Jurnal Teknoinfo

CNN EfficientNetB0 Approach Based on Transfer Learning for Deepfake Detection and Disinformation Mitigation

Muhammad Erico Revaldo — 2026-07-24

Rapid advances in deep learning technology have led to the emergence of artificial intelligence (AI) media that is very similar to reality, called deepfakes, which have the potential to pose a serious threat to information integrity and public trust. Although detection methods using Convolutional Neural Networks (CNN) have been developed, most still struggle with generalization, particularly in distinguishing modern deepfakes from non-standard original images such as selfies, which often leads to high false positive rates. This study introduces a robust detection model based on the EfficientNetB0 architecture implemented through transfer learning techniques. To improve generalization capabilities and minimize bias, we compiled a large and balanced combined dataset by combining three different public datasets (including classic deepfakes, face swaps, and many authentic selfies). The model was trained using a two-stage strategy: first for feature extraction, then refinement with a very low learning rate. The model's performance was thoroughly evaluated on stratified test data using five key metrics. The results of the experiment showed outstanding performance, achieving 99.81% accuracy and a Macro F1 score of 99.81%. Additionally, the reliability metrics ROC-AUC, Average Precision (AP), and True Positive Rate (TPR) all reached 99.99%, while the False Positive Rate (FPR) remained strictly at 1%. As proof of concept, this optimized model was implemented in a web prototype built using the Django framework, allowing users to upload images and receive classification results in real-time.

Designing a Web-Based Enterprise Resource Planning System Using the CodeIgniter Framework for Business Process Optimization

Joko Riyanto — 2026-07-24

Enterprise Enterprise Resource Planning (ERP) is an integrated system designed to improve efficiency and accuracy in company information management. This study aims to design and implement a web-based ERP system at CV Triraksa Jaya Mandiri as a solution to various operational issues, such as manual data processing, reporting delays, and a lack of integration between divisions. The system development method applied is Extreme Programming (XP), as it is flexible to changes and supports rapid iterations. The system consists of seven main modules: Finance, Human Resources, Supply Chain Management, Customer Relationship Management, Warehouse Management System, Document Management, and Business Intelligence. The technologies used include the CodeIgniter 4 framework, Bootstrap 5, Chart.js, PhpSpreadsheet, DomPDF, and MySQL as the database. The implementation results show that the system streamlines business processes, improves operational efficiency, and produces accurate real-time reports. The Business Intelligence module is capable of presenting financial and operational data in the form of graphical visualizations and insights based on actual data trends. The system has been tested using User Acceptance Testing and has demonstrated a high level of user satisfaction. This study proves that a web-based ERP is an effective solution to support the digitalization of business processes in small and medium enterprises (SMEs).

An Iot-Based Flood Early Warning System Integrated with the Sesaga Approach

Rachmat Kasim — 2026-07-24

Floods are among the natural disasters that significantly impact human safety, infrastructure, and the economy, particularly in flood-prone areas. Therefore, an early detection system capable of providing fast, accurate, and adaptive information is required to support risk mitigation efforts. This study develops a Flood Early Detection System based on the SESaGa approach (Systematic, Exploratory, SWOT, and GIS Analysis), which integrates Systematic Literature Review (SLR), Exploratory Data Analysis (EDA), SWOT, and GIS. Research data were obtained from a flood monitoring prototype installed at several flood-prone points in Gorontalo. The implementation results show that the SESaGa approach enhances early detection accuracy and provides valuable spatial information for mitigation planning. In addition, the integration of SWOT analysis generates strategic recommendations for strengthening the system and disaster management policies. The developed system was also tested using the blackbox testing method to ensure that the software functionalities meet user requirements. The test results demonstrate that all key features, including sensor data monitoring, warning notifications, and GIS-based visualization, functioned properly without critical errors. Thus, this system has the potential to serve as an essential instrument in supporting community resilience against flood threats and improving the effectiveness of mitigation measures in vulnerable regions.

Application of COBIT 5 Framework in the Evaluation of Information Technology Infrastructure to Improve Education Effectiveness (Case Study of SMK YPT Pringsewu)

Agus Ginanjar — 2026-07-24

The development of information technology (IT) in the world of education requires educational institutions to have effective and structured IT infrastructure governance. SMK YPT Pringsewu as a technology-based vocational school faces various problems related to IT infrastructure, especially in computer and network laboratories, such as inadequate device specifications, unstable internet connections, and suboptimal system management and monitoring. This study aims to evaluate the condition of information technology infrastructure in supporting the effectiveness of education using the ISACA framework, namely COBIT 5. This study uses a quantitative approach with a survey method through the distribution of questionnaires to 227 respondents consisting of students and teachers majoring in DKV and TKJ. The domains used in the evaluation are DSS01 (Manage Operations), BAI02 (Manage Requirements Definition), and MEA01 (Monitor, Evaluate and Assess Performance and Conformance). The analysis was carried out by measuring the maturity level based on the COBIT 5 capability model and conducting gap analysis between the current conditions and the expected conditions. The results show that the level of maturity of IT infrastructure governance at SMK YPT Pringsewu is at the Defined Process level (Level 3), but there are still some gaps in aspects of operational management, performance monitoring, and definition of IT needs. Therefore, improvements are needed in the form of the preparation of more structured standard operating procedures, improved device maintenance, optimization of infrastructure monitoring, and more targeted IT development planning. This research is expected to be a reference for SMK YPT Pringsewu in increasing the effectiveness of IT infrastructure management and supporting the improvement of the quality of the learning process.

