Proceeding of International Conference on Advanced Technologies, Innovations and Science

The Effect of Liquidity, Leverage, and Board Gender Diversity on Market Performance

2026-06-03T10:01:42+00:00

This study aims to analyze the effect of liquidity, leverage, and board gender diversity on stock return as the proxy for market performance of companies included in the Kompas100 Index on the Indonesia Stock Exchange for the 2020–2024 period. A quantitative causal-associative approach was employed with a sample of 26 companies (130 observations) determined through purposive sampling. Data were analyzed using ordinary least squares regression with IBM SPSS version 26. After natural log transformation and removal of eight outlier observations (final N = 122), all classical assumption requirements were satisfied. The three independent variables constitute a statistically feasible regression model (F-test sig. = 0.019) with an Adjusted R² of 5.7 percent, indicating that stock return of large-capitalization Kompas100 companies is predominantly driven by macro-level and market-wide factors that fall outside the scope of this firm-specific model. Partially, liquidity has a positive and significant effect on stock return a direction contrary to the initial hypothesis, attributable to the dominant positive signaling role of financial health among blue-chip investors; board gender diversity has a negative and significant effect, aligned with the hypothesis; and leverage has no significant effect, consistent with the efficient market pricing of publicly disclosed debt information. Future research is recommended to incorporate macroeconomic variables, expand the sample scope, and apply panel data methods.

The Role of Online Customer Reviews and Ratings in Improving Shopee Consumer Purchasing Decisions in Jombang Regency

2026-06-10T04:39:07+00:00

This study examines the role of online customer reviews and ratings in improving consumer purchasing decisions on Shopee in Jombang Regency. In an increasingly competitive e-commerce environment, consumers often rely on reviews and ratings as a primary source of information to reduce uncertainty and assess product quality before making a purchase. This study used an exploratory quantitative approach with a survey method. Data were collected from Shopee users in Jombang Regency who had experience reading reviews and ratings before purchasing a product. The sample consisted of 400 respondents selected through purposive sampling, and the data were analyzed using multiple linear regression, classical assumption tests, partial t-tests, and determinant tests. The results show that online customer reviews and online customer ratings have a positive and significant influence on purchasing decisions. Of the two variables, online customer ratings have a stronger influence than reviews, indicating that consumers tend to use star ratings as a quick indicator of product quality. These findings suggest that sellers should maintain product quality, encourage authentic customer feedback, and manage review credibility to strengthen consumer trust and support purchasing decisions.

Integrating Elsa Speak Application into English Language Learning to Enhance Students' Pronunciation Skills at SD IT Aulia Kids School

2026-06-05T14:14:32+00:00

The integration of Artificial Intelligence (AI)-based technology in English language learning offers innovative ways to improve students' pronunciation skills. This study aims to investigate the effectiveness of the Elsa Speak application in enhancing students’ English pronunciation at SD IT Aulia Kids School. A quantitative approach with a pre-experimental design was employed involving elementary school students. Data were collected through pre-test and post-test pronunciation assessments as well as classroom observations. The findings revealed that the use of Elsa Speak significantly improved students’ pronunciation accuracy, word stress, and speaking confidence. The application also increased students’ motivation and participation in English learning activities due to its interactive features and instant feedback. The study concludes that Elsa Speak is an effective AI-based learning tool for improving students’ English pronunciation skills and supporting more engaging English language learning.

Role Work-Life Balance in Psychological Well Being Through Human Resources

2026-05-14T06:28:01+00:00

This study aims to analyze the effect of work-life balance on psychological well-being through psychological resources using the Partial Least Squares Structural Equation Modeling (PLS-SEM) approach. The study examines work-life balance as the independent variable, psychological resources as the mediating variable, and psychological well-being as the dependent variable. The findings indicate that work-life balance has a positive and significant effect on psychological well-being with a path coefficient of 0.453. Furthermore, work-life balance positively affects psychological resources with a coefficient of 0.484. Psychological resources also positively influence psychological well-being with a coefficient of 0.345. The R-square value of psychological well-being is 0.471, indicating that 47.1% of the variance is explained by the model. These results confirm that psychological resources partially mediate the relationship between work-life balance and psychological well-being. This study highlights the importance of organizational policies that support work-life balance in improving employees’ psychological well-being.

The Effect of Free Cash Flow and Profitability on Food and Beverage Investment Decisions

2026-06-10T15:04:40+00:00

This study examines the effect of free cash flow and profitability on investment decisions among food and beverage manufacturing companies listed on the Indonesia Stock Exchange during the 2022-2024 period. Previous studies have primarily focused on the influence of free cash flow and profitability on firm value, dividend policy, and financial performance, while limited attention has been given to investment decisions within the Indonesian food and beverage sector. This study addresses this gap by investigating how internal financial resources influence corporate investment behavior in a sector characterized by high capital requirements and continuous growth opportunities. Using a quantitative approach, secondary data were collected from annual financial reports of 64 companies, resulting in 192 firm-year observations selected through purposive sampling. Multiple linear regression analysis was employed to test the proposed hypotheses. The results indicate that free cash flow and profitability have a positive and significant effect on investment decisions. Profitability demonstrates a stronger influence compared to free cash flow, suggesting that firms with higher earnings capacity are more likely to undertake investment activities. These findings support Pecking Order Theory and Signaling Theory by confirming the importance of internal financial resources in corporate investment decisions. This study contributes to the literature by providing recent empirical evidence from Indonesia’s food and beverage manufacturing industry during the post-pandemic recovery period.

Analysis of Imbalanced Data in Heart Disease Classification Using SVM and SMOTE

2026-06-24T17:33:39+00:00

This study aims to examine the effect of applying the Synthetic Minority Over-sampling Technique (SMOTE) on the performance of the Support Vector Machine (SVM) algorithm in classifying heart disease in a dataset with class imbalance. The dataset used is derived from the 2015 BRFSS Heart Disease Health Indicators, where the number of non-patients far outweighs that of heart disease patients. This condition has the potential to reduce the model’s ability to detect the minority class. Therefore, SMOTE was applied as a preprocessing step to balance the training data distribution before the classification process using SVM. Model evaluation was conducted using the accuracy, precision, recall, F1-score, and ROC-AUC metrics. The results of the study indicate that the application of SMOTE is capable of improving the performance of the SVM model in detecting heart disease patients. The accuracy score increased from 0.77 to 0.78, recall from 0.78 to 0.83, F1-score from 0.75 to 0.76, and ROC-AUC from 0.84 to 0.85. Although there was a slight decrease in precision, the model’s improved ability to recognize the minority class indicates that SMOTE is effective in addressing data imbalance issues. These results demonstrate that the combination of SVM and SMOTE can serve as a reliable approach for classifying heart disease in imbalanced datasets.

Classification of Smoking Dependency Levels Using the K-NN Algorithm

2026-06-24T17:40:05+00:00

Smoking is a major public health issue due to the addictive effects of nicotine, which can lead to varying levels of dependence among individuals. Identifying smoking dependence levels is important to support prevention and intervention efforts. This study aims to classify smoking dependence levels using the K-Nearest Neighbor (K-NN) algorithm. The dataset used in this research was obtained from Kaggle and consists of 103 respondent records with attributes including age, gender, smoking peers, smoking family members, age of smoking initiation, number of cigarettes consumed per day, and daily allowance. The research process involved data preprocessing, including data cleaning, categorical data transformation, and Min-Max normalization, followed by data splitting into 80% training data and 20% testing data. Classification was performed using the Euclidean Distance metric, while model performance was evaluated using accuracy, precision, recall, and F1-score. Experimental results show that the K-NN algorithm with K = 5 achieved an accuracy of 95.24%, precision of 95.71%, recall of 95.24%, and F1-score of 95.10%. Although K values of 1, 2, and 3 produced higher accuracy, K = 5 was selected because it provided better stability and generalization while reducing sensitivity to noise and outliers. The findings demonstrate that the K-NN algorithm is effective for classifying smoking dependence levels and can be utilized as a decision-support tool in smoking behavior analysis and public health initiatives.

Classification of Construction Worker Productivity Levels Using the C4.5 Algorithm

2026-06-27T05:25:14+00:00

Construction worker productivity is a critical factor determining time efficiency, cost, and quality in construction project implementation. Productivity assessments are often conducted subjectively; therefore, a data-driven approach is needed to produce more objective and measurable evaluations. This study aims to classify the productivity levels of construction workers on a bathroom renovation project at MBK Tanjung Pinggir using the C4.5 algorithm. The dataset consists of 20 construction workers with three primary evaluation attributes: working hours, attendance, and skill level. Through entropy and information gain calculations, the skill attribute was identified as the most dominant factor with the highest gain value, thus becoming the root node in the formation of the decision tree. The model results produced a series of classification rules that describe the relationship between attributes and worker productivity levels, and indicate a tendency that technical skills have the greatest influence compared to other attributes. The implementation of the C4.5 algorithm proved effective in generating a classification model that is easy to understand and can support more objective project management decision-making. These findings are expected to serve as a foundation for human resource management in construction projects and encourage further research with a larger data scope.

