header

Volume 1, No. 1July 2026

Published June 8, 2026

Issue description

SPECTRA: Journal of Computing, Analytics and Engineering is an international, peer-reviewed, open-access journal publishing high-quality research in computing, data analytics, and engineering. The journal covers emerging topics including artificial intelligence, machine learning, robotics, data analytics, IoT, cloud computing, cybersecurity, digital transformation, energy systems, and industrial engineering, fostering interdisciplinary innovation for sustainable technological advancement.

Articles

  1. Cover Image

    hendranusa Hendra, Very Fernando Very, Nanda Tommy Wirawan Tommy, Risa Nadia Ernes Icha (Author)

    Privacy Preserving Collaborative Data Mining on Electronic Medical Records Using Secure Multi-Party Computation: A Systematic Literature Review

    The rapid adoption of Electronic Medical Records (EMRs) across Indonesian healthcare facilities, mandated by Ministry of Health Regulation (PMK) No. 24 of 2022, has created an urgent need for privacy-preserving mechanisms that enable collaborative data analysis without compromising patient confidentiality. Secure Multi-Party Computation (SMPC) presents a cryptographically sound solution; however, its application in the Indonesian healthcare ecosystem remains critically underexplored. Objective: This systematic literature review aims to synthesize global evidence on privacy-preserving collaborative data mining using SMPC on EMRs, identify implementation challenges and opportunities, and provide a structured framework applicable within Indonesia’s regulatory landscape, particularly in light of the Personal Data Protection Law (UU PDP No. 27 of 2022). Following the PRISMA 2020 guidelines, we conducted a comprehensive search across six major databases (PubMed, IEEE Xplore, Scopus, Web of Science, ACM Digital Library, and Google Scholar) covering publications from January 2020 to December 2024. A total of 1,247 records were identified, and after rigorous screening and quality appraisal using the Mixed Methods Appraisal Tool (MMAT), 42 studies met the inclusion criteria. Results: The reviewed literature revealed three dominant SMPC paradigms applied to EMRs: secret sharing-based protocols (47.6%), garbled circuits (28.6%), and oblivious transfer-based approaches (23.8%). SMPC is frequently combined with Federated Learning (FL) and Homomorphic Encryption (HE) to overcome computational overhead. Key application domains include collaborative disease prediction (38%), genomic privacy analysis (24%), clinical trial management (19%), and cross-institutional diagnostic model training (19%). Identified barriers include high computational latency, communication overhead, interoperability constraints, and limited awareness among healthcare stakeholders. No peer-reviewed study was found implementing SMPC specifically within Indonesia’s health information infrastructure. The convergence of UU PDP, PMK 24/2022, and the SATUSEHAT national health data platform creates a regulatory and technical environment that both necessitates and enables SMPC adoption. A proposed tiered implementation model from bilateral hospital collaborations to national health research networks offers a roadmap for pragmatic deployment. SMPC represents a viable and necessary technology for privacy-preserving healthcare data collaboration in Indonesia. Future research should focus on lightweight protocol adaptations suited to low-resource health facilities, regulatory sandbox development, and capacity building within the Indonesian health informatics community.

  2. Cover Image

    Very Fernando, Kharisma Utomo Mulyodinoto, Soni Irawan (Author)

    Optimization of Power Grid Fault Detection Using Convolutional Neural Networks (CNN) in Smart Grid Systems: A Systematic Literature Review

    The rapid evolution of smart grid infrastructure has created an urgent need for highly accurate and efficient fault detection mechanisms that can operate in real-time across complex, interconnected power networks. Conventional fault detection approaches including relay-based systems, Support Vector Machines (SVM), and Artificial Neural Networks (ANN) have demonstrated limitations in terms of detection accuracy, adaptability to non-stationary signal characteristics, and robustness under noisy operational conditions. This paper presents a Systematic Literature Review (SLR) examining the application of Convolutional Neural Networks (CNN) as an advanced solution for optimizing fault detection in smart grid systems. Following the PRISMA 2020 framework, a comprehensive search across Scopus, IEEE Xplore, ScienceDirect, and Web of Science databases yielded 1,247 initial records, of which 86 peer-reviewed studies published between 2020 and 2025 met the final inclusion criteria. The synthesis of reviewed literature demonstrates that CNN-based models achieve superior detection accuracy consistently ranging from 98.5% to 99.99% compared to conventional SVM models (92–97%) and standard ANN approaches (94–99%). Key findings reveal that CNN architectures, particularly hybrid CNN-LSTM and 2D-CNN variants utilizing wavelet-based scalogram representations, effectively extract both temporal and spatial features from fault signals. The review identifies critical research gaps including limited deployment in real-world heterogeneous grid topologies, data imbalance challenges, and the computational overhead of deep architectures for edge deployment. These findings establish a clear foundation for future development of lightweight, optimized CNN models tailored for real-time smart grid fault detection.

