Journal of Novel Engineering Science and Technology

Optimization of hybrid-based Collaborative Filtering using Matrix Factorization, Feedforward Neural Network, and XGBoost

Filimantaptius Gulo — 2026-05-14

Collaborative filtering recommendation systems are widely used in digital applications; however, they still face challenges such as cold-start and first-rater problems, as well as limited accuracy due to their inability to capture complex user–item relationships. This study proposes a hybrid recommendation model that integrates Matrix Factorization, MLP-based Feedforward Neural Network (MLP) and Extreme Gradient Boosting (XGBoost). Experiments were conducted on two real-world datasets, namely MovieLens (movies) and PT XYZ (hotels), to validate the effectiveness of the proposed approach. The results indicate that the hybrid model consistently outperforms baseline methods such as SGD-based Matrix factorization, Matrix factorization +MLP, and user/item-based Collaborative filtering. Specifically, the integration of nonlinear learning through MLP and feature enhancement via XGBoost significantly improves prediction accuracy while mitigating cold-start and first-rater issues. These findings suggest that hybrid machine learning–based approaches can advance the development of more adaptive, accurate, and personalized recommendation systems.

Evaluating Friction Dampers for Seismic Protection of Non-structural Elements in Hospitals

Yusril Abdurrahman — 2026-06-28

This study evaluates the effectiveness of friction dampers in improving the seismic performance of non-structural components in hospital buildings. A five-story reinforced concrete hospital, intentionally modeled to exceed allowable drift limits, was analyzed using nonlinear time history analysis under three earthquake scenarios: BSE-1E (225-year), BSE-2E (975-year), and BSE-2N (2475-year), assuming soft soil conditions. Non-structural components were classified as drift- or acceleration-sensitive, with damage probabilities assessed using fragility curves and categorized into risk classes. Results show that friction dampers significantly reduced damage probability for acceleration-sensitive components up to 74% for cabinet contents under BSE-1E. However, drift-sensitive elements remained vulnerable, particularly in higher-intensity events, due to the building's flexible design and limited damper activation. While friction dampers improved global structural performance, their effectiveness declined with increasing seismic demand. These findings underscore the potential and limitations of friction dampers in retrofitting hospital buildings and highlight the need for careful damper sizing and consideration of alternative strategies to protect non-structural systems.

Optimizing Pelton Turbine Efficiency through Variable Flow Rates: An Experimental Study

Rahmad Hidayat Boli — 2026-07-08

This study experimentally investigates the efficiency characteristics of a laboratory-scale Pelton turbine under variable water flow rates and mechanical loading conditions. Five flow rates (0.04 m³/s, 0.035 m³/s, 0.029 m³/s, 0.024 m³/s, and 0.021 m³/s) were tested by measuring rotational speed, torque, shaft power, and turbine efficiency. The results reveal a nonlinear relationship between load and efficiency for all tested flow rates, with maximum efficiency occurring at intermediate loading conditions. The highest efficiency of 52.727% was obtained at a flow rate of 0.04 m³/s and a load of 12 kg. Were associated with reduced efficiency, which is commonly linked to decreased jet momentum. These findings provide experimental insight into Pelton turbine performance under non-design operating conditions at laboratory scale.

Data-Driven Marketing Management Competencies: Strategies for Performance Optimization in the Big Data Era

Safitri Nurhidayati — 2026-07-11

The proliferation of big data technology has fundamentally changed the paradigm of marketing management, requiring a new competency framework for optimal performance outcomes. Despite extensive technological advances, significant gaps remain in understanding how data analytic competencies translate into measurable improvements in marketing performance. This study investigates the relationship between data-analytic-based marketing management competencies and performance optimization strategies in contemporary business environments, with particular emphasis on identifying critical competency dimensions and their impact on organizational marketing effectiveness. A mixed-methods approach was used, combining quantitative analysis of 847 marketing professionals from 156 Indonesian companies with qualitative interviews of senior marketing executives. Data collection employed validated instruments measuring analytic competencies, technology adoption, and performance metrics in Q2–Q4 2024. Findings revealed four critical competency dimensions: technical analytic proficiency (β=0.43, p<0.001), strategic data interpretation (β=0.38, p<0.001), cross-functional collaboration (β=0.32, p<0.01), and ethical data governance (β=0.28, p<0.01). Organizations with high analytic competency scores reported 34% higher marketing ROI and 28% greater efficiency in customer acquisition compared to low-competency counterparts. Data analytic competencies significantly influence marketing performance outcomes, with technical proficiency and strategic interpretation serving as primary drivers. This study provides empirical evidence supporting the adoption of competency-based frameworks in marketing management practice.