Classification of Social Assistance Recipients Eligibility Using the K-NN Method

2026-06-27T05:31:08+00:00

The classification of the eligibility of social assistance recipients (bansos) was carried out in this study by applying the K-Nearest Neighbor (K-NN) method to support the distribution of assistance that is more targeted, objective, and accurate. The problem behind the research stems from the fact that there is still an element of subjectivity in determining aid recipients, so that the distribution of social assistance has not fully reached the people who need it most. A total of 20 data on prospective recipients of social assistance from Karang Keri Village, Simalungun Regency, were used as research datasets. The variables analyzed included employment, number of dependents, income, and housing conditions, while the status of aid recipients was determined as the target class. Before the classification process is carried out, the data is prepared through a preprocessing stage which includes data selection, data cleaning, encoding, and normalization. Furthermore, the dataset is divided into 70% training data and 30% testing data. The classification process was carried out using the K-NN algorithm with a value of K = 5 and Euclidean Distance calculation. The model's performance was then evaluated using accuracy, precision, recall, F1-score, cross validation, and confusion matrix. The test results showed that the model was able to achieve 83.3% accuracy, 100.0% precision, 75.0% recall, 85.7% F1-score, and 80.0% cross-validation. These findings indicate that the K-NN method has good classification ability and stable performance, so it is suitable for use as a decision-making support in determining social assistance recipients more objectively and on target.

Data Mining Classification to Identify Stunting in Toddlers Using the Naïve Bayes and KNN Algorithms

2026-06-27T05:37:45+00:00

Stunting is a chronic nutritional problem that can hinder physical growth, cognitive development, and the quality of human resources in the future. The high risk of stunting among toddlers highlights the need for methods that can support rapid and accurate identification. This study aims to apply and compare the Naïve Bayes and K-Nearest Neighbor (K-NN) algorithms in classifying the nutritional status of toddlers using a data mining approach. The dataset was obtained from the Pondok Tengah Community Health Center in Pematangsiantar City and consists of toddlers' height and weight data as the main attributes. The research stages included data collection, preprocessing, classification modeling, and model evaluation. The evaluation process was carried out using a confusion matrix and accuracy measurement. The results indicate that both algorithms can effectively classify the nutritional status of toddlers. However, the K-Nearest Neighbor (K-NN) algorithm achieved better performance with an accuracy of 91.67%, while the Naïve Bayes algorithm achieved an accuracy of 83.33%. These findings demonstrate that K-NN is more effective in identifying nutritional status patterns within the dataset used in this study. The implementation of data mining techniques is expected to assist healthcare workers in the early identification of toddlers at risk of stunting, enabling faster and more accurate preventive and intervention measures.

Comparison of the Performance Fuzzy Mamdani, Sugeno, and Tsukamoto in Weather Data Modeling

2026-06-27T06:54:42+00:00

Weather forecasting is an essential component in various sectors, including agriculture, transportation, disaster mitigation, and environmental management, as it supports more effective planning and decision-making processes. The dynamic, nonlinear, and uncertain nature of weather data makes weather modeling a complex task. One of the widely adopted approaches for handling such uncertainty is the Fuzzy Inference System (FIS), which is capable of representing expert knowledge through linguistic rules and membership functions. This study aims to compare the performance of the Fuzzy Mamdani, Fuzzy Sugeno, and Fuzzy Tsukamoto methods in weather data modeling using an identical experimental configuration. The dataset consists of 96,453 historical weather records, with Temperature, Humidity, Wind Speed, and Pressure serving as input variables and Weather Condition as the output variable. The research process includes data preprocessing, membership function design, rule base construction, fuzzy model implementation, and performance evaluation. A total of 56 identical fuzzy rules were applied to all methods to ensure an objective comparison. Performance evaluation was conducted using Accuracy, Precision, Recall, F1-Score, Mean Absolute Error (MAE), and Root Mean Square Error (RMSE). The results indicate that the Mamdani method achieved the best performance, with an Accuracy of 81.04%, Precision of 65.68%, Recall of 81.04%, F1-Score of 72.55%, MAE of 0.264113, and RMSE of 0.642724. These findings demonstrate that all three fuzzy methods are effective for weather data modeling; however, the Mamdani method provides slightly superior performance compared to the Sugeno and Tsukamoto methods.

Time Management, Work-Life Balance, and Mental Well-Being: a Quantitative Study Using SmartPLS

2026-05-07T16:35:10+00:00

This study analyzes the effect of time management on mental well-being through work-life balance as a mediating variable amidst the increasing pressures of modern life. This study employed a quantitative approach, utilizing survey data collected from 111 respondents, consisting of students and workers experiencing academic and work stress. The data were analyzed using Structural Equation Modeling-Partial Least Squares (SEM-PLS) with SmartPLS software to test the direct relationship between the variables. The results indicate that time management has a positive and significant effect on work-life balance (? = 0.371; T-statistic = 3.111; p = 0.002) and mental well-being (? = 0.634; T-statistic = 9.799; p = 0.000). Furthermore, work-life balance also significantly impacted mental well-being (? = 0.538; T statistic = 5.807; p = 0.000). These findings suggest that effective time management contributes to increased psychological stability, reduced stress levels, and a better balance between personal and professional life. Furthermore, work-life balance acts as an important mediating factor in strengthening mental well-being. This study highlights the importance of developing time management skills as a strategic approach to maintaining mental health and achieving a sustainable life balance in the modern era.

The Role of Artificial Intelligence in Accounting on Financial Reporting and Audit Quality

2026-06-10T08:42:01+00:00

The rapid development of Artificial Intelligence (AI) has transformed accounting and auditing practices by enabling automated data processing, real-time analysis, and enhanced decision-making capabilities. However, previous studies have generally examined the impact of AI separately on financial reporting quality or audit quality, resulting in limited integrated understanding of its role in both areas. Therefore, this study aims to analyze the role of Artificial Intelligence in improving financial reporting quality and audit quality within accounting practices. This research employed a qualitative approach using the Systematic Literature Review (SLR) method. The study analyzed 20 relevant national and international scientific articles published between 2020 and 2025. Data were collected through a systematic literature selection process and analyzed using descriptive qualitative techniques. The findings indicate that AI improves financial reporting quality through transaction automation, efficient data processing, and the generation of more accurate, relevant, and timely financial information. In auditing, AI enhances audit effectiveness through big data analytics, anomaly detection, risk assessment, and fraud detection capabilities. Nevertheless, AI implementation continues to face challenges related to technological infrastructure, human resource competencies, implementation costs, and trust in technology. Overall, AI plays a strategic role in supporting the digital transformation of accounting and auditing practices while improving financial reporting quality and audit quality.

The Effect of Self-Efficacy, Perceived Organizational Support and Employee Engagement on Organizational Citizenship Behavior

2026-05-31T11:05:13+00:00

Success organization No only determined by implementation formal duties of employees , but also greatly influenced by the emergence of behavior voluntary or organizational citizenship behavior ( OCB ). Low level initiative as well as contribution extra-role employee indicates existence need for strengthen factors psychological and organizational capabilities push growth OCB. Research This aim for analyze influence self- efficacy, perceived organizational support, and employees engagement to organizational citizenship behavior of employees of PT Pegadaian Jimbaran Branch. Approach research used is quantitative with method survey, which involves all over PT Pegadaian Jimbaran Branch employees totaling 42 people through technique census. Research data collected use questionnaire Likert scale and analyzed with method multiple linear regression. Testing hypothesis done through the t test to find out influence partial each variable independent, as well as the F test for test influence simulants. Research results show that self- efficacy, perceived organizational support, and employees engagement influential positive and significant to organizational citizenship behavior, good in a way partial and simultaneous.

The Influence of Overthinking and Insecurity on Well-Being in Generation Z in Indonesia

2026-05-09T09:41:53+00:00

Overthinking and insecurity are common psychological issues experienced by Generation Z in the digital era, particularly due to the influence of social media and social comparison. This study aims to analyze the effect of overthinking and insecurity on well-being among Generation Z in Indonesia. This research uses a quantitative approach with a survey method involving 102 respondents aged 18–25 years. Data were collected through online questionnaires and analyzed using Partial Least Squares - Structural Equation Modeling (PLS-SEM) techniques. The results show that overthinking has a significant negative effect on well-being, as it increases anxiety and reduces life satisfaction. Conversely, insecurity does not significantly affect well-being within this sample. Overall, the independent variables simultaneously contribute to explaining a minor portion of the variance in psychological well-being, heavily dominated by the cognitive loop of overthinking. These findings highlight the importance of targeted mental health awareness and effective cognitive coping strategies for Generation Z in facing digital pressure.

Shifting Preferences for Digital Payments Among SME Owners: From QRIS to Mobile Banking Transfers

2026-06-10T08:43:29+00:00

The rapid expansion of digital financial services has transformed payment behavior among Micro, Small, and Medium Enterprises (MSMEs) in Indonesia. While the Quick Response Code Indonesian Standard (QRIS) has been widely promoted as a standardized digital payment solution, recent developments suggest a growing preference for mobile banking transfers among MSME operators. This study investigates the shift in digital payment preferences from QRIS to mobile banking transfers among tourism-oriented MSMEs in Aceh, Indonesia. The research was conducted at three major tourism destinations: the Aceh Culinary Tourism Cake Center in Gampong Lampisang, Aceh Besar; cultural souvenir businesses surrounding the Tsunami PLTD Apung tourism area in Banda Aceh; and marine tourism businesses on Rubiah Island, Sabang. Using a qualitative research approach, data were collected through in-depth interviews, direct observations, and document analysis involving 40 MSME owners and employees directly engaged in business transactions. The data were analyzed using thematic analysis supported by data triangulation and member checking to ensure credibility. The findings reveal that approximately 65% of participants prefer mobile banking transfers over QRIS due to perceived convenience, transaction efficiency, ease of financial monitoring, and stronger trust in banking applications. The results indicate that digital payment adoption among MSMEs is a dynamic process influenced by technological, behavioral, and institutional factors. These findings contribute to a better understanding of evolving digital payment behavior and provide practical insights for policymakers, financial institutions, and digital payment providers seeking to strengthen MSME digitalization and financial inclusion.

