cover
Article Per Year (5 Year)
All
Home Page
OAI Link
Editorial Team
Contact
Reviewer
Google Scholar
Contact Name
Budi Hermawan
Contact Email
-
Phone
+62081703408296
Journal Mail Official
info@kdi.or.id
Editorial Address
Jl. Flamboyan 2 Blok B3 No. 26 Griya Sangiang Mas - Tangerang 15132
Location
Kab. tangerang,
Banten
INDONESIA
bit-Tech
Published by Komunitas Dosen Indonesia
ISSN : 2622271X     EISSN : 26222728     DOI : https://doi.org/10.32877/bt
Core Subject : Science,
Computer Science & IT
The bit-Tech journal was developed with the aim of accommodating the scientific work of Lecturers and Students, both the results of scientific papers and research in the form of literature study results. It is hoped that this journal will increase the knowledge and exchange of scientific information, especially scientific papers and research that will be useful as a reference for the progress of the State together.
Arjuna Subject : Ilmu Komputer (semua kategori) - Ilmu Komputer (Lain-Lain)
Articles 642 Documents
 45   46   47   48   49 
Machine Learning-Based Early Detection of Stunting and Intervention Recommendations Heti Mulyani; Musawarman Musawarman; Rifialdi Faturrohman; Daris Hafiz Permana
bit-Tech Vol. 8 No. 2 (2025): bit-Tech
Publisher : Komunitas Dosen Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32877/bt.v8i2.3213

Abstract

The stunting rate in Indonesia remains very high, one of the contributing factors being parents lack of knowledge about the symptoms of stunting and how to prevent it. Toddler checkups are only conducted once a month when health center staff are available, so parents are unable to detect stunting early on. This machine learning-based early stunting detection system offers a solution that allows parents to check for stunting in their toddlers at any time at home without having to wait for health center staff. Several previous studies have been conducted on stunting using machine learning, but they have not been integrated with expert systems for nutritional recommendations. The purpose of this study is to develop a machine learning-based early stunting detection system using a decision tree to quickly and accurately identify children at risk of stunting based on anthropometric indicators, namely height, age, and additional attributes such as gender. This study also aims to incorporate the knowledge of medical experts or nutritionists in the process of recommending interventions that parents should take. The model evaluation was conducted using the Confusion Matrix. Based on the research results from hybrid data obtained from the Sukasari community health center and Kaggle, the accuracy of the stunting classification model using decision trees was 98.7%. This model has been successfully implemented into a mobile-based application. Although the accuracy of this study is already high, it is hoped that future studies can be further improved by comparing other algorithms.
Implementation of the Weighted Product Method for Determining Poor Households Alya Izzah Zalfa Rihadah Ramadhani Nirwana Putri; Afina Lina Nurlaili; Muhammad Muharrom Al Haromainy
bit-Tech Vol. 8 No. 2 (2025): bit-Tech
Publisher : Komunitas Dosen Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32877/bt.v8i2.3214

Abstract

Many village-level poverty programs still depend on manual deliberation, which is slow to audit and difficult to reproduce across localities. This study addresses that gap by delivering an end-to-end, transparent implementation of the Weighted Product (WP) method for ranking poor households in Prunggahan Kulon, Tuban Regency. We assess whether a clearly specified WP pipeline complete with documented polarity (benefit/cost), normalized weights, and run logs can convert heterogeneous village records into reproducible preferences suitable for operational targeting. Household data supplied by the village and the Social Office were coded on a 0–1 scale for eight agreed criteria; expenditure (C2) was treated as a cost while others were benefits. Equal weights were used in this initial deployment for clarity and explainability. The method was implemented in a Laravel-based system that records bases, signed exponents, the multiplicative score , and normalized preferences . A five-household subset (A1–A5) is reported for illustration, with the full system supporting larger lists. The computation yielded a clear ordering (A4 > A1 > A2 > A3 > A5). The multiplicative rule preserved penalties for critical shortfalls and prevented strong indicators from masking severe deprivations, while the software artifacts ensured traceability from inputs to final . The dataset comprised 491 households encoded across eight criteria, with one cost criterion and seven benefits. Compared with prior WP applications, our contribution is an end-to-end, district-ready pipeline with explicit polarity, documented weights, and preserved run logs enabling third-party replication. This design measurably improves transparency and reproducibility for local poverty targeting.
Performance Evaluation of YOLOv5su and SVM With HOG Features for Student Attendance Face Recognition Achmad Rozy Priambodo; Achmad Junaidi; Muhammad Muharrom Al Haromainy
bit-Tech Vol. 8 No. 2 (2025): bit-Tech
Publisher : Komunitas Dosen Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32877/bt.v8i2.3215

