cover
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
ISSN : 2622271X     EISSN : 26222728     DOI : https://doi.org/10.32877/bt
Core Subject : Science,
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.
Articles 642 Documents
A SMART-Based Multi-Criteria Decision Support System for Oil Palm Fertilizer Selection Wahyu, Meidy Fajar; Mahfudza, Liza Wardhatul
bit-Tech Vol. 8 No. 3 (2026): bit-Tech
Publisher : Komunitas Dosen Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32877/bt.v8i3.3683

Abstract

The selection of appropriate fertilizers is a critical factor influencing productivity and cost efficiency in oil palm plantations. In practice, fertilizer decisions are frequently based on subjective experience and fragmented information, while existing decision approaches rarely integrate agronomic suitability, economic considerations, and local operational constraints within a single transparent framework. This gap limits consistent and evidence-based fertilizer selection, particularly for plantation managers and smallholders operating under price volatility and uneven input availability. To address this limitation, this study proposes a multi-criteria Decision Support System (DSS) based on the Simple Multi-Attribute Rating Technique (SMART) for fertilizer selection in oil palm cultivation. The proposed model incorporates four evaluation criteria: water solubility, price, expert recommendation, and local availability. Criterion weights were determined through expert consultation, and seven fertilizer alternatives were evaluated using normalized utility values and weighted aggregation to generate preference rankings. The results show that Phonska fertilizer achieved the highest preference score, followed by Pelangi and Mutiara fertilizers, indicating that SMART effectively structures multi-dimensional decision problems into interpretable outcomes. To operationalize the model, a web-based DSS was developed to support automated computation and user interaction. Functional and logical testing confirmed the accuracy and reliability of the system. From a practical perspective, the DSS provides an accessible and transparent tool that supports plantation managers and smallholder farmers in selecting fertilizers that balance agronomic performance, economic feasibility, and local availability. Overall, this study demonstrates the applicability of a lightweight, SMART-based DSS for improving rational fertilizer management decisions in oil palm plantations.
Stacked LSTM Integrated with Big Data Pipelines for Automated Food Beverage Stock Price Prediction Asfiani, Ilil Musyarof; Prasetya, Dwi Arman; Trimono, Trimono
bit-Tech Vol. 8 No. 3 (2026): bit-Tech
Publisher : Komunitas Dosen Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32877/bt.v8i3.3687

Abstract

Stock price volatility in the Food and Beverage (F&B) sector presents persistent challenges for investors and decision-makers, particularly in emerging markets. This study proposes an automated stock price prediction framework whose primary contribution lies in the system-level integration of a Stacked Long Short-Term Memory (LSTM) model with a scalable big data orchestration pipeline, rather than in introducing a new forecasting algorithm alone. The system targets three Indonesian F&B companies PT Indofood CBP Sukses Makmur Tbk, PT Mayora Indah Tbk, and PT Garudafood Putra Putri Jaya Tbk using historical daily stock price data. The dataset spans multiple years of trading records retrieved from the Yahoo Finance API, and predictions are generated for a seven-day forecasting horizon. Methodologically, the approach combines a multi-layer LSTM architecture with Apache Spark for distributed data preprocessing, Apache Airflow for automated workflow orchestration, and PostgreSQL for structured data storage. This integration enables scheduled data ingestion, reproducible model training, and continuous forecasting within an end-to-end analytics pipeline. Model performance is evaluated using error-based metrics, including Mean Absolute Error (MAE), Root Mean Square Error (RMSE), and Mean Absolute Percentage Error (MAPE), and is benchmarked against a conventional single-layer LSTM without pipeline orchestration. Empirical results show that the proposed pipeline-based Stacked LSTM achieves lower prediction error, with MAPE values ranging between approximately 1.1% and 2.2% across the evaluated stocks, indicating improved stability and accuracy. Overall, the findings demonstrate enhanced forecasting reliability and deployment readiness through automated pipelines.
Optimizing Raw Material Inventory for Culinary MSMEs under Data Scarcity: A DR-ARMA Forecasting Approach Lauren, Venicia; Willay, Thommy; Tjen, Jimmy
bit-Tech Vol. 8 No. 3 (2026): bit-Tech
Publisher : Komunitas Dosen Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32877/bt.v8i3.3689