From the Laboratory to the Market: A Review of the Research Commercialisation Ecosystem in Nigeria

Oluremi Nurudeen Olaleye — 2026-07-30

Nigeria’s 200-plus universities generate substantial research across agriculture, healthcare, and engineering, yet output rarely translates into commercial products. This policy and literature review synthesises evidence from 47 government documents, peer-reviewed articles, and international reports to diagnose the research commercialisation landscape through National Innovation Systems and Triple Helix lenses. Thematic analysis reveals chronic underfunding (0.22% of GDP on R&D), weak university–industry linkages, absence of professional Technology Transfer Offices, and regulatory ambiguity as systemic barriers. Comparative analysis of South Korea, Finland, Malaysia, Rwanda, and Kenya identifies transferable institutional mechanisms. The review proposes a hybrid model combining strategic state direction in priority sectors with bottom-up entrepreneurial support, offering Nigeria-specific implementation pathways including phased R&D investment, sector-focused intermediary agencies, and diaspora engagement mechanisms. Limitations and critiques of commercialisation-first approaches are discussed. Strategic recommendations target policymakers, university administrators, and development partners.

Open Banking API and Employee Competency: Driving Remittance Performance and Customer Growth in Islamic Banking

Arina Al-Haque — 2026-07-30

This paper analyzes the impact of employee competency and business performance on the effectiveness of remittance services, with Open Banking API implementation acting as a supporting technological enabler. The study focuses on an Islamic bank undergoing digital transformation. A quantitative approach was employed using questionnaire data from 40 remittance employees, analyzed through regression methods, and complemented by a before-and-after system performance evaluation based on operational data from 2022 to 2024. The measurement instruments demonstrated excellent reliability (Cronbach’s α > 0.96) and validity (p < 0.05). The statistical results indicate that employee competency has a significant positive effect on remittance business performance. Concurrently, the implementation of the Open Banking API significantly reduced average transaction processing time by 80% (from 10 seconds to 2 seconds), which directly contributed to a 52.9% growth in active partners. Grounded in the Resource Based View (RBV) theory, these findings highlight that orchestrating complex API integrations requires parallel human resource readiness to maximize organizational performance and active partner growth.

Integrated Prediction Model for Normal and Recycled Aggregate Concrete Strength Using Ensemble Learning Techniques

Sujiat — 2026-07-30

Recycled aggregate concrete (RAC) is a sustainable alternative construction material to reduce natural resource exploitation and manage construction and demolition waste. However, predicting the mechanical performance of RAC remains a challenge due to the high variability of recycled aggregate properties. The purpose of this study is to develop a machine learning model to predict the compressive strength of recycled aggregate-based concrete and compare its performance with normal concrete. The dataset used consists of 2165 samples (1600 normal concrete and 565 recycled aggregate concrete) collected from various scientific publications. Three tree-based machine learning algorithms (Random Forest, XGBoost, and LightGBM) were implemented and optimized using RandomizedSearchCV with 5-fold cross-validation. The results showed that LightGBM provided the best performance with R² = 0.92, MAE = 2.45 MPa, and RMSE = 3.52 MPa on the test set. This model is able to predict the compressive strength of normal concrete (R² = 0.92) and recycled aggregate concrete (R² = 0.91) with almost the same accuracy, indicating strong generalization. Feature importance analysis revealed that curing age, cement content, and water content are the most important factors in compressive strength prediction, while for RAC, recycled aggregate water absorption (WRCA) also makes a significant contribution. Error analysis shows that residuals are random and normally distributed without systematic bias. This model can reliably predict concrete compressive strength in the range of 20-60 MPa with an average error of ±3-4 MPa and can be integrated into mix proportioning design software to improve the efficiency of the design process and support the use of sustainable construction materials.

