eJournal Komunitas Dosen Indonesia
Not a member yet
1406 research outputs found
Sort by
Forecasting Crab Raw Material Inventory in Seafood Culinary Business Using SARIMA and Prophet Methods
Get PDFSeafood based culinary businesses face significant challenges in inventory management due to fluctuating and unpredictable customer demand. Inaccurate forecasting can lead to either excess stock, resulting in waste and increased storage costs, or stock shortages, which disrupt production and reduce customer satisfaction. Among the various raw materials used, crab is the most critical, accounting for 80.6% of total raw material demand, making it essential to forecast its demand accurately to maintain operational efficiency and avoid disruptions. This study addresses the inventory problem by applying two time series forecasting methods SARIMA and Prophet to predict weekly crab demand. The performance of both models was evaluated using RMSE and MAE to assess the accuracy and reliability of their predictions over time. The SARIMA model with parameters (2,1,10)(0,1,2)[12] achieved the best forecasting performance, with an RMSE of 1.32752 and MAE of 1.24207, clearly outperforming the Prophet model, which recorded an RMSE of 1.4623 and MAE of 1.3506. These results demonstrate that SARIMA is more effective in capturing seasonal patterns and demand trends in crab usage data. In conclusion, the SARIMA model offers more precise and reliable forecasts, making it a more suitable tool for supporting raw material inventory decision-making in seafood culinary businesses, particularly when dealing with high-demand ingredients such as crab
Website-Based Sales Transaction Data Monitoring Information System
Get PDFMicro, Small, and Medium Enterprises (MSMEs) contribute significantly to the Indonesian economy, yet many still rely on manual transaction recording using Excel, which often results in errors, delays in reporting, and difficulties in accessing historical sales data. These limitations hinder effective decision-making and reduce operational efficiency. In response to these challenges, this study aims to develop a website-based sales transaction data monitoring information system that enables real-time monitoring and structured data management for various user roles: admin, reseller, and leader. The system was developed using the Rapid Application Development (RAD) method, which emphasizes fast prototyping and user feedback to ensure functionality meets actual needs. Features include automated Excel data uploads, interactive dashboards, and reseller performance visualizations tailored to user roles. Testing was conducted using Black Box Testing and User Acceptance Testing (UAT). The results indicate that all system functions operated according to design specifications and received positive feedback from users regarding usability and effectiveness. The implementation of this system successfully addresses the common problems faced by MSMEs in transaction management by improving accuracy, speeding up reporting, and enhancing monitoring transparency. In conclusion, this system offers a practical solution that is ready to be adopted by MSMEs with similar transaction recording issues, supporting digital transformation and operational efficiency in the MSME sector.Micro, Small, and Medium Enterprises (MSMEs) contribute significantly to the Indonesian economy, yet many still rely on manual transaction recording using Excel, which often results in errors, delays in reporting, and difficulties in accessing historical sales data. These limitations hinder effective decision-making and reduce operational efficiency. In response to these challenges, this study aims to develop a website-based sales transaction data monitoring information system that enables real-time monitoring and structured data management for various user roles: admin, reseller, and leader. The system was developed using the Rapid Application Development (RAD) method, which emphasizes fast prototyping and user feedback to ensure functionality meets actual needs. Features include automated Excel data uploads, interactive dashboards, and reseller performance visualizations tailored to user roles. Testing was conducted using Black Box Testing and User Acceptance Testing (UAT). The results indicate that all system functions operated according to design specifications and received positive feedback from users regarding usability and effectiveness. The implementation of this system successfully addresses the common problems faced by MSMEs in transaction management by improving accuracy, speeding up reporting, and enhancing monitoring transparency. In conclusion, this system offers a practical solution that is ready to be adopted by MSMEs with similar transaction recording issues, supporting digital transformation and operational efficiency in the MSME sector