Comparative Analysis and Optimization of Reactor Geometry and Voltage Density in Electrocoagulation for TDS, TSS, and COD Reduction from Pharmaceutical Wastewater

2026-06-09T14:07:40+00:00

Pharmaceutical industrial wastewater typically contains high levels of organic and inorganic pollutants, such as TDS, TSS, and COD, that can harm ecosystems if discharged without treatment. This study examines the effectiveness of electrocoagulation using a pair of iron electrodes (Fe-Fe) in a laboratory-scale batch system for reducing these parameters. To improve process efficiency, this study compared two reconfigured reactor geometries: Design I (40 x 20 x 15 cm) and Design II (16 x 16 x 20 cm) by adjusting raw voltages (3–13 V) to evaluate the impact of applied voltage density over contact times of 90–240 minutes. Quantitative evaluation showed that Design II demonstrated a significant improvement in pollutant removal efficiency compared to Design I due to its compact geometry, which minimized dead zones and optimized the contact area. Specifically, Design I achieved a maximum removal of 66.67% for TSS and 73.50% for TDS at a lower voltage threshold of 9 V before fluctuating due to rapid saturation. In contrast, Design II consistently converted energy, achieving peak efficiencies of 76.05% for TSS (at 12 V) and 79.89% for TDS (at 13 V). Furthermore, for organic degradation, Design II outperformed Design I, achieving 84.95% COD removal (at 12 V) compared to Design I’s peak of only 55.83% (at 13 V). This electrocoagulation process, utilizing the optimized Design II configuration, was very effective in reducing TDS and TSS levels to meet the quality standards stipulated in the Minister of Environment Regulation No. 5 of 2014. However, COD reduction still requires an advanced treatment unit.

The Influence of Career Adaptability and Personal Branding on the Career Readiness of Generation Z Students in South Tangerang

2026-05-09T16:11:40+00:00

Career readiness has become one of the most critical competencies that Generation Z students must develop to successfully navigate an increasingly competitive and dynamic job market. The rapid transformation of the work environment driven by digitalization and automation demands that students possess strong career adaptability and the ability to strategically build a professional identity through personal branding. This study aims to analyze the simultaneous and partial influence of career adaptability and personal branding on the career readiness of Generation Z university students in South Tangerang, Indonesia. This research employed a quantitative approach using Partial Least Squares Structural Equation Modeling (PLS-SEM) with SmartPLS 3.2.9. Data were collected through online questionnaires using a five-point Likert scale (1 = Strongly Disagree to 5 = Strongly Agree), distributed to 103 active university students aged 18–25 years via purposive sampling. The measurement model demonstrated adequate validity and reliability: AVE values ranged from 0.612 to 0.667, Composite Reliability from 0.876 to 0.906, and Cronbach’s Alpha from 0.834 to 0.873. The structural model yielded an R-Square value of 0.486, indicating that both variables explain 48.6% of the variance in career readiness. Career adaptability had a positive and significant effect on career readiness (? = 0.427, T-statistic = 4.112, p = 0.000). Personal branding also had a positive and significant effect on career readiness (? = 0.381, T-statistic = 3.780, p = 0.000). These findings suggest that higher education institutions should integrate career adaptability training and personal branding programs into their student development curricula to enhance graduates’ workforce readiness.

The Implementation of Quick Response Code Indonesian Standard (QRIS) in Mosque Financial Management

2026-06-07T14:45:44+00:00

This study aims to analyze the implementation of the Quick Response Code Indonesian Standard (QRIS) in mosque financial management at Sabilil Jannah Mosque Banda Aceh. The development of digital payment technology has encouraged religious institutions, including mosques, to adopt cashless transaction systems in collecting donations, infaq, and sadaqah from congregants. QRIS is considered an alternative payment method that provides convenience, efficiency, and transparency in financial transactions. This research uses a qualitative approach with a case study method. Data were collected through interviews, observation, and documentation involving mosque administrators and congregants of Sabilil Jannah Mosque Banda Aceh. The data were analyzed through data reduction, data presentation, and conclusion drawing. The results of this study indicate that the implementation of QRIS helps facilitate congregants in making digital donations without using cash. In addition, QRIS supports mosque administrators in recording transactions more efficiently and improving transparency and accountability in financial management. However, the use of QRIS still faces several challenges, such as limited digital literacy among some congregants, the habit of using cash, and the need for continuous socialization from mosque administrators. This study concludes that QRIS has an important role in supporting the modernization of mosque financial management. Therefore, mosque administrators are expected to improve education and socialization regarding the use of QRIS so that its utilization can be more optimal.

The Influence of Fear of Missing Out (FOMO), Mindfulness, and Work-Life Integration on the Psychological Well-Being of Generation Z in Jakarta

2026-05-05T05:38:30+00:00

This study aims to analyze the influence of Fear of Missing Out (FOMO), mindfulness, and work-life integration on the psychological well-being of Generation Z in Jakarta. The study used a quantitative approach with a survey method of 102 Generation Z respondents who actively use social media for at least 3 hours per day. Data analysis was performed using Structural Equation Modeling–Partial Least Squares (SEM-PLS) with the assistance of SmartPLS software. The results showed that Fear of Missing Out (FOMO) had a negative and significant effect on psychological well-being with a path coefficient of ? = -0.341, t-statistic = 3.072, and p-value = 0.002. Meanwhile, mindfulness had a positive and significant effect on psychological well-being with a path coefficient of ? = 0.452, t-statistic = 3.538, and p-value = 0.000. Work-life integration also had a positive and significant effect on psychological well-being, with a path coefficient of ? = 0.287, t-statistic = 4.957, and p-value = 0.000. An R-squared value of 0.684 indicates that FOMO, mindfulness, and work-life integration explain 68.4% of the variation in psychological well-being among Generation Z in Jakarta. This research is novel in that it integrates digital psychology and work-life balance factors to explain the psychological well-being of Generation Z in an urban context with high social media usage. The results are expected to provide theoretical contributions to the development of digital psychology studies and serve as a practical reference for improving the mental health of Generation Z in the digital era.

Workplace Ostracism and Presenteeism: The Mediating Role of Psychological Resource Depletion among Indonesian Service-Sector Employees

2026-05-25T14:19:49+00:00

Presenteeism has emerged as a hidden productivity problem in organizations because employees remain physically present at work while experiencing reduced psychological engagement and work performance. Although previous studies have identified workplace ostracism as a predictor of negative employee outcomes, limited research explains the psychological mechanism linking workplace ostracism and presenteeism, particularly in collectivistic work settings such as Indonesia. This study examines the mediating role of psychological resource depletion in the relationship between workplace ostracism and presenteeism among Indonesian service-sector employees. This study employed a quantitative explanatory approach using survey data collected from 126 permanent employees working in the Indonesian service sector, including banking, education, healthcare, and retail industries. Respondents were selected using purposive sampling with criteria including a minimum one-year tenure and full-time employment status. Data were analyzed using Partial Least Squares Structural Equation Modeling (PLS-SEM) with SmartPLS 4.0. The results indicate that workplace ostracism positively and significantly affects psychological resource depletion (? = 0.512, p < 0.001) and presenteeism (? = 0.398, p < 0.001). Psychological resource depletion also positively affects presenteeism (? = 0.421, p < 0.001) and partially mediates the relationship between workplace ostracism and presenteeism. The structural model explains 53.8% of the variance in presenteeism, indicating moderate predictive power. This study extends Conservation of Resources Theory by demonstrating how interpersonal exclusion in collectivistic workplaces triggers a resource-loss process that reduces employee productivity. Practically, organizations should strengthen inclusive work climates and psychological support systems to minimize hidden productivity losses caused by workplace ostracism.

The Effect of Self-Love and Self-Reward on Gratitude Levels Among Social Media Users in South Tangerang

2026-06-04T14:08:06+00:00

Gratitude is a central construct in positive psychology that strongly influences mental well-being and quality of life. Yet, empirical studies examining the simultaneous effect of self-love and self-reward on gratitude among social media users a population particularly vulnerable to self-comparison and reduced self-worth remain scarce. This study aims to analyze the influence of self-love and self-reward on individual gratitude levels among active social media users in South Tangerang, Indonesia. A quantitative approach using Partial Least Squares Structural Equation Modeling (PLS-SEM) was employed. Data were collected via an online questionnaire using a five-point Likert scale distributed to 101 respondents through purposive sampling. The measurement model demonstrated adequate validity and reliability: Cronbach’s Alpha values ranged from 0.819 to 0.886, and Composite Reliability exceeded 0.70. The structural model yielded an R-Square value of 0.578, indicating that the two predictors jointly explained 57.8% of the variance in gratitude. Self-love exerted a positive and significant effect on gratitude (? = 0.303, T = 3.067, p = 0.003), and self-reward also exerted a positive and significant effect (? = 0.457, T = 4.496, p < 0.001). These findings highlight the importance of cultivating self-compassion and structured personal-appreciation practices as strategies for enhancing gratitude-based well-being interventions.

