal Informatika dan Komputer) Ternate adalah jurnal ilmiah diterbitkan oleh Program Studi Teknik Informatika Universitas Khairun sebagai wadah untuk publikasi atau menyebarluaskan hasil - hasil penelitian dan kajian analisis yang berkaitan dengan bidang Informatika, Ilmu Komputer, Teknologi Informasi, Sistem Informasi dan Sistem Komputer. Jurnal Informatika dan Komputer (JIKO) Ternate terbit 2 (dua) kali dalam setahun pada bulan April dan Oktober
Arjuna Subject : -
Articles 312 Documents
 28   29   30   31   32 
COMPARISON OF LINEARSVC AND COMPLEMENT NAIVE BAYES ON CORETAX SENTIMENT ANALYSIS USING INDOBERT PSEUDO-LABELING AND ADASYN Riza Febyana Shollis; Ucta Pradema Sanjaya
JIKO (Jurnal Informatika dan Komputer) Vol 9 No 2 (2026)
Publisher : Program Studi Teknik Informatika Universitas Khairun

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33387/jiko.v9i2.11615

Abstract

Coretax is an Indonesian digital tax service platform that has attracted extensive public discussion on social media, particularly on X/Twitter. Most discussions are related to system stability and user experience when accessing the platform. This study collected sentiment data from December 1, 2024, to December 20, 2024, resulting in 10,140 tweets, which were filtered into 7,900 valid data points. The dataset underwent standard text preprocessing and was represented using TF-IDF features. Sentiment labeling was performed automatically (pseudo-labeling) using the IndoBERT model, producing an imbalanced class distribution (Negative 63.20%; Neutral 30.24%; Positive 6.56%). To address this imbalance, ADASYN was applied to the training data, and two classification models were compared: Complement Naïve Bayes and SVM (LinearSVC). Evaluation using an 80:20 train–test split showed that LinearSVC + ADASYN achieved the best performance with an accuracy of 85.25%, F1-Macro of 0.7423, MCC of 0.7120, ROC-AUC of 0.9428, and Hamming Loss of 0.1475, outperforming ComplementNB + ADASYN with an accuracy of 78.67%. Furthermore, the McNemar test confirmed that the performance difference between the two models is statistically significant. These findings indicate that LinearSVC is more effective in distinguishing sentiments related to technical complaints and procedural inquiries in the context of digital tax services.
SYSTEMATIC REVIEW: COMPARISON OF ARTIFICIAL INTELLIGENCE METHODS FOR PAIN DETECTION Fiona Angeline; Sherli Sherli; Devin Tanadi; Rendi Winata; Christnatalis HS
JIKO (Jurnal Informatika dan Komputer) Vol 9 No 2 (2026)
Publisher : Program Studi Teknik Informatika Universitas Khairun

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33387/jiko.v9i2.11648

Abstract

Objective pain detection remains a major challenge in healthcare because conventional assessments rely on subjective self-reports. Advances in artificial intelligence (AI) have enabled more objective approaches by analyzing biological signals and human expressions. Facial expressions and functional Near-Infrared Spectroscopy (fNIRS) are widely studied due to their complementary characteristics. Facial expressions are non-invasive, easy to capture, and strongly associated with visible pain responses, making them suitable for real-time applications. In contrast, fNIRS measures brain activity related to pain perception, providing a more objective physiological perspective. This study follows PRISMA guidelines for systematic literature review. Studies were identified from major databases, screened for relevance, assessed for eligibility, and included for final analysis. A total of 45 studies were selected. Previous research shows that AI, especially deep learning, is effective in analyzing facial expressions and fNIRS signals for pain detection. However, few studies systematically compare these modalities in a unified framework. This review highlights a shift from single-modality to hybrid and multimodal approaches integrating facial and fNIRS data. Deep learning models, particularly CNNs for facial analysis and hybrid machine learning–deep learning methods for physiological signals, dominate recent studies. Multimodal fusion consistently outperforms single-modality approaches, improving accuracy and robustness in pain detection tasks
COMPARATIVE ANALYSIS OF SUPPORT VECTOR MACHINE AND RANDOM FOREST METHODS BASED ON RANDOMIZED SEARCH OPTIMIZATION IN HOAX NEWS CLASSIFICATION Vincent Lawrence; Frans Mikael Sinaga
JIKO (Jurnal Informatika dan Komputer) Vol 9 No 2 (2026)
Publisher : Program Studi Teknik Informatika Universitas Khairun