Building a Digital Lexical Resource for Banyumasan Javanese: A Low-Resource Language Approach

Nisrina Hanifa Setiono — 2026-07-24

Banyumasan Javanese, widely recognized through the Ngapak dialect, remains culturally significant but is still underrepresented in reusable computational resources. This study develops a digital Banyumasan-Indonesian lexical corpus and frames it as a reusable research artifact rather than a static appendix. The corpus was constructed from a Banyumasan-Indonesian dictionary, normalized into a structured bilingual dataset, and packaged as an installable Python resource so that it can be used directly in computational experiments. The implemented resource supports dataset loading, Banyumasan lookup, Indonesian lookup, simple translation, structured translation analysis, batch translation, and corpus statistics. The resulting corpus contains 2,000 lexical pairs, 1,996 unique Banyumasan forms, 1,444 unique Indonesian equivalents, and 4 duplicated Banyumasan headwords that preserve lexical ambiguity from the source material. To demonstrate practical utility, the study includes a 100-sentence implementation example in which Banyumasan text is translated with the published banyumasan-corpus package and evaluated against Indonesian ground truth using the Indonesian-focused embedding model LazarusNLP/all-indo-e5-small-v4. The average semantic similarity rises from 0.4833 for direct Banyumasan-versus-ground-truth comparison to 0.6427 after translation, producing an absolute gain of 0.1594 and a relative improvement of approximately 33.0% over the baseline. These findings indicate that a structured lexical corpus, when distributed in a directly reusable computational form, can strengthen both resource accessibility and small-scale downstream experimentation for a low-resource regional language.

Feature Optimization for a Content-Based Music Recommendation System on Spotify

Viga Laksa Hardjanto — 2026-07-24

The growth of music streaming services like Spotify has encouraged users to explore millions of songs, making effective recommendation systems essential. This study examines content-based music recommendation systems by comparing feature configurations and similarity metrics. The system uses 11 Spotify audio features, both with and without genre features encoded using one-hot encoding. Similarity is calculated using cosine similarity and Euclidean distance on normalized features (MinMaxScaler) to generate the top 10 recommendations. Recommendations are considered relevant if the recommended song is by the same artist as the searched song. Performance is measured using Precision@10 and Recall@10 on 100 samples. Using audio features alone yields Precision@10 of 3.80–3.90% and Recall@10 of 6.36–6.45%. The addition of genre features improved performance to Precision@10 of 9.00–9.10% and Recall@10 of 9.06–9.89%. These results show that genre features significantly improve the relevance of recommendations, while both similarity metrics perform similarly when the features have been well normalized. This study contributes by demonstrating that feature representation plays a more critical role than similarity metrics in content-based music recommendation.

Airport Clustering in Indonesia Based on Domestic Air Traffic Using the K-Means Algorithm

Mohammad Rohmad Nurkhoirofiq — 2026-07-24

Indonesia, as an archipelagic country, relies heavily on air transportation to support interregional connectivity. Airports play an important role in supporting community mobility, economic activities, and equitable regional development. However, the level of air transportation activity at each airport in Indonesia varies and may change over time. This study aims to cluster airports in Indonesia based on domestic air transportation activity using the K-Means algorithm. The dataset used in this study consists of historical data from the period 2015–2023, including aircraft movements, passenger traffic, baggage volume, cargo volume, and mail shipments obtained from the HUBNET system of the Ministry of Transportation. The analysis process includes exploratory data analysis, data preprocessing, feature selection, and clustering modeling using the K-Means algorithm. The number of clusters is determined using the Elbow and Silhouette methods, resulting in candidate values of K = 2, 3, and 5. The evaluation results show that K = 2 has the best quality with a Silhouette Score of 0.6254, a Calinski Harabasz Index of 3543.93, and a Davies Bouldin Index of 0.5938. However, K = 3 is chosen as it is more representative, consisting of 260 low-activity airports, 1063 medium, and 627 high. Temporal analysis shows 233 cluster transitions across 123 airports, with the highest increase in 2016 (31 airports) and the highest decrease in 2018 (23 airports). The results indicate that airport activity is dynamic and that clustering can be used as a basis for data-driven airport planning and evaluation.

Analysis of Public Sentiment about Sea Games eSport on Twitter using the Adaboost Algorithm

Muhammad Syifa — 2026-07-24

This research aims to analyze public sentiment towards the Sea Games eSports as reflected in conversations on the Twitter platform. The two main issues in focus are how the public sentiment towards the Sea Games eSports event and whether their responses tend to be positive or negative towards the event. The research method used is the Adaboost Algorithm, a technique in data mining that aims to improve classification accuracy. Adaboost was used to analyze sentiment from Twitter conversation data by using feature selection to select weak classification functions, then combining them into a new classification function. The results showed that the highest evaluation was achieved in the 2nd test with the use of training data by 90% and testing by 10%, which resulted in an accuracy of 98%. Sentiment analysis of the Sea Games eSports showed that the majority of Twitter users expressed 111 (95.7%) positive sentiments, while only 5 (4.3%) negative sentiments. This indicates that people tend to give positive responses to the Sea Games eSports event. This research illustrates that sentiment analysis using the Adaboost Algorithm is able to provide accurate results in mapping public responses to the Sea Games eSports event on Twitter. The implication is that the strong support from the community towards this event can be an important basis for further development and promotion of Sea Games eSports.