User Satisfaction Analysis of Financial Technology Applications Using the End User Computing Satisfaction Model
Get PDFThe rapid evolution of digital technology has significantly transformed many sectors, especially financial services, with the rise of fintech innovations such as Flip. Flip offers interbank transfer services free of administrative fees, simplifying transactions like e-wallet top-ups, bill payments, and international money transfers, attracting a substantial user base. However, despite its popularity, users have raised concerns about transaction delays, slow customer service responses, and design issues, highlighting a gap between user expectations and actual service delivery. This study aims to evaluate user satisfaction with the Flip application using the End User Computing Satisfaction (EUCS) model, which assesses satisfaction across five key dimensions: Content, Accuracy, Format, Ease of Use, and Timeliness. Data from 415 respondents were analyzed using Partial Least Squares Structural Equation Modeling (PLS-SEM) to investigate the factors influencing satisfaction. The findings reveal that user satisfaction with Flip is generally positive, with significant impacts from Content, Format, Ease of Use, and Timeliness. However, Accuracy, while positively related, had no statistically significant effect. These results emphasize the need for improving transaction timeliness and refining content and interface formats to enhance the user experience. The study provides essential insights for Flip developers to enhance service quality and offers a data-driven foundation for future research on fintech applications.The rapid evolution of digital technology has significantly transformed many sectors, especially financial services, with the rise of fintech innovations such as Flip. Flip offers interbank transfer services free of administrative fees, simplifying transactions like e-wallet top-ups, bill payments, and international money transfers, attracting a substantial user base. However, despite its popularity, users have raised concerns about transaction delays, slow customer service responses, and design issues, highlighting a gap between user expectations and actual service delivery. This study aims to evaluate user satisfaction with the Flip application using the End User Computing Satisfaction (EUCS) model, which assesses satisfaction across five key dimensions: Content, Accuracy, Format, Ease of Use, and Timeliness. Data from 415 respondents were analyzed using Partial Least Squares Structural Equation Modeling (PLS-SEM) to investigate the factors influencing satisfaction. The findings reveal that user satisfaction with Flip is generally positive, with significant impacts from Content, Format, Ease of Use, and Timeliness. However, Accuracy, while positively related, had no statistically significant effect. These results emphasize the need for improving transaction timeliness and refining content and interface formats to enhance the user experience. The study provides essential insights for Flip developers to enhance service quality and offers a data-driven foundation for future research on fintech applications
Multimodal Detection of Covert Online Gambling Advertisements Using Faster R-CNN and Tr-OCR
Get PDFThe increasing prevalence of online gambling advertisements on social media has led to the use of covert strategies, such as embedding visual watermarks and employing euphemistic language, to bypass traditional detection methods, rendering manual moderation ineffective. This study proposes an AI-based automated detection system designed to identify both explicit and obfuscated gambling content. The system operates in three stages: (1) Object detection: Faster R-CNN, using a ResNet-50 backbone and Feature Pyramid Network (FPN), detects gambling-related visual elements, such as watermarks and logos; (2) Text extraction: A Transformer-based Optical Character Recognition (TrOCR) model is employed