Publication Search

Agentic artificial intelligence (AI) systems capable of autonomous goal-directed behavior, multi-step planning, tool use, multi-agent coordination, and iterative self-correction represent a transition from passive clinical AI tools toward systems that can participate in complex healthcare workflows. However, empirical evidence remains fragmented across clinical decision support, patient monitoring, and administrative applications, and no systematic synthesis has evaluated which agentic principles have been technically demonstrated and which have accumulated sufficient evidence to support responsible clinical deployment. We conducted a PRISMA-informed systematic review of peer-reviewed empirical studies published between January 2025 and April 2026. Searches across five bibliographic databases and Google Scholar, supplemented by citation tracking, identified 443 unique records for screening, of which 25 met the predefined PICOS and quality appraisal criteria. Evidence was synthesized using an evidence-informed seven-principle framework derived from the integration of agentic AI, clinical AI, and healthcare governance literature. This framework provides a structured lens for examining how agentic principles are evaluated individually and in combination, enabling a deployment-readiness perspective that extends beyond capability-focused assessments alone. The evidence base was concentrated on technical capability principles, whereas human oversight, safety, compliance, and equity-related evaluation received comparatively limited attention. Most studies remained at the laboratory, benchmark, or proof-of-concept stage, and none reported demographic-stratified performance outcomes. Overall, the findings suggest a structural asymmetry in agentic healthcare AI: empirical research is advancing agentic capabilities more rapidly than it is generating evidence for the oversight, safety, equity, and governance mechanisms required for responsible clinical translation.

Malware represents an evolving cybersecurity threat that demands more effective detection methods. Conventional signature-based detection systems have limitations in identifying new variants, driving the development of deep learning-based approaches. This research implements and evaluates four variants of the YOLOv11 algorithm (n, s, m, l) for malware classification based on visual image representation. The dataset consists of 22,056 malware and benign images, divided into 70% training, 15% validation, and 15% testing across 8 classes (adware, backdoor, benign, downloader, spyware, trojan, virus, worm). Each model was trained for 100 epochs with batch size 32 using Google Colab with GPU support. Results demonstrate that all variants achieve high accuracy (97.8%-98.1%) with YOLOv11m as the best performer (98.1%). YOLOv11n offers optimal balance between accuracy (97.9%) and efficiency (1.5M parameters, 0.3 ms/img inference) ideal for real-time applications. This research surpasses previous methods such as K-NN (97.18%) and hybrid CNN (96.55%) with superior inference speed (0.3-0.9 ms/img vs tens to hundreds of ms/img), proving the effectiveness of YOLOv11 for fast, accurate, and scalable malware detection.

Online maps applications have become an essential tool for modern society in finding the fastest and most efficient routes. However, these applications often fail to detect current road conditions such as flooding, demonstrations, accidents, or public events, causing users to get stuck in problematic routes. This study aims to develop a prototype of a community-based road condition reporting system, visualized through a web-based digital map. The system allows users to directly report road conditions by providing photo evidence, descriptions, and event categories. It is also equipped with features for designing event routes such as carnivals and suggesting alternative paths based on community reports. The development process was carried out using a simulation-based approach with scenario testing that reflects real field conditions, without involving direct user data. The implementation results show that all core features work properly. The technologies used include Leaflet.js, OpenStreetMap, and the Nominatim geolocation API. This research produces an adaptive community-based GIS model that can be further developed as an intelligent navigation solution at the city scale

This study investigates the performance of Internet of Things (IoT)-based monitoring systems using a mobile hotspot and IoT sensors for temperature and humidity data transmission. The research is based on the IoT concept, which enables electronic devices to communicate and exchange data through internet networks without direct human intervention. System performance was evaluated using standard Quality of Service (QoS) parameters, including throughput, packet loss, delay, and jitter. The experimental setup utilized a NodeMCU ESP32 microcontroller and a DHT22 sensor, with measurements conducted at various transmission distances through wireless communication media. The objective was to determine the reliability of hotspot connectivity and sensor communication in supporting IoT applications. The results indicate that the optimal performance was achieved at a distance of 20 meters using a 40-lambda variation. Furthermore, the communication signal between the ESP32 device and the mobile hotspot remained detectable up to a maximum distance of 32 meters. These findings demonstrate the effectiveness of the proposed IoT system for environmental monitoring applications within specific transmission ranges.