Abstract

Culinary MSMEs struggle with inventory management because raw materials perish quickly and daily demand fluctuates unpredictably. Most forecasting tools require extensive historical data, often unavailable in kitchens with sparse, intermittent sales records. To address this gap, this study develops and validates a Demand Response Auto-Regressive Moving Average (DR-ARMA) model that performs reliably under severe data constraints. DR-ARMA extends classical ARMA through three stages: baseline ARIMA modeling, moving-average trend detection, and adaptive calibration that incorporates forecast errors directly into safety stock computation via an RMSE-buffered adjustment. This mechanism treats safety stock as endogenous to the forecasting workflow rather than a post hoc decision, representing the core methodological innovation. The model simultaneously enhances forecast accuracy and safety stock reliability. We validated DR-ARMA using a three-month daily sales dataset from an Indonesian culinary business, comprising 90 observations, with over 30% of days with zero sales. Results demonstrate that DR-ARMA achieves a Mean Absolute Percentage Error of 24.64%, substantially outperforming Simple Moving Average (42.70%) and marginally improving upon the Naïve benchmark (24.99%). In this zero-inflated context, even modest gains in forecast stability directly reduce spoilage and stockouts. The integrated safety stock buffer provides an empirical service level of 80%, with tighter inventory bounds that prioritize waste reduction. Finally, we embedded the model into a desktop system, converting predictions into daily procurement lists. This study confirms DR-ARMA as a practical, theoretically grounded solution for inventory optimization in data-scarce culinary settings.
Job Vacancy Recommendation System Based on Text Description Analysis Using Word Embedding and Cosine Similarity Riski, Fathur; Auliasari, Karina; Orisa, Mira
bit-Tech Vol. 8 No. 3 (2026): bit-Tech
Publisher : Komunitas Dosen Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32877/bt.v8i3.3690

Abstract

The rapid expansion of digital recruitment platforms has intensified information overload, making it increasingly difficult for job seekers to identify vacancies aligned with their skills and professional interests. In response to this challenge, this study develops a semantic-based job recommendation system that leverages word embedding and cosine similarity to enhance retrieval relevance within the Indonesian labor market context. The primary contribution lies in the empirical examination of embedding-driven semantic ranking applied to Indonesian job descriptions, with a focus on ranking coherence and contextual alignment rather than binary classification accuracy. The proposed framework transforms both user-entered skill keywords and job vacancy descriptions into dense vector representations within a shared embedding space. Semantic similarity is then computed using cosine similarity, enabling the system to rank job postings according to their contextual proximity to the user query. The recommendation output is presented in a Top-N format, prioritizing vacancies with the highest semantic correspondence. Experiments conducted on a dataset of 523 job postings demonstrate that the system consistently produces semantically coherent ranking patterns, where vacancies emphasizing relevant competencies are positioned at higher ranks. Qualitative evaluation further indicates stable ranking behavior across repeated queries, suggesting robustness in similarity-based ordering. These findings support the feasibility of embedding-based semantic retrieval as a practical and interpretable solution for content-driven job recommendation in dynamic digital recruitment environments.
IoT-Based Automatic Traffic Light Prototype Using ESP32-CAM Antarina, Selvi; Gunanto, Sigit
bit-Tech Vol. 8 No. 3 (2026): bit-Tech
Publisher : Komunitas Dosen Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32877/bt.v8i3.3698

Abstract

Traffic congestion at signalized intersections remains a persistent challenge due to the dominance of fixed-time traffic lights that cannot respond to short-term variations in vehicle demand. Although recent advances in computer vision and Internet of Things (IoT) technologies enable adaptive traffic control, many existing solutions depend on costly hardware platforms or simulation-based validation, limiting their applicability in resource-constrained contexts. This study proposes and evaluates a low-cost AIoT-based automatic traffic light prototype that integrates visual sensing, real-time vehicle detection, and adaptive signal control within an end-to-end operational framework. The system utilizes an ESP32-CAM as a vision sensor, a server-side You Only Look Once (YOLO) model for vehicle detection, and an ESP32 microcontroller for traffic light actuation, with communication implemented via WiFi using the HTTP protocol. Experimental validation is conducted under controlled prototype scenarios with traffic densities ranging from zero to three vehicles per lane to examine feasibility under hardware and network constraints. The results indicate reliable vehicle counting performance, achieving 100% accuracy for low to moderate densities and 93.3% accuracy at higher prototype density levels. Compared with a fixed-time strategy, the adaptive mechanism dynamically adjusts green light durations, reducing idle green time under low demand and increasing service time as vehicle density rises. The findings provide empirical insights into the feasibility and performance limits of low-cost vision-based adaptive traffic control systems.
Implementation of Cheng’s Fuzzy Time Series Method for Rice Price Forecasting Rosni, Rosni; Afifathuzahwa, Fauziah; Sylfia Dewi, Karina; Cahyaning Baiti, Putri Isnaini; Azzanina, Nanda
bit-Tech Vol. 8 No. 3 (2026): bit-Tech
Publisher : Komunitas Dosen Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32877/bt.v8i3.3700