to extract textual content from images and video frames, even in the presence of visual distortions; and (3) Text classification: A BERT-based Natural Language Processing (NLP) model is used to identify gambling-related language within the extracted text. The dataset, manually collected and annotated, was augmented with Roboflow to improve model robustness and generalization. Experimental results show that the Faster R-CNN model achieved an average precision of 98.1%, TrOCR demonstrated a Character Error Rate (CER) of 4.6% and a Word Error Rate (WER) of 29%, while the BERT classifier reached an impressive 99% accuracy with high precision and recall. The system was integrated into a Flask-based web application that allows real-time analysis of both image and video inputs. This system presents strong potential to support automated content moderation and curb the spread of online gambling advertisements on digital platforms, contributing to safer online spaces.The increasing prevalence of online gambling advertisements on social media has led to the use of covert strategies, such as embedding visual watermarks and employing euphemistic language, to bypass traditional detection methods, rendering manual moderation ineffective. This study proposes an AI-based automated detection system designed to identify both explicit and obfuscated gambling content. The system operates in three stages: (1) Object detection: Faster R-CNN, using a ResNet-50 backbone and Feature Pyramid Network (FPN), detects gambling-related visual elements, such as watermarks and logos; (2) Text extraction: A Transformer-based Optical Character Recognition (TrOCR) model is employed to extract textual content from images and video frames, even in the presence of visual distortions; and (3) Text classification: A BERT-based Natural Language Processing (NLP) model is used to identify gambling-related language within the extracted text. The dataset, manually collected and annotated, was augmented with Roboflow to improve model robustness and generalization. Experimental results show that the Faster R-CNN model achieved an average precision of 98.1%, TrOCR demonstrated a Character Error Rate (CER) of 4.6% and a Word Error Rate (WER) of 29%, while the BERT classifier reached an impressive 99% accuracy with high precision and recall. The system was integrated into a Flask-based web application that allows real-time analysis of both image and video inputs. This system presents strong potential to support automated content moderation and curb the spread of online gambling advertisements on digital platforms, contributing to safer online spaces
IoT-Based Alcohol Presence Detection in Soy Sauce Using MQ-3 and ESP32
Get PDFEthanol is a compound that naturally forms during the fermentation process of products such as soy sauce, raising concerns about their halal status for Muslim consumers. This research develops an alcohol detection system based on the Internet of Things (IoT) that integrates an MQ-3 sensor, ESP32 microcontroller, and the Blynk mobile application. The MQ-3 sensor detects ethanol vapor, and the ESP32 processes the sensor data and transmits it via Wi-Fi, enabling real-time monitoring through both a 16x2 LCD display and the Blynk app. The system’s calibration process involves standard ethanol solutions with concentrations ranging from 0% to 10%. The sensor output is converted from analog-to-digital (ADC) values to voltage, parts per million (ppm), and percentage estimates. A regression analysis of the sensor data yielded the equation y = 684.59x + 3198.9, with an R2 value of 0.7288, indicating a moderate correlation between ethanol concentration and sensor readings. Using solutions with ethanol concentrations of 1% and 3%, a detection threshold of 5300 ppm was established. Testing on commercial soy sauce samples (0% and 3.08% ethanol) confirmed the system's ability to distinguish between products with and without detectable ethanol, validating its effectiveness. While not designed for precise quantitative analysis, this system offers a practical, economical, and portable solution for initial screening of alcohol in fermented food products, making it a valuable tool for halal product monitoring.Ethanol is a compound that naturally forms during the fermentation process of products such as soy sauce, raising concerns about their halal status for Muslim consumers. This research develops