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33387/jiko.v9i2.11650

Abstract

In today’s digital era, the spread of hoax news has increasingly escalated alongside the ease of access to information through social media and online news portals. This phenomenon has caused negative impacts such as public confusion, social conflict, and a decline in public trust toward information accuracy. Therefore, an effective classification method is needed to accurately detect hoax news. This study aims to analyze and compare the performance of the Support Vector Machine (SVM) and Random Forest (RF) algorithms in classifying hoax news, by applying the Randomized Search optimization technique to enhance model performance. The dataset used in this research was obtained from Kaggle, titled Indonesia Fact and Hoax Political News, consisting of news titles and narratives as input attributes, and hoax or factual labels as outputs. The results show that the SVM algorithm without optimization achieved an accuracy of 83.92%, which increased to 84.28% after optimization using Randomized Search. Meanwhile, the Random Forest algorithm without optimization achieved an accuracy of 85.93%, which increased to 86.05% after optimization. Based on these findings, it can be concluded that the application of Randomized Search successfully improved the accuracy, sensitivity, and stability of the classification models, with the Random Forest algorithm providing the best performance in detecting hoax news in this study.
AN INTEGRATED AHP-MFEP DECISION SUPPORT MODEL FOR IMPROVING PRODUCT RECOMMENDATION ACCURACY IN FASHION RETAIL Cici Anggriani; Riki Andri Yusda; Maulana Dwi Sena
JIKO (Jurnal Informatika dan Komputer) Vol 9 No 2 (2026)
Publisher : Program Studi Teknik Informatika Universitas Khairun

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33387/jiko.v9i2.11704

Abstract

The development of decision support systems has become essential in assisting businesses to make accurate and objective decisions, especially in product selection processes. However, many retail businesses still rely on manual decision-making, which is often subjective and inconsistent. This study aims to develop a decision support system using the Analytical Hierarchy Process (AHP) and Multi-Factor Evaluation Process (MFEP) methods to improve the accuracy and consistency of product recommendations. The AHP method is used to determine the weight of each criterion through pairwise comparisons, while the MFEP method is applied to evaluate and rank product alternatives based on these weights. The criteria used in this study include sales, profit, and customer preferences. Data collection was conducted through observation, interviews, and documentation. The results show that the system is able to generate structured and objective product rankings. The highest score obtained is 4.6, indicating the most recommended product alternative. Furthermore, the evaluation results show that the system achieves an accuracy of 92% and a precision value of 90%, indicating high relevance of the recommendations. In addition, the Spearman Rank Correlation value of 0.95 indicates a very strong agreement between the system results and expert judgment. Therefore, the proposed system is effective in improving decision-making accuracy, reducing subjectivity, and providing reliable product recommendations.
ANALYSIS OF SEISMICITY ANOMALIES IN SULAWESI USING DBSCAN BASED ON USGS DATA (2021–2026) Muhammad Dzaky Hasyim; Muhammad Wahyu Hikmalsyah; Gabriel Sebastian Santoso; Ken Ditha Tania; Allsela Meiriza; Ahmad Rifai
JIKO (Jurnal Informatika dan Komputer) Vol 9 No 2 (2026)
Publisher : Program Studi Teknik Informatika Universitas Khairun

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33387/jiko.v9i2.11728

Abstract

Sulawesi is one of Indonesia's most tectonically complex regions, situated at the meeting point of major plates and active fault systems, which results in significant seismic activity. This study aims to analyze seismicity anomalies in Sulawesi from 2021 to 2026 using the Density-Based Spatial Clustering of Applications with Noise (DBSCAN) algorithm applied to United States Geological Survey (USGS) data. The methodology involves a hybrid approach for parameter optimization, utilizing both the K-Distance Graph (KneeLocator) and Grid Search to ensure results are geologically representative. From an initial dataset of 859 events, 839 earthquake records were processed after geographic filtering, showing an average magnitude of 4.80 and an average depth of 70.15 km. The analysis reveals that the optimal parameters (ε = 0.2897 and MinPts = 3) produced 25 distinct clusters that align with real geological structures such as the Palu-Koro Fault and the Molucca Sea subduction zone. The algorithm successfully identified 66 points (7.87%) as seismic anomalies or noise, representing sporadic tectonic activity outside primary density zones. Furthermore, Convex Hull visualization identified a significant "seismic gap" between the Palu-Koro segment and Southeast Sulawesi, indicating a high-risk zone for potential future energy release. These findings demonstrate that density-based clustering is highly effective for mapping seismic hazards in complex tectonic regions, providing vital data for sustainable disaster mitigation planning in Sulawesi.
A DEEP LEARNING-BASED SMART MOBILE APPLICATION FOR AUTOMATED CITRUS FRUIT QUALITY CLASSIFICATION Armando Sitorus; Laskar Eltriman Gulo; Titien The Lawren Pasaribu; Josi Leonardo Davinsi Saragih; Adya Zizwan Putra
JIKO (Jurnal Informatika dan Komputer) Vol 9 No 2 (2026)
Publisher : Program Studi Teknik Informatika Universitas Khairun