  3. Rusbiansyah, Sofian, Very Fernando, Muhammad Hilman, Hendra Nusa Putra (Author)

    Real Time Fraud Detection in Indonesian Fintech Transactions Using Apache Kafka and Streaming Machine Learning with Concept Drift Adaptation: A Systematic Literature Review

    The proliferation of digital financial transactions in Indonesia has created an urgent need for robust, real-time fraud detection systems capable of adapting to continuously evolving fraud patterns. Conventional batch-processing fraud detection models are inadequate for dynamic streaming environments, particularly in the context of Indonesian fintech, where local transaction behaviors exhibit unique temporal and behavioral patterns not well represented in global datasets. This Systematic Literature Review (SLR) investigates the intersection of three critical domains: real-time data streaming architectures (specifically Apache Kafka), streaming machine learning algorithms, and concept drift adaptation mechanisms, applied to fraud detection in Indonesian fintech ecosystems. Following PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines, we conducted a comprehensive search across six academic databases—Scopus, IEEE Xplore, Web of Science, ScienceDirect, ACM Digital Library, and MDPI—covering publications from 2020 to 2024. From an initial pool of 847 articles, 42 studies met the inclusion criteria after rigorous screening. Our analysis reveals that while Apache Kafka has been increasingly adopted as a high-throughput event streaming backbone for fraud detection, its integration with adaptive streaming machine learning models that explicitly address concept drift remains critically underexplored, especially for emerging market contexts like Indonesia. Findings indicate that Adaptive Random Forest (ARF) and Hoeffding Adaptive Trees (HAT), combined with drift detectors such as ADWIN and EDDM, demonstrate superior performance in non-stationary financial streaming environments. We identify a significant research gap: the absence of locally calibrated, concept-drift-aware streaming ML frameworks for Indonesian fintech transaction patterns. This review contributes a structured research taxonomy, identifies key challenges including data imbalance, regulatory constraints (OJK/BI-FAST standards), and latency requirements, and proposes a conceptual architecture for future empirical work.

  4. Ii Sopiandi, Hendra Nusa Putra, Very Fernando, Nanda Tommy Wirawan (Author)

    Graph-Based Big Data Analysis for Community Detection and Disinformation Diffusion on Indonesian Social Media Platforms Using GraphX/GraphFrames

    Disinformation on Indonesian social media platforms has emerged as a serious threat to socio-political stability, with more than 77 million pieces of content identified in 2023. Nevertheless, large-scale graph analysis (>100 million nodes/edges) for the Indonesian language on distributed infrastructure remains critically limited, creating a significant research gap. This study presents the Graph-Based Disinformation Detection for Indonesia (GBD-ID), a distributed analytic framework built upon Apache Spark GraphX and GraphFrames. We constructed a multi-platform heterogeneous graph comprising 587.6 million nodes and 3,484.6 million edges, sourced from Indonesian Twitter/X, Facebook, TikTok, Instagram, and YouTube over the period of January–December 2023. A custom community detection algorithm was developed by integrating a Label Propagation Algorithm (LPA) optimized with BERT-Indonesian for sentiment analysis of Indonesian-language content. GBD-ID successfully identified 2,847 disinformation communities with an F1-Score of 0.915, outperforming the best baseline method by 6.8 percentage points. Disinformation diffusion patterns exhibited scale-free network characteristics with an exponential coefficient of γ = 2.34. Five dominant disinformation clusters were detected, encompassing health-related hoaxes (34.2%), political narratives (28.7%), identity-based hate content (18.9%), financial fraud (11.4%), and environmental disinformation (6.8%). The system achieved a throughput of 18.2 million edges per second on a 32-node Apache Spark cluster. This study demonstrates the feasibility of big data graph analysis at a scale exceeding 100 million nodes for Indonesian-language disinformation detection. The GBD-ID framework provides a scalable and efficient solution that can be adopted by social media platforms and Indonesian government agencies in real-time disinformation mitigation efforts.

  5. Nanda Tommy Wirawan, Risa Nadia Ernes, Very Fernando, Imam Syaifuddin (Author)

    Serverless Edge Computing for Real-Time Processing of Wearable Medical Sensor Data

    The proliferation of Internet of Things (IoT)-enabled wearable medical devices has generated unprecedented volumes of real-time biosignal data, creating significant challenges for conventional cloud-based processing architectures. Latency, bandwidth consumption, and data privacy remain critical barriers in clinical deployment. This paper proposes a serverless edge computing framework specifically designed for the real-time processing of wearable medical sensor data, including electrocardiogram (ECG), blood oxygen saturation (SpO2), body temperature, and tri-axial accelerometer signals. The proposed architecture leverages AWS Lambda@Edge and Cloudflare Workers deployed at geographically distributed edge nodes to execute lightweight inference models and anomaly detection algorithms within milliseconds of data acquisition. Experimental results demonstrate a mean end-to-end latency of 42 ms an 86.9% improvement over traditional cloud pipelines while maintaining 99.7% data accuracy and reducing bandwidth consumption by 73.2%. The framework also ensures HIPAA-compliant data handling through on device preprocessing and selective data forwarding. This work advances the feasibility of real-time clinical decision support systems powered by wearable technology in resource-constrained environments.