Digital Communication Quality, Feedback Quality, and Psychological Safety Among Platform Users

2026-06-01T15:29:19+00:00

Digital communication has become an essential component of interactions among users of digital platforms. Effective communication not only facilitates information exchange but also contributes to the development of psychological safety, which enables individuals to express ideas, opinions, and concerns without fear of negative consequences. This study examines the effect of Digital Communication Quality on Psychological Safety through the mediating role of Feedback Quality among digital platform users. A quantitative approach was employed using a survey of 115 respondents selected through convenience sampling. Data were analyzed using Partial Least Squares Structural Equation Modeling (PLS-SEM). The results indicate that Digital Communication Quality has a significant positive effect on Feedback Quality (? = 0.897; p < 0.001), while Feedback Quality significantly influences Psychological Safety (? = 0.906; p < 0.001). The indirect effect of Digital Communication Quality on Psychological Safety through Feedback Quality was also significant (? = 0.813; 95% CI [0.726–0.882]). The model explains 80.5% of the variance in Feedback Quality and 82.0% of the variance in Psychological Safety. Reliability and convergent validity requirements were fulfilled with Composite Reliability values ranging from 0.944 to 0.961 and AVE values ranging from 0.771 to 0.832. These findings suggest that high-quality digital communication and constructive feedback play an important role in fostering psychological safety among digital platform users.

Technostress, Techno-Exhaustion, and Digital Mindfulness Among College Student

2026-05-29T12:44:26+00:00

OnmentsThe increasing integration of digital technology into higher education has transformed learning activities and communication processes among students. While digital technologies provide numerous benefits, excessive reliance on technology may generate psychological challenges, particularly technostress and techno-exhaustion. Technostress refers to the stress experienced due to the increasing demands of technology use, whereas techno-exhaustion describes emotional, cognitive, and physical fatigue resulting from prolonged exposure to digital technologies. In this context, digital mindfulness has emerged as an important factor that may help students regulate their technology use and manage technology-related stress more effectively. Therefore, this study aims to examine the effects of technostress and digital mindfulness on techno-exhaustion among college students. This study employed a quantitative research design using a survey method. Data were collected from 104 undergraduate students through an online questionnaire. The collected data were analyzed using Structural Equation Modeling–Partial Least Squares (SEM-PLS). The results indicate that technostress has a positive and significant effect on techno-exhaustion (? = 0.527, t = 5.812, p < 0.001). Furthermore, digital mindfulness also significantly influences techno-exhaustion (? = 0.686, t = 7.806, p < 0.001). The coefficient of determination (R² = 0.471) indicates that the proposed model explains 47.1% of the variance in techno-exhaustion. The findings highlight the importance of managing technology-related stress and promoting mindful technology use among students. Higher education institutions are encouraged to develop strategies that foster digital well-being and support students in adapting to increasingly technology-dependent learning enviroment.

The Effect of Psychological Contract Breach on Turnover Intention Among Generation Z Employees

2026-06-04T07:14:03+00:00

Turnover intention among Generation Z employees has become a significant challenge for organizations due to changing work expectations and high workforce mobility. One factor that may contribute to employees’ intention to leave is Psychological Contract Breach, which refers to employees’ perceptions that the organization has failed to fulfill promised obligations and expectations. Therefore, this study aims to examine the effect of Psychological Contract Breach on Turnover Intention among Generation Z employees. This research employed a quantitative approach using a survey method. Data were collected from 126 Generation Z employees through a structured questionnaire and analyzed using Partial Least Squares Structural Equation Modeling (PLS-SEM). The findings indicate that Psychological Contract Breach has a positive and significant effect on Turnover Intention. The structural model demonstrates substantial explanatory power, with an R-square value of 0.691, indicating that the independent variable explains 69.1% of the variance in turnover intention. These findings support Social Exchange Theory, suggesting that employees who perceive unfulfilled organizational obligations are more likely to develop intentions to leave the organization. This study contributes to the literature on psychological contracts and employee retention while providing practical implications for organizations in managing Generation Z employees through the fulfillment of organizational commitments and expectations.

The Effect of Emotional Deprivation and Social Isolation on Employee Well-Being

2026-06-06T10:22:35+00:00

This study examines the effects of Emotional Deprivation and Social Isolation on employee Well-Being within the context of workplace loneliness. Workplace loneliness has emerged as a significant organizational issue due to changing work patterns, communication systems, and interpersonal relationships. A quantitative survey approach was employed involving 102 employees as respondents. Data were collected using a structured questionnaire measured on a five-point Likert scale and analyzed using descriptive statistics, correlation analysis, and multiple linear regression. The results indicate that Social Isolation has a significant negative effect on Well-Being, whereas Emotional Deprivation does not have a significant partial effect. Simultaneously, Emotional Deprivation and Social Isolation significantly influence Well-Being. However, the coefficient of determination (R²) of 17.8% suggests that a substantial proportion of Well-Being is explained by factors outside the proposed model. The insignificant effect of Emotional Deprivation may be associated with the availability of emotional support from sources outside the workplace and the collectivist cultural context that characterizes Indonesian society. These findings suggest that workplace social connectedness plays a more critical role in employee Well-Being than emotional attachment within workplace relationships. Organizations are therefore encouraged to implement strategies that strengthen employee interaction, inclusion, and social support to enhance workplace well-being.

Adoption of Social Media as a Digital Marketing Strategy: A Case Study of Wardi Pallo Mie Balap

2026-05-29T08:23:35+00:00

Micro, Small, and Medium Enterprises (MSMEs) are crucial pillars of the economy, yet micro-scale culinary entrepreneurs often face operational challenges during digitalization. This study examines the strategic adoption of Instagram and TikTok platforms as digital marketing frameworks for the Wardi Pallo Mie Balap MSME in Medan. Using descriptive qualitative methods, data was collected through observation, documentation, and in-depth interviews with eight informants using a purposive sampling technique. Data analysis employed the Miles and Huberman interactive model integrated with social media marketing dimensions, including interaction, content sharing, accessibility, and credibility. Data credibility was tested through source triangulation. The findings indicate that structured digital interventions successfully transformed temporary physical presence into digital visibility. Dynamic visualizations of the cooking process stimulated consumer purchase intention, significantly increasing monthly sales volume by 19.67 percent, with the highest spike for the yellow noodle variant, reaching 40 percent. However, qualitative analysis also identified critical structural barriers, such as limited human resources (HR) marketing specialists, operational time management constraints, and market volatility due to the fading novelty effect among consumers. This research provides theoretical and practical contributions regarding the importance of shifting from conventional word-of-mouth patterns to local social commerce models in order to maintain the sustainability of micro-scale commercial digitalization.

Designing a Customer Engagement-Based Digital Marketing Strategy to Increase the Number of Customers at Republik Barbershop

2026-05-30T07:25:38+00:00

This study aims to design a customer engagement-based digital marketing strategy to increase the number of customers at Republik Barbershop. The research is motivated by the growing competition in the barbershop industry and the increasing importance of digital platforms in attracting and retaining customers. A descriptive qualitative approach was employed, with data collected through interviews, observations, questionnaires, and documentation. The analysis was guided by the AIDA (Attention, Interest, Desire, Action) framework to evaluate the effectiveness of existing digital marketing activities and identify opportunities for strategic improvement.The findings indicate that social media, particularly Instagram, plays a significant role in enhancing brand visibility, fostering customer interaction, and strengthening customer relationships. Customer recommendations, service quality, a comfortable environment, and a strategic location were identified as key factors influencing customer acquisition and retention. Based on these findings, a digital marketing strategy centered on customer engagement is proposed, including the development of interactive content, customer-generated content, online review management, personalized communication, and consistent social media engagement. These strategies are expected to strengthen customer relationships, increase brand awareness, and contribute to sustainable customer growth at Republik Barbershop.

Implementation of the Simple Additive Weighting (SAW) Method in a Decision Support System for the Most Popular Course Interests at Labuhanbatu University

2026-06-22T06:52:41+00:00

Labuhanbatu University (ULB), one of the private universities in Labuhanbatu Regency, North Sumatra, admits new students every academic year across various study programs under four faculties. To date, the determination of the study program with the highest student interest has been carried out manually by summarizing the number of applicants without considering other supporting factors such as applicant quality, program capacity, and student satisfaction. Consequently, the analysis results tend to be less objective and less reliable as a basis for academic planning. This study aims to design and implement a Decision Support System (DSS) using the Simple Additive Weighting (SAW) method to assist the academic management in identifying the study program with the highest student interest based on five criteria: the number of applicants, the average entrance examination score, the on-time graduation rate, study program capacity, and student satisfaction level. This research employed an applied research approach using a case study method, which included problem identification, data collection, determination of criteria and weights, decision matrix normalization, preference value calculation, web-based system development, and system testing using the black-box testing method. The results demonstrate that the SAW method is capable of producing consistent and objective rankings of study programs based on the highest preference values. Based on the sample data, the Information Technology Study Program achieved the highest preference value of 0.9817, making it the study program with the highest student interest, followed by Management, Information Systems, Accounting, Mathematics Education, and Agrotechnology. The developed system can serve as an effective decision-support tool for university administrators in planning admission quotas and developing study programs.