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33387/jiko.v9i2.11750

Abstract

This study aims to develop a digital image-based citrus fruit quality detection system using the Convolutional Neural Network (CNN) method with the MobilenetV2 architecture and implement it into a cross-platform mobile application. The dataset used is a combination of public datasets and local citrus fruit datasets with five quality classes, namely fresh, half-ripe, black spots, damaged, and rotten. The total data is 5001 data, with 80% as training data and 20% as testing data. The CNN model was trained for 10 epochs and evaluated using accuracy, precision, recall, and F1 score metrics. The test results show that the CNN model is able to classify citrus fruit quality with high and consistent performance, indicated by precision, recall, and F1 score values in the range of 0.96-0.97. The trained model is integrated into a Flutter-based mobile application through the Flask backend, enabling real-time citrus fruit quality detection through a smartphone camera. The results of the study prove that the integration of CNN and cross-platform mobile applications can be an effective and objective solution in automatically detecting citrus fruit quality.
PERFORMANCE ANALYSIS OF ROUTING PROTOCOLS IN LARGE-SCALE COMPUTER NETWORKS USING SIMULATION METHOD Hafidzun Alim; Arafat Febriandirza
JIKO (Jurnal Informatika dan Komputer) Vol 9 No 2 (2026)
Publisher : Program Studi Teknik Informatika Universitas Khairun

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33387/jiko.v9i2.11853

Abstract

Routing protocols are essential components in computer networks that function to determine the best path in the data transmission process. In large scale networks the selection of the appropriate routing protocol highly influences the Quality of Service or QoS and network stability. This research aims to analyze and compare the performance of OSPF, EIGRP and BGP routing protocols based on QoS parameters which include throughput, delay, jitter and packet loss also convergence time and system resource utilization. The research method used is network simulation using Network Simulator 3 (NS-3) with a quantitative experimental approach. Testing is conducted on a large-scale network with variations in the number of nodes, data traffic patterns, and network topologies. The simulation result data is analyzed to determine the performance differences of each routing protocol. The research results indicate that OSPF has the most optimal performance in large-scale internal networks with high throughput, low latency, and fast convergence time. EIGRP demonstrates stable and balanced performance, while BGP excels in the scalability aspect but has higher convergence time and latency. The conclusion of this research asserts that the selection of routing protocols must be adjusted to the needs and characteristics of the applied network. 
A RECALL-ORIENTED STACKING ENSEMBLE FOR EXTREME RAINFALL EARLY WARNING IN TERNATE Achmad Fuad; Muhammad Sabri Ahmad; Muhammad Ridha Albaar; Yasir Muin
JIKO (Jurnal Informatika dan Komputer) Vol 9 No 2 (2026)
Publisher : Program Studi Teknik Informatika Universitas Khairun

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33387/jiko.v9i2.11857

Abstract

Extreme-rainfall detection in small tropical islands is difficult because extreme events are rare and false negatives have substantial disaster-mitigation consequences. This study aims to (1) compare imbalance-aware machine-learning and deep temporal models, (2) determine whether an out-of-fold stacking ensemble can maximize sensitivity, and (3) explain model behavior using SHAP. ERA5 hourly data for a representative Ternate grid point were collected from 1 January 2005 to 6 February 2026 and aggregated into 7,707 daily records; only 12 days met the extreme threshold of 75 mm/day. A 30-day sequence was used to predict the next-day class, and three expanding temporal test folds were evaluated. Weighted loss, focal loss, balanced tree learning, validation-based threshold selection, and leakage-controlled stacking were applied. Across 2,229 pooled test days containing four extreme events, the stacking model detected all events (recall 1.000) but produced low precision (0.0023), F2-score 0.0115, specificity 0.2256, and 1,723 false positives. Balanced Random Forest provided the best overall trade-off, with recall 0.750, F2-score 0.0316, specificity 0.7951, and MCC 0.0570. Therefore, stacking is useful as a high-sensitivity screening layer, but it is not yet suitable as a standalone operational warning model. Local observations and additional event samples are required for calibration.
ANALYSIS ON MACHINE LEARNING MODELS ROBUSTNESS AGAINST NOISE AND CONCEPT DRIFT IN THE CICIDS2017 DATASET Azhar Bintang Pramudyanto; Chyntia Raras Ajeng Widiawati; M. Syaiful Amin
JIKO (Jurnal Informatika dan Komputer) Vol 9 No 2 (2026)
Publisher : Program Studi Teknik Informatika Universitas Khairun