Abstract

The rapid evolution of Artificial Intelligence (AI) and Computer Vision has revolutionized conventional attendance systems by introducing automated and intelligent alternatives. Traditional approaches such as manual entry and fingerprint-based systems are often inefficient, error-prone, and unsuitable for large-scale student management. This study evaluates a hybrid face recognition framework that combines You Only Look Once version 5 su, Histogram of Oriented Gradients (HOG), and Support Vector Machine (SVM) to automate student attendance. The YOLOv5su algorithm performs fast and lightweight face detection, while HOG extracts gradient-based facial descriptors classified by SVM. Experiments were conducted using a facial image dataset consisting of 500 original images from 10 classes (50 images per class), which were augmented to 3,500 images with variations in pose, expression, and illumination. The proposed YOLOv5sU–HOG–SVM model achieved 97.1% detection accuracy and 97% recognition accuracy, with mean precision, recall, and F1-score values of 0.98, outperforming conventional CNN-based hybrid models in both accuracy and computational efficiency. These results demonstrate that the combination of YOLOv5su, HOG, and SVM provides a novel balance between detection speed and recognition robustness, making it suitable for real-time academic attendance management. Future work should integrate transformer-based facial feature extraction to further enhance robustness under extreme conditions and larger-scale datasets.
Forecasting Financial Sector Stock Price and Loss Risk Using the ARIMAX and Value-at-Risk Methods Amanda Aulia; Trimono Trimono; Kartika Maulida Hindrayani
bit-Tech Vol. 8 No. 2 (2025): bit-Tech
Publisher : Komunitas Dosen Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32877/bt.v8i2.3219

Abstract

Stock price volatility remains a persistent challenge in financial forecasting, as traditional ARIMA-based models often neglect the role of macroeconomic forces, leading to limited predictive robustness. Addressing this methodological gap, this study uniquely integrates the Autoregressive Integrated Moving Average with Exogenous Variables (ARIMAX) model and the Value-at-Risk (VaR) framework to simultaneously predict stock prices and quantify investment risk. This dual approach advances prior forecasting literature by merging predictive modeling and risk assessment within a single analytical structure. Using daily data from PT Bank Central Asia Tbk (BBCA) and the USD/IDR and SGD/IDR exchange rates from January 2019 to September 2024, model identification through ACF, PACF, and the Akaike Information Criterion (AIC) identifies ARIMAX(0,1,1) as optimal. The model achieves a Mean Absolute Percentage Error (MAPE) of 2.19%, indicating very high predictive accuracy. Although forecasted movements appear smoother than observed fluctuations, the model effectively captures short-term market trends influenced by exchange rate dynamics. Historical simulation at a 95% confidence level estimates a daily Value-at-Risk (VaR) of 1.71%, implying a potential loss of approximately Rp17,144 per Rp1,000,000 invested. These results demonstrate that integrating ARIMAX with VaR not only enhances statistical precision but also provides practical value for investors and policymakers. The combined framework enables evidence-based decision-making, portfolio optimization, and risk mitigation in volatile capital markets, offering a replicable and data-driven model for financial forecasting under macroeconomic uncertainty.
Chatbot for Clothing Color Recommendations Based on Skin Tone Rahayu Ramli; Anna Dina Kalifia
bit-Tech Vol. 8 No. 2 (2025): bit-Tech
Publisher : Komunitas Dosen Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32877/bt.v8i2.3220

Abstract

In the era of global digital fashion and personalized styling technologies, the ability to match clothing colors with individual skin tones has become increasingly important for enhancing self-expression and confidence. Many people struggle to choose clothing colors that match their skin tone due to limited knowledge of proper color combinations. As a result, they often select outfits based on trends or personal taste without considering compatibility with their skin tone, which affects confidence and comfort in appearance. This study aims to develop a novel Natural Language Processing (NLP)–based chatbot that uniquely interprets textual descriptions of skin tone to recommend suitable clothing colors. Users input their skin tone, and the chatbot analyzes it, classifying the input into appropriate skin tone and undertone categories. The data were obtained from interviews with personal color analysts and color theory in fashion. The research involves system requirement analysis, chatbot architecture design, and the creation of flowcharts, use case diagrams, and activity diagrams to describe user–system interactions. Quantitative evaluation shows that the implemented chatbot achieves over 80% accuracy in recognizing textual skin tone descriptions and delivers responses within an average of 1.8 seconds, demonstrating strong empirical performance. It can also suggest matching clothing colors and indicate those to avoid. This system enables users to obtain suitable clothing color recommendations quickly and interactively. The study highlights the growing role of AI-driven interaction design in modern fashion systems and positions the model as a bridge between linguistic input and aesthetic recommendation technologies.
Implementation of a Mobile Application for GPS-Based Reporting of Water Pollution and Invasive Species Muchammad Hasbi Ashshiddiqi; Muhammad Fahrie
bit-Tech Vol. 8 No. 2 (2025): bit-Tech
Publisher : Komunitas Dosen Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32877/bt.v8i2.3222