an alcohol detection system based on the Internet of Things (IoT) that integrates an MQ-3 sensor, ESP32 microcontroller, and the Blynk mobile application. The MQ-3 sensor detects ethanol vapor, and the ESP32 processes the sensor data and transmits it via Wi-Fi, enabling real-time monitoring through both a 16x2 LCD display and the Blynk app. The system’s calibration process involves standard ethanol solutions with concentrations ranging from 0% to 10%. The sensor output is converted from analog-to-digital (ADC) values to voltage, parts per million (ppm), and percentage estimates. A regression analysis of the sensor data yielded the equation y = 684.59x + 3198.9, with an R2 value of 0.7288, indicating a moderate correlation between ethanol concentration and sensor readings. Using solutions with ethanol concentrations of 1% and 3%, a detection threshold of 5300 ppm was established. Testing on commercial soy sauce samples (0% and 3.08% ethanol) confirmed the system's ability to distinguish between products with and without detectable ethanol, validating its effectiveness. While not designed for precise quantitative analysis, this system offers a practical, economical, and portable solution for initial screening of alcohol in fermented food products, making it a valuable tool for halal product monitoring
Pengaruh Kompensasi dan Lingkungan Kerja Terhadap Produktivitas Kerja Karyawan Pada PT Bumi Siak Pusako (PT. BSP)
Get PDF
This study aims to understand the extent to which compensation and work environment affect employee productivity at PT Bumi Siak Pusako (PT BSP). This is relevant because in the era of globalization, fierce business competition requires companies to focus more on increasing employee productivity as one of the key factors to achieve company sustainability and growth. The decline in the productivity of PT BSP employees in recent years has raised concerns, where productivity decreased from 102% in 2019 to 86% in 2023. Most likely, this is due to delays in providing compensation and inadequate working environment conditions. A quantitative approach was used by a survey method of 73 operational employees as respondents. Analysis technique with multiple linear regression. The findings found that compensation and the work environment significantly had a positive impact on work productivity. Of the two variables, compensation has a more dominant influence than the work environment. PT BSP needed to improve the compensation system by ensuring that it was timely and in accordance with the workload. Improving the quality of the work environment is also needed to create safer and more comfortable conditions for employees. Employee productivity can be increased to support the achievement of the company's targets
Pengaruh NIM, BOPO, dan LDR terhadap Return on Assets pada Bank BJB Tahun 2017-2025
Get PDFTujuan dari penelitian ini adalah untuk menguji bagaimana Return on Assets PT Bank Pembangunan Daerah Jawa Barat dan Banten Tbk (Bank BJB) dipengaruhi oleh NIM, BOPO, LDR terhadap ROA pada tahun 2017 sampai dengan triwulan pertama tahun 2025. Dengan menggunakan strategi eksplanatori berdasarkan data sekunder yang diperoleh dari laporan keuangan terbaru, Penelitian menggunakan pendekatan kuantitatif. Analisis Regresi linier berganda adalah teknik analisis yang digunakan, dan dilengkapi dengan pengujian asumsi dan hipotesis tradisional (uji-t dan uji F). Temuan ini menunjukkan bahwa LDR tidak memiliki dampak yang nyata terhadap ROA, meskipun NIM dan BOPO memiliki dampak parsial yang cukup besar. Ketiga variabel independen secara signifikan mempengaruhi ROA secara bersamaan. NIM merupakan faktor yang paling mempengaruhi ROA dari ketiganya, yang menunjukkan betapa pentingnya pengelolaan aset produktif yang efisien. Menariknya, BOPO memiliki pengaruh positif, yang mengindikasikan kondisi efisiensi relatif dimana biaya operasional yang lebih tinggi dapat dikompensasi oleh peningkatan laba operasional. Penelitian ini menyoroti betapa pentingnya pengendalian marjin bunga dan efektivitas operasional untuk meningkatkan profitabilitas bank pembangunan daerah
Pengaruh Kecerdasan Intelektual, Self Efficacy, dan Stereotip Gender Terhadap Pilihan Profesi Sebagai Konsultan Pajak Dengan Lingkungan Kerja Sebagai Variabel Moderating