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33387/jiko.v9i2.11875

Abstract

Intrusion Detection Systems (IDS) based on machine learning face significant challenges in real-world deployment due to noise in network data and concept drift caused by evolving attack patterns. This study analyzes the robustness of three machine learning models Logistic Regression, Decision Tree, and Random Forest against noise and concept drift using the CICIDS2017 dataset. The experimental design includes baseline testing, noise robustness testing with three intensity levels (5%, 10%, 20%), and concept drift evaluation through temporal data splitting. Performance is measured using accuracy, precision, recall, and F1-score metrics, with robustness score calculated as a weighted average (40% baseline, 30% noise, 30% drift). Results show Logistic Regression achieves the highest robustness score (60.07%) due to excellent noise tolerance (1.09% F1-score degradation), followed by Random Forest (59.65%) and Decision Tree (57.60%). However, all models are categorized as NOT ROBUST against concept drift with degradation ranging from 29.64% to 30.34%, indicating sudden drift between Thursday and Friday data. The findings reveal that noise robustness and concept drift robustness are independent characteristics that do not correlate. Logistic Regression is recommended for practical deployment due to its optimal combination of robustness, interpretability, and computational efficiency. This research contributes to understanding model stability under non-ideal data conditions and emphasizes the necessity of implementing adaptive mechanisms such as periodic retraining and drift detection in operational IDS architecture.
PUBLIC SENTIMENT ANALYSIS TOWARDS THE DISCOURSE OF MILITARY SERVICE FOR PROBLEMATIC CHILDREN USING INDOBERT ON SOCIAL MEDIA X Windu Abdillah; Estu Sinduningrum
JIKO (Jurnal Informatika dan Komputer) Vol 9 No 2 (2026)
Publisher : Program Studi Teknik Informatika Universitas Khairun

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33387/jiko.v9i2.11890

Abstract

This research aims to analyze public sentiment towards the discourse of military service for problematic children using data from social media X. The approach used is deep learning based on Transformer with the IndoBERT model. The dataset was obtained through a scraping process and produced 421 data after the preprocessing and labeling stages. The research stages include text preprocessing, tokenization, sentiment labeling, model training, and performance evaluation using accuracy, precision, recall, and F1-score metrics. The research results show that the IndoBERT model is capable of achieving an accuracy of 68.23%. The sentiment distribution is dominated by negative sentiment (55.34%) which reflects public resistance to the policy. The contribution of this research lies in the integration of non-standard language preprocessing with IndoBERT as well as the analysis of public opinion on specific social issues 

Page 31 of 32 | Total Record : 312

 28   29   30   31   32 

Filter by Year

2017 2026


Reset
Filter By Issues
All Issue Vol 9 No 2 (2026) Vol 9 No 1 (2026) Vol 8 No 3 (2025) Vol 8 No 2 (2025) Vol 8, No 1 (2025) Vol 8 No 1 (2025) Vol 7, No 3 (2024) Vol 7 No 3 (2024) Vol 7, No 2 (2024) Vol 7 No 2 (2024) Vol 7, No 1 (2024) Vol 7 No 1 (2024) Vol 6, No 2 (2023): JIKO (Jurnal Informatika dan Komputer) Vol 6, No 1 (2023) Vol 5, No 3 (2022) Vol 5, No 2 (2022) Vol 5, No 1 (2022) Vol 4, No 3 (2021) Vol 4, No 2 (2021) Vol 4, No 1 (2021) Vol 3, No 3 (2020) Vol 3, No 2 (2020) Vol 3, No 1 (2020) Vol 2, No 2 (2019) Vol 2, No 1 (2019) Vol 1, No 2 (2018) Vol 1, No 1 (2018) Vol 1, No 2 (2017): Jurnal Informatika dan Komputer (JIKO) Ternate Vol 1, No 1 (2017): Jurnal Informatika dan Komputer (JIKO) Ternate More Issue