The dissemination of personal data through digital media has increased significantly alongside the growing use of Quick Response (QR) Codes for various purposes, such as electronic tickets, certificates, and digital identities. Conventional QR Codes are open and can be easily scanned, copied, or manipulated by unauthorized parties. The personal data referred to in this study includes sensitive information such as full name, identity number (NIK/National ID), date of birth, address, phone number, and email address. This research proposes a layered security system that combines the Advanced Encryption Standard (AES) cryptographic algorithm with steganography using the Discrete Cosine Transform (DCT) method. The process begins with encrypting personal data using AES, converting the encrypted result into a QR Code, and embedding the QR Code into a digital image using DCT, hiding it in the image’s frequency domain. The digital images used are of fixed size and formats that preserve visual quality. System evaluation is carried out by testing the visual quality of the stego image, the success rate of QR Code extraction, and the integrity of the encrypted data. The results are expected to conceal sensitive information visually while maintaining its confidentiality, with potential applications in electronic ID cards, digital certificates, e-tickets, and other confidential documents.

Computers in one of the network companies at PT. Estrada uses the Fortinet operating system. The final result expected through this implementation is to comprehensively see the capabilities of the firewall on Fortinet in overcoming the problem of blocking social media applications and streaming platforms during working hours. Blocking the application in question is the ability to filter web processes such as Facebook, Instagram, YouTube, etc. In the tests carried out, web filtering was able to block applications on social media and streaming platforms, which proves that the performance of web filtering is quite good. In analyzing web filtering performance, use the office hour rule tool by carrying out the rule schedule in the Fortinet network and displaying all the information in detail. The final result obtained in the network application filtering simulation process using Fortinet is that every network sent cannot be entered (blocked) on both social media applications and streaming platforms.

Usability evaluation plays a crucial role in ensuring the quality of digital systems, particularly in terms of comfort, effectiveness, and ease of use. Instruments such as the System Usability Scale (SUS), User Experience Questionnaire (UEQ), and Heuristic Evaluation (HE) are widely used in modern usability studies. This research conducts a Systematic Literature Review (SLR) to identify patterns and trends in the use of these instruments. A total of 27 initial studies were collected, and 16 were selected through the PRISMA screening procedure. The findings show that UEQ is the most frequently used instrument, especially in Learning Management Systems (LMS) and academic platforms, while SUS is commonly applied to mobile applications and digital libraries for rapid usability assessment. HE is effective in revealing fundamental interface issues such as non-intuitive navigation and layout inconsistencies. Overall, digital systems perform well in Efficiency and Perspicuity, but consistently show low scores in Novelty. This study provides an integrative knowledge map that highlights cross-instrument insights and supports the development of more intuitive, innovative, and user-centered digital systems

  Red dragon fruit contains bioactive compounds such as vitamin C, flavonoids, and polyphenols that have antioxidant potential and possess natural pigments that can be used as cosmetic colorants. This study aims to formulate and evaluate a blush-on powder preparation from dragon fruit pulp extract as an alternative natural colorant. The evaluation includes organoleptic testing, homogeneity, pH, irritation, spreadability, fineness, moisture content, and antioxidant activity. The results showed that the formulation met most of the requirements, exhibiting a pink color, a smooth texture, no odor, homogeneity, a pH of 6 suitable for the skin, no irritation, even spreadability within 4 applications, a fineness degree of 97.6%, and a moisture content of 6.91%. Antioxidant activity showed an IC₅₀ value of 0.3036 ppm (very strong), although the relative inhibition percentage was low (18.35%). Thus, dragon fruit extract has the potential as a natural ingredient in blush-on powder that provides both color and antioxidant benefits.