Abstract

Indonesia is an agricultural country where rice is the primary staple and plays a crucial role in maintaining economic stability. However, rice price fluctuations, driven by internal and external factors, often create uncertainty for both producers and consumers. Therefore, accurate forecasting of rice prices is essential to support effective food price monitoring and policy planning. This study aims to forecast rice prices in Bandung City using Cheng’s Fuzzy Time Series (FTS) method. The novelty of this study lies in applying the Cheng FTS approach to analyze recent monthly rice price data and evaluate its forecasting performance in capturing short-term price fluctuations. The dataset consists of monthly average rice prices in Bandung City from January 2022 to June 2025, obtained from the Consumer Price Survey (SHK) published by the Badan Pusat Statistik (BPS). The modeling process involves data preprocessing, interval determination, fuzzification, construction of fuzzy logical relationships, and defuzzification to generate forecasting values. Forecasting performance is evaluated using the Mean Absolute Percentage Error (MAPE). The experimental results show that the Cheng FTS model achieved an MAPE value of 1.54%, indicating very high forecasting accuracy. The predicted rice prices closely track actual price movements, with the average forecast for the next period at Rp15,719. These findings demonstrate that the Cheng Fuzzy Time Series method delivers reliable forecasting performance and can serve as an alternative approach for predicting rice price movements. Furthermore, the proposed model may provide policymakers and related stakeholders with useful insights to support rice price monitoring and stabilization strategies in Bandung City.
Evaluating Web Application Security Using OWASP Top 10 and NIST SP 800-115 Vierino, Farrel Tiuraka; Wahanani, Henni Endah; Junaidi, Achmad
bit-Tech Vol. 8 No. 3 (2026): bit-Tech
Publisher : Komunitas Dosen Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32877/bt.v8i3.3702

Abstract

Cybersecurity assurance for public-facing government websites remains critical amid accelerating digital transformation. This study adopts an exploratory–evaluative research design to systematically examine and validate the security posture of the Surabaya Public Slaughterhouse (RPH Surabaya) website through an integrated application of OWASP Top 10 (2021) as a vulnerability taxonomy and NIST SP 800-115 as a procedural testing framework. The methodology follows structured planning, discovery, attack, and reporting phases. Discovery combined reconnaissance tools (Nslookup, Whois, Nmap, Dirsearch, Wappalyzer, and Google Dorking) with OWASP ZAP scanning, while attack validation employed Burp Suite, SQLMap, and browser-based developer analysis within a controlled Kali Linux environment. Thirteen potential vulnerabilities were detected, of which ten were empirically confirmed after manual verification. Confirmed weaknesses were predominantly categorized as Security Misconfiguration, including missing Anti-CSRF protections, directory browsing exposure, absent Content Security Policy and anti-clickjacking headers, outdated JavaScript libraries, insecure cookie attributes (missing HttpOnly and SameSite), lack of Strict-Transport-Security and X-Content-Type-Options headers, and user-controllable HTML attributes. The contribution lies in demonstrating a reproducible dual-framework validation pipeline that distinguishes scanner alerts from confirmed exploitability, thereby strengthening methodological rigor in public-sector web security assessment. These findings indicate systemic configuration-level risk exposure that may elevate susceptibility to XSS, CSRF, clickjacking, and injection-related threats relative to comparable public-institution websites. However, the assessment is limited to a single institutional website and an unauthenticated testing scope, constraining generalizability and deeper application-layer analysis.
Design and Development of an IoT-Based Rain Intensity Prediction System Using LoRa Arif, M.; Idhom, Mohammad; Wahanani, Henni Endah
bit-Tech Vol. 8 No. 3 (2026): bit-Tech
Publisher : Komunitas Dosen Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32877/bt.v8i3.3704