Abstract

Water pollution and the spread of invasive species threaten Indonesian river ecosystems, yet public participation in monitoring remains limited, often due to the lack of a centralized and accessible reporting tool. This study develops and validates a mobile platform to address this gap. Its primary contribution is a novel dual-reporting model, allowing the public to report both pollution incidents and invasive species findings using a single GPS-based application. This integrated approach provides a more holistic dataset than single-issue tools, critically enabling the future analysis of ecological links between pollution hotspots and invasive species outbreaks. The application was developed using Flutter for cross-platform accessibility and Supabase as a Backend as a Service to ensure scalability and rapid development. The main finding, confirmed through Black Box testing, is a functional prototype where all core user-side features, such as registration, login, and GPS report submission, performed successfully. This outcome validates the system's technical feasibility and the practical viability of using non-experts for data collection. While user-side functionality is proven, the report management functionality for administrators remains in the early stages. This research provides a viable tool for agencies to gather real-time field data. It also supports future development focused on a comprehensive admin dashboard, which is essential for report validation, data aggregation, and trend analysis, ultimately enhancing the system's effectiveness.
Comparative Performance Analysis Between the MQTT and WebSocket Protocols Muhammad Faishal Tsaqief; Joko Sutopo
bit-Tech Vol. 8 No. 2 (2025): bit-Tech
Publisher : Komunitas Dosen Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32877/bt.v8i2.3223

Abstract

In the era of smart monitoring and real-time 3D visualization, communication delay remains a critical challenge for resource-constrained IoT devices. This study aims to determine which protocol MQTT or WebSocket delivers superior performance for real-time sensor data streaming in such latency-sensitive contexts. A controlled experimental setup was implemented using an ESP32-S3 microcontroller and an MPU-6050 inertial sensor, streaming 1,000 data samples per protocol to a Python-based server hosting a local Mosquitto MQTT broker. The evaluated performance metrics included mean latency, jitter (standard deviation of latency), and packet loss rate. Results indicate that under the tested LAN environment with a local broker, MQTT (using QoS level 0) significantly outperformed WebSocket. MQTT demonstrated substantially lower mean latency (11.040 ms vs. 41.544 ms), markedly reduced jitter (0.201 ms vs. 18.824 ms), and a superior packet loss rate (0.501% vs. 0.937%), showcasing exceptional stability and timing consistency for real-time data delivery. In contrast, WebSocket exhibited significantly higher latency and jitter, which would be detrimental to motion-sensitive applications. These findings challenge the common assumption favoring WebSocket for low-latency tasks and suggest that MQTT offers a more robust and suitable communication protocol for real-time 3D visualization on resource-constrained IoT devices within local network conditions. The superior performance of MQTT in this context provides a strong rationale for its adoption in edge computing-based visualization systems, where timing consistency is paramount.
Evaluating Multi-Party Verification for Land Certificates Using Hyperledger Fabric Rizqulloh Brilliant 'Ainur Rofiq Rofiq; Joko Sutopo
bit-Tech Vol. 8 No. 2 (2025): bit-Tech
Publisher : Komunitas Dosen Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32877/bt.v8i2.3224

Abstract

Blockchain technology provides a secure and transparent framework for managing land ownership records. However, most existing systems lack mechanisms that allow multiple authorities to verify transactions collaboratively before they are finalized. This study evaluates the implementation of a multi-party verification model for land certificate transactions using Hyperledger Fabric, a permissioned blockchain platform that supports configurable endorsement policies to simulate multi-signature behavior. The system architecture involves three endorsing organizations: BPN (National Land Agency), PPAT (Land Deed Officials), and the Local Government, each acting as an independent validator within the blockchain network. A series of performance tests were conducted to measure transaction latency, endorsement success rate, and throughput under single-, two-, and three-party endorsement configurations. Results show that the average latency increased from 1.13 seconds under a single-party policy to 2.43 seconds under a three-party policy, while throughput declined from 63.4 to 37.2 transactions per second. Despite this reduction, the system maintained a 100 percent endorsement success rate across all configurations, indicating consistent policy enforcement by Hyperledger Fabric’s validation system chaincode. The findings highlight the potential of Hyperledger Fabric as a foundation for transparent and auditable land registration. This approach directly supports Indonesia’s digital governance framework by providing a technical model for secure inter-agency coordination, which is critical for the national transition to electronic land certification and enhancing institutional accountability. Key limitations of this study, representing directions for future work, include the controlled simulation environment and the need for integration with Indonesia’s national Public Key Infrastructure (PKI) to ensure full legal compliance.
IoT-Based Automated Catfish Feeder Applicatoin: A Case Study at Dompon Sejahtera Fishery Kisan Rozin Asrigen; Rr. Hajar Puji Sejati
bit-Tech Vol. 8 No. 2 (2025): bit-Tech
Publisher : Komunitas Dosen Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32877/bt.v8i2.3228