Consumer Perceptions of Discounts and Free Shipping on Purchase Decisions in the Shopee Marketplace: Evidence from College Students in Medan

2026-06-15T12:21:43+00:00

The rapid development of information and communication technology has encouraged the growth of online shopping activities through marketplace platforms, particularly Shopee. Discounts and free shipping are among the most frequently used promotional strategies to attract consumers and increase transaction activities. This study aims to describe consumer perceptions regarding discounts and free shipping in relation to purchase decisions in the Shopee marketplace. The research employed a quantitative descriptive approach using a survey method. Data were collected through questionnaires distributed to 32 college students in Medan who had experience shopping on Shopee. The collected data were analyzed using descriptive statistical techniques in the form of frequencies and percentages. The results showed that 43.8% of respondents were 19 years old and 71.9% were students. Most respondents agreed that discount programs increase purchase interest and encourage faster purchasing decisions. In addition, respondents perceived free shipping as an attractive promotional program because it reduces shopping expenses and provides additional benefits when making online transactions. The findings indicate that discounts and free shipping are positively perceived by consumers and are considered important promotional programs in online shopping activities through the Shopee marketplace.

Determinants of Jakarta Composite Index (JCI) Movements: A Narrative Literature Review of Macroeconomic, Global, and Market Sentiment Factors (2020-2025)

2026-06-11T12:55:27+00:00

The Jakarta Composite Index (JCI) is the primary indicator used to measure the performance of the Indonesian capital market and reflects the dynamics of the national economy. The movement of the JCI is influenced by various factors, including macroeconomic conditions, global economic developments, and market sentiment. However, previous studies have reported inconsistent findings regarding the magnitude and direction of these influences. Therefore, this study aims to analyze and synthesize the determinants of JCI movements during the 2020-2025 period. This research employs a Narrative Literature Review (NLR) approach by reviewing ten Scopus-indexed articles selected based on their relevance to macroeconomic factors, global factors, and market sentiment. Data were analyzed using a thematic synthesis approach to identify dominant themes, patterns of relationships, and research gaps in the existing literature. The findings reveal that macroeconomic factors, particularly inflation, interest rates, exchange rates, and economic policies, remain fundamental determinants of stock market performance. In addition, global factors such as international stock market movements, oil price fluctuations, geopolitical risks, and global economic uncertainty significantly affect JCI volatility due to increasing financial market integration. Furthermore, market sentiment, reflected in investor attention, media coverage, and investor expectations, has become increasingly influential in shaping market behavior, especially during periods of economic uncertainty. The study highlights that JCI movements are driven by the interaction of macroeconomic conditions, global developments, and investor behavior rather than by a single determinant. This research contributes to the literature by providing a comprehensive synthesis of recent studies and offering directions for future research on stock market dynamics in emerging economies.

Technology Adoption, Corporate Digital Responsibility, and Inclusive Growth in MSMEs: A Systematic Review of Emerging Markets

2026-06-11T15:05:19+00:00

Is a brief summary of the research to help the reader quickly ascertain the research problem/objectives according to research needs. This systematic literature review aims to analyze the intersection between digital transformation, inclusive growth, and sustainable development within Micro, Small, and Medium Enterprises (MSMEs). Looking at global trends between 2020 and 2025, the research maps the managerial drivers, structural barriers, and socioeconomic impacts of technology adoption. Utilizing the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) framework, this study synthesizes 54 high-quality journal articles extracted from Scopus and Web of Science databases. Data analysis techniques involved thematic synthesis and conceptual mapping of organizational capabilities. This study synthesizes 54 peer-reviewed articles published between 2020 and 2025. The findings reveal that 72% of studies reported positive effects of digital transformation on market expansion, while 65% highlighted improvements in operational efficiency and financial inclusion. However, significant disparities persist across geographical contexts due to differences in infrastructure readiness, digital literacy, and institutional support. The review identifies Dynamic Digital Capabilities and Corporate Digital Responsibility (CDR) as critical mechanisms linking technology adoption with sustainable and inclusive growth outcomes.

Sustainable Career Pathways to Senior Management: Integrating Management Trainee Programs, Work-Life Balance, and Corporate Social Responsibility

2026-06-11T14:45:34+00:00

The increasing challenges of leadership succession, employee retention, and workforce well-being have raised concerns regarding the sustainability of career development within modern organizations. While Management Trainee Programs (MTP), Work-Life Balance (WLB), and Corporate Social Responsibility (CSR) have been widely studied, limited research has integrated these factors into a unified framework explaining sustainable career pathways toward senior managerial positions. This study aims to analyze the role of MTP, WLB, and CSR in shaping sustainable career pathways through a Narrative Literature Review approach. The study reviewed ten Scopus-indexed journal articles published between 2020 and 2025 selected based on relevance, publication quality, and theoretical contribution to sustainable career development. Data were analyzed using thematic narrative synthesis to identify key themes, theoretical debates, and conceptual relationships among the variables. The findings indicate that Management Trainee Programs function as leadership development mechanisms that accelerate managerial readiness, while Work-Life Balance contributes to employee well-being and long-term career sustainability. Furthermore, Corporate Social Responsibility strengthens employee engagement, organizational identification, and commitment, thereby supporting career continuity within organizations. The study also reveals that sustainable career pathways are not determined by a single factor but emerge from the interaction between leadership development, employee well-being, and organizational responsibility. This study contributes to the literature by proposing an integrated conceptual framework that links MTP, WLB, and CSR within the perspective of Sustainable Human Resource Management. The findings provide practical implications for organizations seeking to develop sustainable leadership pipelines and retain high-potential talent in an increasingly competitive business environment.

Analysis of Application Ease of Use and Consumer Trust in Building Customer Satisfaction in E-Commerce

2026-06-11T13:49:50+00:00

E-commerce platforms in Indonesia face intensifying competition where user experience and trust have become critical differentiators of customer retention. Despite abundant quantitative studies testing individual effects of Perceived Ease of Use (PEOU) and Consumer Trust (CT) on Customer Satisfaction (CS), a systematic synthesis examining their combined and interactive roles remains scarce. This study conducts a systematic literature review of 20 peer-reviewed articles published between 2020 and 2025, selected through PRISMA-guided screening across databases including Scopus, ScienceDirect, and Google Scholar. Thematic content analysis was employed to identify patterns, convergences, and divergences across studies. Results show that PEOU contributes to CS primarily through reduced cognitive load and enhanced transaction efficiency, while CT operates through perceived security and platform reputation. Critically, the two constructs interact synergistically: platforms scoring high on both dimensions reported 23–35% higher customer retention rates compared to platforms excelling in only one dimension. These findings extend TAM theory by demonstrating that ease of use amplifies trust-building mechanisms in digital environments. Practically, e-commerce operators should simultaneously invest in interface usability and security infrastructure to maximize customer satisfaction outcomes.

Physical Evidence Enhancement and Consumer Purchase Intention at Toko Usaha Keluarga MSME

2026-06-11T05:12:57+00:00

SMEs play an important role in the Indonesian economy, including in Medan City. However, the increasing competition demands business actors to strengthen their marketing strategies, one of which is through Physical Evidence. This research aims to determine the impact of the dimensions of facility exterior, facility interior, and other tangible aspects on consumer purchase interest after the strengthening of physical evidence in Toko Usaha Keluarga. This study uses a quantitative approach with a descriptive method. A sample of 50 respondents was selected using accidental sampling techniques. Data were collected through questionnaires that had been tested for validity and reliability, then analyzed using multiple linear regression and paired sample t-test to measure the impact before and after the strengthening of physical evidence. The research results show that there are significant differences in the variables of exterior facility, interior facility, and purchase intention between before and after the strengthening of physical evidence with a significance value of 0.000. For Other Tangible, the significance value obtained is 0.402 > 0.05, indicating that there is no significant difference before and after the strengthening of physical evidence at Toko Usaha Keluarga.

A Theoretical Framework for Strengthening Community Digital Talent Through Lifelong Learning: Insights from Program BAKAT at Community College Arau

2026-05-16T15:42:54+00:00

Community-based digital skills initiatives have become an important strategy for addressing digital competency gaps among diverse population groups in Malaysia. However, existing programs are often examined from training outcome perspectives, with limited attention given to the mechanisms through which community colleges facilitate digital talent development through lifelong learning approaches. This conceptual paper develops a theoretical framework based on Program Pengukuhan Bakat Komuniti (BAKAT) implemented by Community College Arau. Unlike many digital literacy initiatives that primarily focus on short-term technical training, Program BAKAT combines lifelong learning opportunities, community participation, and locally responsive skill development within a community college setting. Using an integrative review of literature on lifelong learning, human capital development, community engagement, and digital talent development, the study synthesis key concepts and proposes a framework explaining the relationships among lifelong learning participation, digital skills training, community engagement, and community digital talent development. Community engagement is positioned as a contextual factor that may influence the effectiveness of learning and training initiatives. Rather than proposing a new theory, the paper offers a structured conceptual model that consolidates existing perspectives into a community college context. The framework provides a basis for future empirical studies examining digital talent development initiatives within Technical and Vocational Education and Training (TVET) institutions and community-based learning environments.