Scientific writing is one of the essential skills that college students need to support research activities and the development of knowledge. However, students still encounter various challenges during the writing process. This study aimed to identify students’ responses toward scientific activities at FMIPA Universitas Negeri Surabaya and to examine the difficulties experienced by students in preparing scientific papers. The study employed a qualitative descriptive method using a purposive sampling technique. The research subjects consisted of 36 students from various study programs at FMIPA Universitas Negeri Surabaya who had been involved in research activities and scientific proposal writing. Data were collected through student response questionnaires consisting of closed-ended and open-ended questions. The data were analyzed descriptively and presented in tables and narrative descriptions. The results showed that most students had understood the existing student research programs and were familiar with the use of research guidelines. However, students still faced difficulties in systematically organizing research ideas, determining research variables and methods, finding relevant literature reviews, and connecting theories from various references. In addition, time management and the use of citation management applications were also obstacles for some students. Therefore, more structured assistance is needed through scientific writing training, reference searching training, and the use of citation management applications to improve the quality of students’ scientific writing.

This study aims to examine the influence of trust on the perception of transaction security in the use of online shopping applications among millennials. The development of digital technology and the increasing use of online shopping applications have influenced how users assess security risks, personal data protection, and the reliability of transaction systems. This study employs a literature review method by examining relevant previous studies on trust, transaction security, digital risk, and user experience in online shopping. The findings indicate that user trust plays an important role in shaping perceptions of transaction security. Such trust is influenced by the platform’s ability to protect personal data, provide secure payment systems, and ensure smooth transaction processes. The higher the level of user trust in system security, the greater their tendency to use online shopping applications for transactions. Therefore, trust and transaction security are important factors in increasing millennials’ interest and comfort in online shopping.

Non-performing loans represent one of the risks faced by banks in their lending activities, particularly in housing loan (KPR) products. This study aims to analyze the procedures for resolving problematic housing loans and to identify the factors causing the decline in debtors’ repayment ability, as well as the obstacles encountered in the implementation of such procedures at PT Bank Tabungan Negara (Persero) Tbk, Tasikmalaya Branch Office. This research employs a qualitative method with a descriptive approach. Data were collected through direct observation and interviews with relevant parties within the bank. The results show that the procedures for resolving problematic loans are carried out through several stages, including submission of restructuring applications, document verification, analysis of the debtor’s repayment capacity, determination of restructuring schemes, approval, agreement signing, and post-restructuring monitoring. These procedures are supported by the application of the 3R concept, namely rescheduling, reconditioning, and restructuring, which are implemented flexibly according to the debtor’s condition. The findings also indicate that the decline in debtors’ repayment ability is mainly caused by decreased income, job loss, increased living expenses, and unstable economic conditions. In practice, several obstacles were identified, such as incomplete documentation, lack of debtor cooperation, and issues related to collateral ownership that has been transferred from the original debtor. Therefore, improved supervision, better communication, and stronger coordination between the bank and debtors are necessary to ensure the effectiveness of loan resolution procedures.

This study analyzes the implementation of public service innovation through the E-TTSP application at the Investment and One-Stop Integrated Services Agency (DPMPTSP) of Banten Province. Using a descriptive qualitative method, data were collected through interviews, observation, and document analysis. The results show that this digital system has significantly improved service quality and efficiency. Processing time was reduced from 7–14 days to only 2–5 days, while bureaucratic procedures were simplified. The system also enhances transparency, accountability, and accessibility, allowing users to apply online without visiting the office. Consequently, the investment climate improved, marked by a 28% increase in permit applications and a 35% rise in investment value within two years. However, challenges remain, including uneven network infrastructure, low digital literacy, and limited human resource capacity. It is recommended to improve infrastructure, provide continuous training, and intensify socialization. In conclusion, E-TTSP has delivered tangible benefits, yet sustainable development is necessary to ensure equal and optimal services for all.