Get PDFPenelitian ini bertujuan untuk menganalisis pengaruh kecerdasan intelektual, self-efficacy, dan stereotip gender terhadap pemilihan profesi sebagai konsultan pajak perempuan, dengan lingkungan kerja sebagai variabel moderasi. Metode penelitian yang digunakan adalah kuantitatif dengan pendekatan survei. Jumlah responden sebanyak 100 orang perempuan yang bekerja atau memiliki minat dalam bidang perpajakan di wilayah Tangerang dan Jakarta. Pengumpulan data dilakukan melalui kuesioner daring, dan analisis data menggunakan Partial Least Square Structural Equation Modeling (PLS-SEM) melalui perangkat lunak SmartPLS 4. Hasil penelitian menunjukkan bahwa kecerdasan intelektual dan self-efficacy berpengaruh positif dan signifikan terhadap pemilihan profesi konsultan pajak. Stereotip gender memiliki pengaruh positif namun tidak signifikan secara statistik. Lingkungan kerja tidak berperan sebagai moderator pada ketiga hubungan variabel bebas terhadap variabel dependen. Nilai R² sebesar 0,484 menunjukkan bahwa model menjelaskan 48,4% variabilitas keputusan memilih profesi sebagai konsultan pajak. Implikasi penelitian ini menggarisbawahi pentingnya pengembangan kapasitas intrapersonal perempuan dalam mendukung partisipasi profesional di bidang perpajakan, serta perlunya intervensi kelembagaan yang lebih mendalam guna menciptakan lingkungan kerja yang mendukung
Pengaruh E-WOM dan E-Trust Terhadap E-Loyalty Yang Dimediasi E-Customer Satisfaction Pada Pengguna ShopeeFood (Studi Kasus Gen Z di Kota Semarang)
Get PDFPeran E-WOM dan kepercayaan elektronik (e-trust) dalam membentuk loyalitas elektronik (e-loyalty), dengan kepuasan pelanggan elektronik (e-customer satisfaction) bertindak sebagai mediator, dieksplorasi di kalangan pengguna ShopeeFood Generasi Z di Kota Semarang. Studi ini menggunakan metodologi penelitian kuantitatif, dengan pengumpulan data melalui kuesioner yang diberikan kepada 160 pengguna aktif ShopeeFood. Data dianalisis menggunakan PLS-SEM, didukung oleh perangkat lunak SmartPLS. Temuan menunjukkan bahwa E-WOM secara signifikan meningkatkan baik kepuasan pelanggan elektronik maupun loyalitas elektronik. Demikian pula, kepercayaan elektronik memiliki dampak yang berarti pada kedua variabel tersebut. Selain itu, kepuasan pelanggan elektronik ditemukan memiliki efek yang signifikan pada loyalitas elektronik. Penemuan penting adalah bahwa pengaruh E-WOM dan kepercayaan digital terhadap loyalitas digital ditransmisikan secara tidak langsung melalui kepuasan pelanggan digital. Koefisien jalur yang kuat dan nilai R-square yang tinggi mengonfirmasi keandalan model struktural dalam menjelaskan loyalitas konsumen digital.. Dengan demikian, penelitian ini memberikan kontribusi teoritis dalam memperkuat model loyalitas pelanggan di era digital, serta implikasi praktis bagi pengembangan strategi digital marketing ShopeeFood agar lebih fokus pada peningkatan kepercayaan dan kepuasan pelanggan melalui pengelolaan ulasan online yang efektif
Pengaruh Social Media Marketing dan Customer Satisfaction Terhadap Purchase Decision Dimediasi Oleh Repurchase Intention Pada E-Commerce Shopee
Get PDFPenelitian ini difokuskan untuk mengeksplorasi sejauh mana Social Media Marketing dan Kepuasan Pelanggan memengaruhi Keputusan Pembelian konsumen, dengan Niat Pembelian Ulang (Repurchase Intention) berperan sebagai variabel mediasi. Studi ini dilakukan dalam konteks platform e-commerce Shopee, yang dikenal sebagai salah satu pemain utama dalam industri belanja daring di Indonesia. Penelitian ini menggunakan metode kuantitatif dengan pendekatan eksplanatori. Sampel berjumlah 100 responden yang merupakan mahasiswa aktif pengguna Shopee di wilayah Semarang. Pemilihan sampel dilakukan secara purposive sampling berdasarkan kriteria tertentu. Teknik pengumpulan data dilakukan melalui penyebaran kuesioner, kemudian dianalisis menggunakan metode Structural Equation Modeling (SEM) dengan bantuan software SmartPLS. Temuan penelitian mengindikasikan bahwa Social Media Marketing memberikan dampak positif dan signifikan pada Niat Pembelian Kembali dan Keputusan Pembelian. Kepuasan Pelanggan juga terbukti memberi dampak positif pada kedua variabel itu. Repurchase Intention secara signifikan menjadi mediator hubungan antara Social Media Marketing dan Customer Satisfaction dalam mempengaruhi Purchase Decision. Temuan ini menegaskan bahwa strategi pemasaran melalui media sosial yang efektif serta peningkatan kepuasan pelanggan dapat memperkuat niat pembelian ulang dan mendorong keputusan pembelian secara berkelanjutan