Abstract

Conventional catfish aquaculture in Indonesia often suffers from inefficient feed management due to manual feeding, leading to feed waste, poor water quality, and inconsistent fish growth. To address these limitations, this study develops a user-friendly Internet of Things (IoT)-based automatic catfish feeding system that integrates automation, real-time monitoring, and simplified configuration. The system employs an ESP32 microcontroller as the core hardware and Firebase as a cloud-based backend for instantaneous data synchronization. A key innovation is the inclusion of a local web server on the ESP32, enabling farmers to configure Wi-Fi networks directly through a browser without reprogramming the device. The companion mobile application, “Feedy Finn,” built using Flutter, allows users to remotely set feeding schedules, control operations manually, and monitor device performance in real time. Black Box testing demonstrated stable two-way communication with a latency of under 500 milliseconds and a feed dispensing consistency exceeding 95%, reducing feed waste by approximately 20% compared with manual methods. The results confirm that the system enhances operational efficiency, ensures precise feeding intervals, and simplifies device setup, making it accessible to small and medium-scale fish farmers. Beyond technical performance, the study contributes to sustainable aquaculture by minimizing environmental impact and promoting digital inclusion in rural communities. This integration of affordability, usability, and automation highlights the system’s potential as a scalable model for advancing smart aquaculture and supporting Indonesia’s transition toward technology-driven fisheries.
Indonesian Sign Language (SIBI) Recognition from Audio Mel-Spectrograms Using LSTM Architecture Enryco Hidayat; Mohammad Idhom; Afina Lina Nurlaili
bit-Tech Vol. 8 No. 2 (2025): bit-Tech
Publisher : Komunitas Dosen Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32877/bt.v8i2.3229

Abstract

Persistent communication barriers continue to challenge Deaf and Hard of Hearing (DHH) individuals in accessing spoken language, underscoring the need for effective and inclusive translation technologies. Existing audio-to-sign language systems typically employ multi-stage pipelines involving speech-to-text transcription, which may propagate recognition errors and fail to preserve acoustic nuances. Addressing these limitations, this study developed and evaluated a deep learning framework for translating spoken Indonesian audio directly into classifications of the Indonesian Sign Language System (SIBI), eliminating explicit text conversion. The dataset comprised 495 eight-second WAV recordings (22,050 Hz) representing five SIBI phrase classes, augmented through time stretching, pitch shifting, and noise addition to improve generalization. Mel-Spectrogram features were extracted and input to a stacked Long Short-Term Memory (LSTM) network implemented in TensorFlow/Keras, trained to learn temporal–spectral mappings between audio patterns and SIBI categories. Evaluation on a held-out test set demonstrated robust performance, achieving 98 % accuracy with consistently high precision, recall, and F1-scores. The trained model was further integrated into a prototype web application built with Flask and React, confirming its feasibility for real-time assistive communication. While results highlight the viability of direct Mel-Spectrogram-to-LSTM translation for SIBI recognition, current findings are constrained by the limited dataset size and restricted speaker diversity. Future research should therefore expand the dataset to include more speakers, varied acoustic environments, and continuous-speech inputs to ensure broader applicability and real-world robustness.

Page 47 of 65 | Total Record : 642

 45   46   47   48   49 

Filter by Year

2018 2026


Reset
Filter By Issues
All Issue Vol. 8 No. 3 (2026): bit-Tech Vol. 8 No. 2 (2025): bit-Tech Vol. 8 No. 1 (2025): bit-Tech Vol. 7 No. 3 (2025): bit-Tech Vol. 7 No. 2 (2024): bit-Tech Vol. 7 No. 1 (2024): bit-Tech Vol. 6 No. 3 (2024): bit-Tech Vol. 6 No. 2 (2023): bit-Tech Vol. 6 No. 1 (2023): bit-Tech Vol. 5 No. 3 (2023): bit-Tech Vol. 5 No. 2 (2022): bit-Tech Vol. 5 No. 1 (2022): bit-Tech Vol. 4 No. 3 (2022): bit-Tech Vol. 4 No. 2 (2021): bit-Tech Vol. 4 No. 1 (2021): bit-Tech Vol. 3 No. 3 (2021): Remote Delivery Vol. 3 No. 2 (2020): Datamining Vol. 3 No. 1 (2020): Distance Learning Vol. 2 No. 3 (2020): Pandemik ICT Vol. 1 No. 3 (2019): Learning Synchronous and Asynchronous in the Industry 4.0 Vol. 2 No. 2 (2019): Support System Vol. 2 No. 1 (2019): Data Mining and Green Technology Vol. 1 No. 2 (2018): Data and Information Quality Vol. 1 No. 1 (2018): First Impression on Computing More Issue