Abstract

An Internet of Things (IoT)–based system for rain intensity monitoring and next-day prediction is presented by integrating low-power wide-area communication using LoRa with cloud-based processing for outdoor and rural environments. This study evaluates the feasibility of LoRa communication and the end-to-end operational reliability of an IoT–cloud pipeline, while positioning machine learning as a supporting decision-aid module. A low-cost sensing node equipped with temperature, humidity, and wind-speed sensors is connected to a LoRa-based gateway that forwards measurements to an Amazon EC2 cloud server via MQTT for centralized storage, processing, and notification delivery. The system is evaluated through a 10-day single-node real-world outdoor deployment, focusing on sensor data acquisition reliability, LoRa link quality, and end-to-end operation from data acquisition to user notifications. The classification module achieves an overall accuracy of 0.74 with a weighted F1-score of 0.71, while minority-class performance remains limited due to class imbalance. LoRa communication remains stable with RSSI values of −80.91 to −79.19 dBm, SNR values of 9.86–9.95 dB, and packet loss rates below 3%. By jointly evaluating LPWAN communication reliability and cloud-side predictive services within a single field deployment, the results demonstrate the practicality of LPWAN-based IoT sensing with cloud integration for rain intensity monitoring in resource-constrained environments, while highlighting the need for future improvements in minority-class prediction and multi-node scalability.
Modeling the Open Unemployment Rate in West Java: A Comparison of Panel Data Regression Models Maulana, Mohamad Ibnu Fajar; Damaliana, Aviolla Terza; J. S., Wahyu Syaifullah
bit-Tech Vol. 8 No. 3 (2026): bit-Tech
Publisher : Komunitas Dosen Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32877/bt.v8i3.3705

Abstract

The Open Unemployment Rate (OUR) across regencies and municipalities in West Java Province reflects substantial structural heterogeneity associated with divergent local socio-economic dynamics. This study addresses the central question of which socio-economic factors systematically explain within-region variations in unemployment over time when unobserved, time-invariant regional heterogeneity is explicitly controlled. Using annual panel data for 27 regencies/cities over the period 2019–2024 (162 observations), a panel regression framework is implemented through a One-Way Fixed Effects (OWFE) model estimated via the Least Squares Dummy Variable (LSDV) approach. The explanatory variables include Regency/City Minimum Wage (UMK), Labor Force Participation Rate (TPAK), and Human Development Index (IPM). Beyond conventional fixed-effects applications, the analysis integrates a backward elimination procedure within the OWFE framework to derive a parsimonious specification; this refinement is treated as an exploratory model-selection strategy and interpreted cautiously with respect to potential sample sensitivity. Model comparison based on the Chow and Hausman tests confirms the superiority of OWFE over pooled and random specifications. The final model demonstrates substantial explanatory power (R² = 0.813) and acceptable predictive accuracy (MAPE = 10.73%), indicating that a large proportion of within-region unemployment variation is captured. Diagnostic tests show no evidence of autocorrelation (Durbin–Watson = 1.777) or heteroskedasticity under the implemented procedures. Empirically, TPAK and IPM exhibit significant negative associations with unemployment, while UMK shows a positive relationship, highlighting human capital, participation dynamics, and wage–employment trade-offs in regional labor markets.
Performance Analysis of Reasoning Models in RAG-Based Question Answering System for University Admission Services Setiawan, Muhammad Surya Adhi; Pratama, Arista; Ananto, Prasasti Karunia Farista
bit-Tech Vol. 8 No. 3 (2026): bit-Tech
Publisher : Komunitas Dosen Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32877/bt.v8i3.3707

Abstract

Access to accurate, relevant, and timely information is crucial for prospective university students; however, conventional information services often struggle with high query volumes and the risk of generative hallucinations in automated systems. This study investigates whether reasoning-oriented large language models provide measurable improvements in response quality within a Retrieval-Augmented Generation (RAG) architecture for university admission services. The study hypothesizes that internal chain-of-thought reasoning improves factual grounding compared with non-reasoning models under identical retrieval conditions. A vector-based institutional knowledge base was constructed from 30 official admission sources using VoyageAI embeddings and evaluated on a multilingual dataset of 353 real-world inquiries in Indonesian, English, and Javanese dialects. To isolate the effect of reasoning capabilities, retrieval outputs and prompt configurations were controlled across all models. Performance was evaluated using the RAGAS framework across six models categorized as reasoning (DeepSeek-R1, Gemini-2.5-Flash, o4-mini) and non-reasoning (DeepSeek-V3, Gemini-2.0-Flash, GPT-4o-mini). The results show that reasoning models achieved a higher average RAGAS score (0.7772) than non-reasoning models (0.7289), representing a 6.63% improvement, with the largest gain observed in factual correctness (+15.95%). Additional multilingual benchmarking confirmed that reasoning models maintain more stable performance across languages. Gemini-2.5-Flash achieved the highest composite score (0.8207) while maintaining favorable cost efficiency. These findings indicate that reasoning-enabled models significantly improve factual reliability in domain-specific RAG systems, although overall system performance remains strongly dependent on retrieval quality.