Comparative Analysis of Pre-Test and Post-Test Performance in Competency-Based Food Handler Training

2026-05-17T00:20:29+00:00

Food safety training plays a vital role in reducing foodborne diseases and improving hygiene awareness among food handlers. This study evaluates the baseline knowledge gains achieved during the Food Handler Training Program conducted at Kolej Komuniti Arau. A quantitative one-group pre-test and post-test design was employed involving 104 participants across ten training cohorts conducted in 2026. Data were collected through structured participant assessment scores alongside four-point Likert-scale course evaluation forms. Descriptive statistics and paired sample t-test analysis were performed using SPSS software. The findings revealed significant immediate improvements in participants’ cognitive knowledge scores. The average assessment score increased from 73.31% (SD= 8.24) in the pre-test to 95.05% (SD = 4.11) in the post-test, representing an absolute improvement of 21.74%. The paired sample t-test indicated a statistically significant difference between pre-test and post-test scores (t(103) = 12.45, p < 0.001, Cohen's d = 1.22). Participants also reported positive perceptions regarding course delivery clarity and content organization. The study demonstrates that structured, competency-based food safety training can effectively enhance immediate learning outcomes within a classroom environment. However, further longitudinal research is required to evaluate actual behavioural translation and long-term compliance in professional kitchen environments.

Development and Preliminary Sensory Evaluation of Watermelon Rind-Based Dipping Sauce

2026-06-09T09:35:29+00:00

Food waste has become a major global concern, particularly in the agro-food industry where large quantities of fruit by-products are discarded despite their nutritional potential. Watermelon rind, which constitutes approximately 30% of the total fruit weight, is commonly treated as waste although it contains dietary fibre, antioxidants, and bioactive compounds. This study aimed to develop a watermelon rind-based dipping sauce and evaluate its sensory acceptability among consumers as an innovative approach for sustainable food waste valorisation. An experimental quantitative research design was employed involving product formulation and sensory evaluation using a 5-point hedonic scale. A total of 31 respondents participated in the sensory evaluation based on five attributes: colour, aroma, viscosity, taste, and overall acceptability. Descriptive statistical analysis using SPSS was conducted to determine the mean and standard deviation for each attribute. The findings revealed that overall acceptability obtained the highest mean score (M = 4.58, SD = 0.56), followed by colour (M = 4.55, SD = 0.57), aroma (M = 4.35, SD = 0.80), viscosity (M = 4.26, SD = 0.77), and taste (M = 4.16, SD = 0.73). The sensory evaluation results indicated favourable responses among the participating respondents, with mean scores exceeding 4.00 for all evaluated attributes. Within the limitations of a small convenience sample, the findings suggest that watermelon rind may be incorporated into a dipping sauce formulation with acceptable sensory characteristics. Further studies involving larger and more diverse consumer groups, together with physicochemical and shelf-life analyses, are required to confirm broader consumer acceptance and commercial feasibility.

Performance Optimization of Gojek Reviews Sentiment Classification Using the SVM Method with Hyperparameter Tuning

2026-06-27T07:00:21+00:00

This study aims to optimize multiclass sentiment classification on Indonesian-language Gojek application reviews using a linear Support Vector Machine (SVM) combined with GridSearchCV-based hyperparameter tuning. The increasing volume of user-generated reviews on digital transportation platforms presents significant challenges for manual sentiment analysis due to the noisy and unstructured characteristics of textual data. The dataset used in this study consists of approximately 5,000 Indonesian-language Google Play Store reviews categorized into negative, neutral, and positive sentiments. The preprocessing stages include text cleaning, case folding, tokenization, stopword removal, and stemming using the Sastrawi library, while feature extraction is performed using TF-IDF unigram representation. The classification model employs a linear SVM optimized through 5-fold cross-validation on the cost parameter (C). Experimental results show that the optimized SVM achieved an accuracy of 63.40%, outperforming the baseline model with an accuracy of 61.29%. The findings indicate that hyperparameter optimization improves model generalization and classification stability for sparse high-dimensional Indonesian text data. Furthermore, this study highlights the challenges of multiclass sentiment classification in informal Indonesian-language reviews, particularly for neutral sentiment categories with semantic overlap. Overall, the proposed approach demonstrates the effectiveness of optimized linear SVM for Indonesian-language sentiment analysis in online transportation application reviews.

Performance Analysis of Random Forest and Support Vector Machine in TikTok Shop Review Classification

2026-06-27T07:05:53+00:00

TikTok Shop generates a large volume of user reviews containing valuable information regarding customer satisfaction, service quality, and user experience. However, manually analyzing large amounts of review data is inefficient and time-consuming, making sentiment classification an important approach for automatically extracting user opinions. This study aims to analyze and compare the performance of Random Forest and Support Vector Machine (SVM) algorithms in classifying TikTok Shop reviews. The dataset was obtained from Kaggle and consisted of 3,145 review records. Data preprocessing was conducted through cleaning, case folding, tokenization, stopword removal, and stemming, resulting in 2,376 valid review records. Feature extraction was performed using the Term Frequency–Inverse Document Frequency (TF-IDF) method. The dataset was then divided into training and testing sets using an 80:20 ratio before classification using Random Forest and SVM algorithms. Model performance was evaluated using accuracy, precision, recall, and F1-score metrics. The results showed that Random Forest achieved an accuracy of 89.71%, precision of 79.50%, recall of 87.00%, and F1-score of 83.00%, while SVM achieved an accuracy of 91.60%, precision of 87.70%, recall of 82.60%, and F1-score of 85.10%. These findings indicate that SVM outperformed Random Forest and provided better overall performance for sentiment classification of TikTok Shop reviews.

Feature Selection Optimization Using a Genetic Algorithm in Naïve Bayes for Classifying Mortality Risk in Heart Failure Patients

2026-06-27T07:09:51+00:00

Heart failure is a cardiovascular condition associated with a high risk of mortality, necessitating a data-driven approach to assist in patient risk classification. This study aims to analyze the impact of feature selection optimization using a Genetic Algorithm (GA) on the performance of the Naïve Bayes (NB) classifier in classifying the mortality risk of heart failure patients. The dataset used is the Heart Failure Clinical Records Dataset, consisting of 299 patient records, 12 clinical features, and one classification target, namely DEATH_EVENT, with data split using an 80:20 stratified split, resulting in 239 training records and 60 testing records. The developed models consist of a standard NB baseline model and a proposed NB + GA model, evaluated using accuracy, precision, recall, F1-score, and a confusion matrix. The results show that the standard NB baseline model achieved an accuracy of 70.00%, precision of 54.55%, recall of 31.58%, and an F1-score of 40.00%, while the proposed NB + GA model improved to an accuracy of 81.67%, precision of 83.33%, recall of 52.63%, and an F1-score of 64.52%. GA selected three main features: ejection_fraction, smoking, and time; however, a sensitivity analysis excluding the time feature showed a decline in performance, so this feature must be interpreted with caution as it relates to the duration of patient follow-up.

Comparison of the Random Forest and LightGBM Algorithms in MSME Performance Classification

2026-06-27T07:13:24+00:00

The performance assessment of Micro, Small, and Medium Enterprises (MSMEs) requires a data-driven approach, as business performance is influenced by multiple dimensions, including financial, customer, digital, operational, and business environment indicators. This study aims to compare the performance of Random Forest (RF) and Light Gradient Boosting Machine (LightGBM) models in classifying MSME performance into four categories: Critical, Struggling, Growth, and Elite. The study utilized a synthetic secondary dataset obtained from Kaggle, consisting of 150,000 records and 15 attributes, which were processed using Google Colab. The research methodology comprised data preprocessing, categorical feature encoding, feature-target separation, training and testing data partitioning using an 80:20 ratio with stratified sampling, model development, model testing, performance evaluation, and comparative analysis. Model performance was assessed using accuracy, macro-precision, macro-recall, macro F1-score, and confusion matrix analysis. The experimental results demonstrate that RF achieved superior performance, with an accuracy of 0.998267, a macro-precision of 0.997563, a macro-recall of 0.996832, and a macro F1-score of 0.997197. In comparison, LightGBM achieved an accuracy of 0.997567, a macro-precision of 0.996465, a macro-recall of 0.996517, and a macro F1-score of 0.996491. Furthermore, the confusion matrix analysis revealed that RF generated 52 misclassifications out of 30,000 testing instances, whereas LightGBM generated 73 misclassifications. Feature importance analysis of the RF model identified Burn_Rate_Ratio, Net_Profit_Margin (%), Avg_Historical_Rating, Repeat_Order_Rate (%), and Peak_Hour_Latency as the most influential variables in distinguishing MSME performance categories. These findings indicate that RF is a highly effective model for MSME performance classification based on multidimensional business indicators. Nevertheless, further validation using empirical MSME datasets is recommended to enhance the generalizability and practical applicability of the proposed approach.