Cosmetic Packaging Quality Analysis Using the Six Sigma and House of Quality Methods at PT XYZ

Roland Y.H. Silitonga — 2026-07-31

Product quality is a critical factor in the success of manufacturing industries, particularly in the cosmetic packaging sector, where packaging serves both protective and marketing functions. PT XYZ experienced a defect rate of 4.09% in cosmetic packaging products, resulting in increased production costs and reduced customer satisfaction. This study aims to reduce product defects by applying the Six Sigma DMAIC (Define, Measure, Analyze, Improve, Control) methodology integrated with Failure Mode and Effects Analysis (FMEA) and the House of Quality (HoQ). DMAIC was employed to identify defect sources, measure process performance, analyze root causes, implement improvements, and establish process control. FMEA was used to evaluate potential failure modes and prioritize corrective actions based on Risk Priority Numbers, while HoQ translated improvement priorities into practical technical actions. Data were collected through direct observation and interviews with relevant personnel at PT XYZ. The implementation of the proposed improvements reduced overall defects by 2,321.43 defects per million opportunities (DPMO), equivalent to 2.32%, and increased the process sigma level by 0.21. Specifically, the “Dirty Glass/Grepes” defect decreased by 11,616.07 DPMO (1.16%) with a sigma improvement of 0.58, whereas the “Scratch” defect declined by 5,660.71 DPMO (0.57%) with a sigma improvement of 0.25. Root cause analysis revealed that Dirty Glass/Grepes defects were primarily associated with human and method factors, while Scratch defects were related to human, equipment, and method factors. Seven improvement alternatives were identified, with two successfully implemented: adding a blower stage during final assembly and inspecting the cover pull-force testing instrument.

Development of a Visual Attention Game Using Difference Detection Mechanics: Design and Usability Evaluation

Bambang Robi'in — 2026-08-06

The increasing use of mobile devices has led to the widespread adoption of digital games as interactive systems. However, most mobile games are primarily designed for entertainment and rarely incorporate structured interaction mechanisms to support cognitive training. This study presents the design and development of a visual attention-based mobile game that uses difference-detection mechanics as its core interaction model. The game, named FindX, was developed using the Game Development Life Cycle (GDLC) framework, including initiation, pre-production, production, and testing stages, and implemented using the Unity engine on the Android platform. The system incorporates structured gameplay elements, including progressive difficulty levels, time constraints, and penalty mechanisms for incorrect interactions, to encourage focused visual processing. Functional testing was conducted using a black-box approach to ensure system reliability. Expert evaluation using the Learning Object Review Instrument (LORI) was performed to assess media quality and content suitability. In addition, usability testing was conducted with 23 participants using the System Usability Scale (SUS). The results indicate that the system achieved a media quality score of 80% and a content feasibility score of 93.3%. The usability evaluation produced an average SUS score of 74.1, categorized as Good and Acceptable. These findings indicate that the proposed system is functional, of good media and content quality, and user-friendly, providing a usable interactive environment designed to support visual attention training through structured game mechanics. Efficacy in improving visual attention was not measured in this study and is identified as future work.

Temperature-Stable 437.5 MHz Current-Starved Ring Oscillator with Bandgap Reference for Nanosatellite Beacon Transmitters

Thit Waso Khine — 2026-08-07

This paper presents a 437.5 MHz 3-stage CMOS current-starved ring oscillator (CSRO) with bandgap reference (BGR) design including process, voltage, and temperature (PVT) analysis. The circuit was designed and simulated in LTspice software using BSIM3 models, Level 8, 180 nm technology. The main objective is to design a simple, low-power, low-cost, high-temperature-stable (437 MHz) current-starved CMOS ring oscillator for a nanosatellite Morse code beacon transmitter. For practical implementation, the circuit was designed with standard E12 series resistor values, resulting in a final frequency of 437.5 MHz. The oscillator consumes 0.669 mW from a 1.8 V supply voltage, achieving phase noise of 95.7 dBc/Hz at 1 MHz offset and a figure-of-merit (FOM) of 150.27. The CSRO exhibits a frequency variation of 1.31% over the temperature range of -40℃ to 125℃, demonstrating excellent robustness. A supply voltage variation of 5% results in 7.02 % of frequency change. In this work, a bandgap reference circuit (BGR) was separately designed and simulated, which gave the temperature coefficient of 10.68 ppm/ across the temperature of 40 to 125 and line regulation of 0.18% across the supply voltage variation of 10%. The design includes PVT analysis across worst-case corners, temperature extremes, and ±5% supply voltage variation on operating frequency and power consumption to ensure results closely match measurements.