The increasing complexity of urban road networks demands intelligent navigation systems capable of determining optimal routes efficiently. This research implements the Dijkstra Shortest Path algorithm to optimize route search on a location navigation system in Medan City. The system models a road network as a weighted graph comprising 57 strategic locations and over 90 road connections, represented using adjacency list data structures. The Dijkstra algorithm, implemented in Python using the heapq module for priority queue management, achieves an optimal time complexity of O((V+E) log V). The system features five main functions: shortest route search, popular routes, location listing, dynamic location addition, and dynamic road connection addition. System testing using a case study from Kualanamu Airport to the University of North Sumatra (USU) yielded an optimal route of 16.5 km through 4 road segments. Results demonstrate that the system successfully determines the most efficient route, provides accurate distance and travel time information for multiple transport modes (motorcycle, car, walking), and presents step-by-step journey guidance. This research contributes as a practical reference for applying shortest path algorithms in urban areas and serves as a foundation for developing more complex navigation applications in the future.

Suspicious urban sound recognition is a critical component of intelligent public safety and urban monitoring systems, enabling the automated identification of anomalous acoustic events such as gunshots, sirens, and other security-sensitive sounds. However, existing deep learning approaches often struggle to simultaneously capture long-range temporal dependencies and global contextual relationships, particularly under noisy and acoustically complex urban conditions. This limitation can reduce reliability in safety-critical scenarios where missed detections carry significant risk. To address these challenges, this study proposes a Multi-Branch Bidirectional Long Short-Term Memory (BiLSTM) framework with Multi-Head Self-Attention (MHSA) for enhanced sequential and contextual feature modeling. Mel-frequency cepstral coefficients (MFCCs) are extracted from a curated subset of the UrbanSound8K dataset, comprising five suspicious sound classes, and used as input to the proposed architecture. The multi-branch design enables complementary temporal representations, while the self-attention mechanism provides lightweight contextual weighting of BiLSTM outputs. Experimental results demonstrate that the proposed model achieves a test accuracy of 95.59%, outperforming conventional Dense and LSTM-based baseline models under identical experimental settings. An ablation study further confirms the contribution of multi-branch integration and attention-based enhancement to overall performance. Class-wise evaluation reveals consistently high recall across all sound categories, particularly for safety-critical classes such as gunshots and sirens. These findings indicate that the proposed framework provides robust and reliable performance, making it suitable for real-time smart city surveillance and public safety applications.

This study aims to design and develop an Android-based public infrastructure damage reporting application on Lepat Island, South Bangka Regency, as an effort to support the Smart Island concept. The background of this study is the continued use of a manual reporting system that causes delays in handling and a lack of transparency. The method used is Agile which includes the stages of system design, implementation, testing and periodic evaluation. Data collection techniques are carried out through observation, interviews, and literature studies. The results of the study are an Android-based application that allows the public to report infrastructure damage with features to upload photos, descriptions, and GPS-based locations. In addition, the application provides report history and status monitoring features to increase transparency. The implementation results show that the use of mobile technology can improve service efficiency, speed up the reporting process, and increase community participation. This application is expected to support the implementation of Smart Governance and Smart Island in the region.

Accurate classification of rainfall intensity patterns is important for early warning systems, hydrometeorological risk assessment, and water resource management. Surface rain gauges have limited spatial coverage, so this study uses NOAA NEXRAD Level II radar data from the KTLX station in 2023. K-Means clustering was applied to identify rainfall intensity patterns from 30 randomly selected days, with scans stratified into four daily time intervals. Seven features were extracted from each radar sweep, including reflectivity statistics, convective and stratiform ratios, and rainfall coverage. The data were normalized and balanced before clustering. The optimal cluster count was determined through a combined evaluation of the Elbow Method, Silhouette Score, and Davies-Bouldin Index, yielding K=5 as the most representative configuration. Evaluation results demonstrated a Silhouette Score of 0.3871 and a Davies-Bouldin Index of 0.8599, indicating moderate cluster cohesion that reflects the inherent overlapping nature of rainfall intensity transitions in radar reflectivity data. The clusters represent rainfall regimes from non-precipitating conditions to intense convective events. These results support the use of K-Means for automated rainfall pattern recognition and flood forecasting applications. 