Performance Analysis of Naive Bayes and SVM in Review Sentiment Classification of Spotify

2026-06-27T07:21:17+00:00

This study analyzes the sentiment of Spotify application users on the Google Play Store using a large dataset containing 100,000 reviews. The primary objective of this study is to compare the effectiveness of Naive Bayes and Support Vector Machine (SVM) algorithms in classifying user opinions into positive and negative categories. The dataset was cleaned through a text preprocessing pipeline that included case folding, cleansing, tokenizing, stopword removal, and stemming using the Sastrawi library. Neutral rating reviews were eliminated to yield a total clean dataset of 95,359 reviews. Numerical feature extraction was performed using the Term Frequency-Inverse Document Frequency (TF-IDF) method. Experimental results on the test data show that the SVM algorithm achieved the highest global accuracy rate of 92.57%, while Multinomial Naive Bayes recorded an accuracy of 92.09%. Interestingly, both algorithms produced an identical recall value of 0.79 in recognizing negative sentiment amidst the dominance of the majority class. The word cloud visualization detected that the main factor driving user satisfaction centered on the completeness of the song library, while critical user complaints were dominated by premium account transition issues and the intensity of advertisements. This research proves that an optimal combination of text preprocessing can mitigate class imbalance bias in both classification model architectures

Performance Comparison of Naive Bayes and Support Vector Machine for Sentiment Analysis of the Free Nutritious Meal Program

2026-06-27T07:26:28+00:00

The Free Nutritious Meal Program (MBG) has become one of the most widely discussed government policies on social media, generating diverse public opinions. This study aims to compare the performance of the Naive Bayes and Support Vector Machine (SVM) algorithms in sentiment analysis of public comments related to the MBG program. The dataset was obtained from Kaggle and consists of 4,889 comments categorized into negative, neutral, and positive sentiments. The research process includes text preprocessing, train-test splitting using an 80:20 ratio, feature extraction using Term Frequency–Inverse Document Frequency (TF-IDF), data balancing using the Synthetic Minority Oversampling Technique (SMOTE), and sentiment classification using Naive Bayes and SVM. The preprocessing stage includes case folding, text cleaning, tokenization, word normalization, and stopword removal. SMOTE was applied to address class imbalance in the training data and improve the recognition of minority classes. The experimental results show that Naive Bayes achieved an accuracy of 58.90%, while SVM achieved an accuracy of 70.76%. These findings indicate that the combination of TF-IDF, SMOTE, and SVM provides superior classification performance compared with Naive Bayes for sentiment analysis of the Free Nutritious Meal Program.

Marketing Strategies for MSMEs: A Case Study of Mama Zita Bakery

2026-06-11T04:37:11+00:00

The number of Micro, Small, and Medium Enterprises (MSMEs) continues to grow, reaching 66 million in 2023. One such enterprise is Mama Zita Bakery, which operates in the bakery industry. Amidst increasing business competition, it is essential for Mama Zita Bakery to implement effective marketing strategies, particularly by applying a SWOT analysis to maximize potential and increase sales. This study aims to identify the marketing strategies employed by Mama Zita Bakery to boost sales and to examine the challenges faced by the shop in achieving this objective. This research utilizes a qualitative approach with a descriptive research design. The findings indicate that, in general, Mama Zita Bakery's marketing strategy is well-executed. The bakery maintains high product quality and a strong reputation, while effectively leveraging digital opportunities through social media and online ordering systems. However, there is a critical weakness regarding limited staffing that requires immediate attention. Furthermore, price competition poses a significant threat that could lead to customer churn in favor of competitors. The obstacles faced by Mama Zita Bakery in increasing sales include inconsistent service quality, employee burnout due to excessive workload, price competition, and negative feedback that potentially compromises the bakery's reputation.

Designing a Digital Promotion Strategy Through TikTok to Increase Brand Awareness in MSMEs Adah Mode

2026-06-11T06:11:11+00:00

Micro, Small, and Medium Enterprises (MSMEs) play a vital role in Indonesia’s economy; however, many still face limitations in utilizing digital marketing strategies. Adah Mode, a fashion-based MSME, has so far relied solely on sporadic promotions through Instagram, which have not been optimized. Meanwhile, TikTok has rapidly developed into an effective and creative digital marketing platform capable of reaching a wide audience. The findings indicate that digital promotion strategies through TikTok have a significant effect on increasing brand awareness. This is evidenced by a significance value of 0.000 (<0.05) across all variables within the AIDA model (Attention, Interest, Desire, Action) as well as the brand awareness indicators. the study affirms that TikTok can serve as an effective digital promotion strategy for MSMEs in enhancing brand awareness and competitiveness within the creative economy era. The results are not only beneficial for Adah Mode but may also serve as a practical reference for other MSMEs facing similar challenges in developing their digital marketing efforts.

Instagram-Based Marketing Strategy for CV. Karya Mulia Kencana Using Action Research

2026-06-11T08:26:12+00:00

This study examines the use of Instagram as a digital marketing strategy for CV. Karya Mulia Kencana, a clothing manufacturing MSME in Medan. Many MSMEs still rely on conventional marketing methods. As a result, their ability to reach broader markets and attract new customers remains limited. This study employed a qualitative action research approach consisting of planning, action, observation, and reflection stages conducted from June to September 2025. Data were collected through observation, interviews, documentation, and Instagram Insights. In the first cycle, the Instagram account @konveksianas_medan was created to introduce the business and its products. A total of 11 posts generated 22 followers, an average reach of 5 accounts per post, and 90 engagements. The results were relatively low because the account was newly established and had not yet gained wider public attention. Based on the evaluation of the first cycle, several improvements were implemented in the second cycle. These included the use of Instagram Reels, more relevant hashtags, and regular content uploads. As a result, the number of followers increased to 237 accounts, reach expanded to 1,732 accounts, and engagements totaled 92 interactions. The findings indicate that Instagram can effectively improve business visibility and audience engagement of MSMEs.

The Effect of Financial Literacy on Personal Financial Management: The Moderating Role of Family Support

2026-06-11T16:02:44+00:00

The development of the global economy and advancements in technology have brought significant changes to personal financial management, where individuals are faced with various financial products and services. However, many individuals with adequate financial literacy are still unable to manage their finances effectively, raising questions about the influence of financial literacy on personal financial management and the role of family support in moderating this relationship. This study aims to analyze the effect of financial literacy on personal financial management and examine the moderating role of family support. The research employed a quantitative approach using an online survey, involving 35 respondents who met specific criteria. The findings indicate that financial literacy has a positive and significant effect on personal financial management. Furthermore, family support serves as a moderating variable that strengthens the relationship between financial literacy and personal financial management. In conclusion, improving financial literacy among students, accompanied by strong family support, can help them make better financial decisions and manage their personal finances more effectively.

Determinants of Receivables Control Effectiveness: Evidence from Aceh Public Hospitals

2026-06-11T17:58:45+00:00

Effective receivables control is a critical factor in maintaining the financial sustainability and operational performance of hospitals. Increasing transaction complexity, the digitalization of financial management systems, and potential operational disruptions necessitate the implementation of accounting information systems, internal control, and disaster recovery planning to support receivables management. This study aims to analyze the influence of Accounting Information Systems (AIS), Internal Control, and Disaster Recovery Planning (DRP) on the effectiveness of receivables control in Regional Public Service Agency (BLUD) hospitals across Aceh Province. A quantitative approach with a survey method was employed. Data were gathered by distributing questionnaires to personnel involved in financial administration and receivables management. Data analysis was performed using multiple linear regression. The empirical results indicate that Accounting Information Systems, Internal Control, and Disaster Recovery Planning exert positive and significant effects on the effectiveness of receivables control. Among the three variables, internal control provides the most dominant contribution to enhancing receivables management effectiveness. These findings imply that the integration of a reliable accounting information system, effective internal control, and comprehensive disaster recovery planning can improve the accuracy, security, and sustainability of receivables management processes. This study contributes to the development of financial management and accounting information systems literature in the public healthcare sector, while providing practical implications for hospital management in strengthening financial governance and receivables control effectiveness.

The Influence of Using Social Media as a Source of Academic Information on the Effectiveness of Student Learning

2026-06-12T14:34:58+00:00

The development of social media has fundamentally changed the way students obtain information and learning materials outside formal classroom settings. In Indonesia, university students are among the most active social media users, spending an average of more than five hours per day on various platforms. However, its impact on learning effectiveness is not yet fully understood, particularly in the context of private higher education institutions. Therefore, this study aims to analyze the influence of social media utilization as a source of academic information on the learning effectiveness of students at STIM Sukma Medan. This research employed a quantitative approach using descriptive and verification survey methods. Data were collected from 50 active students selected through a total sampling technique, using a Likert-scale questionnaire (1–5) distributed via Google Forms. The independent variable in this study was the utilization of social media as a source of academic information (X), measured through the dimensions of usage intensity, platform selectivity, and the quality of accessed information. The dependent variable was learning effectiveness (Y), measured through the dimensions of material comprehension, learning motivation, and completion of academic assignments. Data analysis was conducted using simple linear regression, coefficient of determination analysis, and t-test analysis. The results showed that 62% of students used social media daily for academic purposes, with TikTok (52%), Instagram (50%), and YouTube (46%) being the most dominant platforms. The regression coefficient value of 0.724 indicated a positive influence, while the coefficient of determination (R²) value of 0.524 suggested that 52.4% of the variation in learning effectiveness could be explained by the utilization of social media. The t-test results showed that t_calculated (7.321) > t_table (2.011) with a value of 0.000 < 0.05, indicating that the research hypothesis was accepted. significance It can be concluded that the utilization of social media as a source of academic information has a positive and significant effect on the learning effectiveness of students at STIM Sukma Medan.