This study aims to test the influence of digital literacy and the use of accounting applications on the effectiveness of MSME information systems with user competence as a mediating variable. The research sample amounted to 100 MSME actors in Gresik City who were selected using purposive sampling techniques according to the research criteria. Data testing was carried out using a quantitative approach using the Structural Equation Modeling–Partial Least Square (SEM-PLS) method through the SmartPLS application. The results of the study show that digital literacy and the use of accounting applications have a positive effect on the effectiveness of MSME information systems. Meanwhile, user competence does not have a significant effect and is unable to mediate the influence of digital literacy and the use of accounting applications on the effectiveness of information systems. This research is expected to provide benefits for MSME actors in increasing the use of digital technology and accounting applications, as well as become an academic reference for future research. The limitations of this study lie in the scope of the MSME scale, which has not considered external factors, and the research model that has not fully explained the role of user competence as a mediating variable.

The increasing demand for electrical energy and the limited availability of fossil fuels have driven the development of renewable energy sources, including marine current energy, which remains underutilized in coastal and remote maritime regions. This study presents the design and realization of a small-scale marine current power generation prototype using a horizontal axis propeller turbine with a NACA S814 blade profile and analyzes the effect of turbine rotational speed on electrical power output. The system converts marine current kinetic energy into mechanical energy through turbine rotation and subsequently into DC electrical energy using a generator, which is stabilized by a Buck–Boost Converter and Maximum Power Point Tracking (MPPT) for charging a 12 VDC battery. Real-time monitoring of electrical and mechanical parameters is implemented using an Internet of Things (IoT)–based system comprising an ESP32 microcontroller, a PZEM-017 sensor, and an RPM sensor. Experimental results demonstrate a positive correlation between water flow rate, turbine rotational speed, and generator output voltage. The system begins operating at a minimum flow rate of 35.2 L/s at 56 RPM, producing 0.2 V, while optimal performance is achieved at 45.3 L/s and 516 RPM, generating up to 13.3 V. These results indicate that the proposed prototype is a viable alternative renewable energy source for marine applications.

The rapid development of Artificial Intelligence (AI) has brought significant changes to the criminal justice system, particularly in criminal investigations and evidentiary processes, while simultaneously raising complex legal and ethical challenges. Objective: This study aims to analyze the legal implications of the use of AI in criminal investigations, focusing on its benefits, risks, and challenges related to the admissibility of AI-based evidence, as well as the need for regulatory frameworks that ensure fairness, transparency, and accountability. Methods: This research employs a normative qualitative approach through the analysis of legal regulations, a review of legal and technological literature, and a comparative approach across jurisdictions, complemented by case studies of AI applications in law enforcement practices. Results: The findings indicate that AI enhances investigative efficiency through data analysis, crime prediction, and digital forensics; however, it also poses risks such as algorithmic bias, human rights violations, and issues concerning the reliability and transparency of evidence. Furthermore, differences across legal systems result in the absence of uniform standards for the admissibility of AI-based evidence. Therefore, adaptive regulatory frameworks grounded in the principles of fairness, transparency, and accountability are required, along with strengthened human oversight to ensure that the use of AI aligns with the principles of justice and human rights protection.

The rapid development of information technology has had a significant impact on the world of education, particularly in the use of mobile applications as learning media. Developing Android-based applications using Flutter is one innovation that can increase efficiency and provide a more personalized and adaptive learning experience. In this context, educators are required to continue innovating to meet the increasingly diverse and dynamic needs of students in the digital era. The Multimedia and Information Technology Professional Association (PPMultindo), an organization that brings together technology professionals, is taking advantage of this opportunity by developing an Android-based learning platform. This platform is designed to provide easier, more flexible, and more interactive access to learning materials, as well as support the process of designing more engaging materials. Features such as learning data analysis, material recommendations, and automatic evaluation are expected to increase the effectiveness and efficiency of the teaching and learning process, while also providing a means for collaboration between members. However, the implementation of this technology still faces various challenges, such as limited technological competence of educators, inadequate infrastructure, and the need for curriculum adjustments. Therefore, collaborative efforts are needed to optimize the use of mobile technology to support educational transformation.