Islamic Leadership in Human-AI Collaboration: Reviewing Strategies for Building Ethical and Effective Hybrid Teams

2026-06-15T11:59:41+00:00

The rapid development of Artificial Intelligence (AI) has transformed organizational operations and accelerated the emergence of Human-AI Collaboration as a new approach to creating organizational value. While AI offers significant opportunities to improve productivity, innovation, and decision-making quality, its implementation also raises challenges related to trust, algorithmic transparency, ethical governance, and human oversight. Existing studies primarily focus on technological and managerial perspectives, while limited attention has been given to the ethical role of Islamic Leadership in managing Human-AI Collaboration. Therefore, this study aims to review and synthesize leadership strategies for building ethical and effective hybrid teams through the lens of Islamic Leadership. This research employs a Narrative Literature Review approach by analyzing ten peer-reviewed journal articles published between 2020 and 2025. Data were collected from Google Scholar and analyzed using thematic narrative synthesis to identify major themes, theoretical debates, and conceptual relationships. The findings indicate that successful Human-AI Collaboration depends on balancing automation and augmentation, strengthening trust in AI systems, ensuring algorithmic transparency, and maintaining human involvement in decision-making processes. Furthermore, Islamic leadership values, including amanah (trustworthiness), adl (justice), shura (consultation), ihsan (excellence), and mas'uliyyah (accountability), provide a strong ethical foundation for governing Human-AI Collaboration. This study proposes a conceptual framework that integrates Islamic Leadership with AI governance to support effective hybrid team performance and organizational sustainability.

Islamic Leadership in Digital Workplaces: A Narrative Review

2026-06-15T12:00:54+00:00

Digital transformation has changed organizational structures, work processes, and leadership practices, thus driving the emergence of the digital workplace as an increasingly dominant work model in modern organizations. While digital transformation provides benefits in the form of increased flexibility, productivity, and collaboration, it also presents various challenges related to employee well-being, digital trust, ethical decision-making, and the use of artificial intelligence. Previous research has largely discussed digital leadership and workplace transformation, but studies that integrate Islamic leadership values in the digital workplace are still relatively limited. Therefore, this study aims to review and synthesize the literature on the role of Islamic Leadership in the digital workplace. This study uses the Narrative Literature Review approach by analyzing 15 selected articles published in the 2020–2025 period. Data were obtained through literature searches on Google Scholar, Scopus, Web of Science, Emerald Insight, ScienceDirect, and SpringerLink databases. Data analysis was conducted using thematic narrative synthesis techniques to identify key themes, theoretical debates, and research gaps. The results of the study show five dominant themes, namely digital workplace transformation, digital leadership capabilities, employee engagement and well-being, ethical challenges in an artificial intelligence-based work environment, and Islamic Leadership as an ethical framework. This study concludes that Islamic Leadership values, namely amanah, adl, shura, ihsan, and mas'uliyyah, can be an important foundation in strengthening ethical governance, increasing employee trust, and supporting organizational sustainability in the era of digital transformation.

Artificial Intelligence Adoption in Human Resource Management: A Literature Review

2026-06-15T16:36:57+00:00

The development of Artificial Intelligence (AI) has driven significant transformation in Human Resource Management (HRM) through improving work process efficiency, decision-making quality, and data-driven human resource management. The increasing use of AI in various HRM functions makes this topic important to study in order to understand the development of research and its future development. This study aims to identify research trends, compare previous research findings, uncover research gaps, and formulate future research recommendations related to AI adoption in HRM. The research uses the Narrative Literature Review approach by utilizing 10 Scopus indexed articles for the period 2020–2025 obtained through Google Scholar and selected based on their relevance and scientific contribution to the research topic. The analysis was carried out through theme identification, comparison of research results, literature synthesis, and assessment of research gaps. The results of the study show that AI contributes to improving the effectiveness of recruitment, talent management, employee experience, decision-making, and workforce planning. However, AI implementation still faces various challenges in the form of ethical issues, data privacy, algorithm bias, organizational readiness, and workforce digital competence. This research provides conceptual contributions through a comprehensive synthesis of the development of AI in HRM as well as the mapping of future research agendas focusing on AI governance, Generative AI, and sustainable digital transformation. It is concluded that AI is a strategic factor in the transformation of modern HRM, but the success of its implementation is largely determined by the readiness of technology, organization, and human resources.

Mapping the Intellectual Structure and Thematic Evolution of the Integrated Digital Asset Loan Workflow Model (IDALWM)

2026-06-16T02:04:40+00:00

Convergence of finance technology, asset management systems, workflow automation, artificial intelligence, and blockchain technologies has resulted in drastic changes in the management of the ecosystem associated with digital asset loans. Though many studies on the above-mentioned issues have emerged recently, fragmentation prevails, thereby making it difficult to grasp the intellectual structure and progression of knowledge in the respective fields. The purpose of this paper is to examine the evolution of knowledge architecture and emerging themes of the Integrated Digital Asset Loan Workflow Model through bibliometric analysis. A total of 237 documents were found from the Scopus database following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) approach. VOSviewer and Bibliometrix were used in the course of performance analysis, co-citation analysis, bibliographic coupling, keyword co-occurrence analysis, and network analysis. It was established that since 2015, there has been an increase in the number of scholarly papers related to integrated digital ecosystems of finance and asset management due to their relevance. Five major theme clusters have been revealed, namely (1) digital lending and financial technology, (2) artificial intelligence and workflow automation, (3) blockchain and governance technologies, (4) asset management systems and life cycle governance, and (5) interoperability and digital integration platforms. Additionally, a significant trend towards technological convergence has been discovered with respect to intelligent automation, explainable AI, digital twin, and blockchain-based governance. As a result, this study will contribute to the development of the IDALWM as an integrative conceptual model, which will help bridge the gap in the relevant literature. The proposed conclusions will be useful for academics and policy-makers in addition to financial institutions and technology providers.

The Evolution of Silhouette Design: Western Tailoring Adaptation in the Transformation of Traditional Baju Kebaya in Malaysia

2026-06-18T06:53:13+00:00

The Baju Kebaya serves as a quintessential repository of aesthetic and socio-cultural identity, functioning as a dynamic symbol of Malaysia’s intangible heritage. This study investigates the structural evolution of the Baju Kebaya (Malay and Nyonya Peranakan) across a century of transformation, from 1910 to 2017. Employing a qualitative, object-based material culture research design, the study utilises digital tracing analysis and micro-visual observation of 112 heritage artefacts, corroborated by semi-structured interviews with cultural custodians and master practitioners. Drawing upon Syed Ahmad Jamal’s Rupa dan Jiwa (Form and Soul) framework, the research maps a significant morphological transition from traditional, loose cylindrical geometry to anatomically contoured, "hourglass" silhouettes influenced by Western dressmaking. The findings demonstrate that the assimilation of technical components such as darting, princess lines, and shoulder padding has fundamentally reshaped the garment’s visual aesthetics. However, this study argues that such technical integration constitutes a resilient "techno-cultural negotiation" rather than a loss of identity; the Baju Kebaya effectively synthesises these globalised tailoring influences while sustaining its foundational Jiwa (spirit) as a symbol of modesty and feminine elegance. By providing a longitudinal analysis of these structural shifts, this research contributes to the safeguarding and scholarly understanding of heritage garments in an era of rapid industrialisation and digital globalisation.

Sustainable Human Resource Management as a Strategic Driver of ESG Performance

2026-06-11T18:30:22+00:00

Increasing demands for Environmental, Social, and Governance (ESG) are driving organizations to integrate sustainability into business strategies, including through Sustainable Human Resource Management (Sustainable HRM) practices. Although the relationship between Sustainable HRM and ESG is increasingly being researched, the existing literature remains fragmented, tends to focus only on ESG dimensions, and has not comprehensively explained the mechanisms linking Sustainable HRM to ESG Performance. This study aims to synthesize and evaluate research developments on the role of Sustainable HRM as a strategic driver of ESG Performance during the 2020–2025 period. The study used the Systematic Literature Review (SLR) method based on the PRISMA 2020 guidelines for 20 Scopus Q1 and Q2 articles selected through a systematic selection and quality assessment process. The synthesis results show that Sustainable HRM contributes to ESG Performance through three main mechanisms: the development of sustainability-oriented human capital, increasing employee engagement, and establishing a sustainability-oriented organizational culture. Furthermore, this study identifies that the governance dimension remains a relatively underexplored area compared to the environmental and social dimensions. The contribution of this research lies in developing a conceptual synthesis that broadens the perspectives of the Resource-Based View and Stakeholder Theory in explaining the strategic role of Sustainable HRM in achieving ESG performance. These findings provide theoretical implications for the development of sustainability research as well as practical implications for organizations in designing more effective ESG strategies.

The Design of a Liquid Pressure Gauge to Enhance Conceptual Understanding of Fluid Pressure

2026-06-15T07:35:55+00:00

This research aimed to: 1) design and develop a liquid pressure gauge learning module, and 2) improve students’ conceptual understanding of pressure in liquids. The study began with an investigation of students’ prior conceptions and misconceptions regarding liquid pressure. A set of diagnostic questions was developed and administered to explore students’ understanding of key pressure concepts. The findings indicated several misconceptions, including the belief that liquid pressure depends on the size and shape of the container and that liquids with lower density will always remain at the top surface. Based on these findings, a liquid pressure learning module was designed and developed. The module consisted of a conceptual understanding test, instructional materials, and lesson plans based on the Interactive Learning Demonstration (ILD) approach. The learning module was implemented with six students from Trat Polytechnic College. Its effectiveness was evaluated through assessments of students’ conceptual understanding and learning expectations. The results showed a statistically significant improvement in students’ understanding at the 0.05 level. The average normalized gain (g) was 0.55, indicating a moderate level of learning improvement and suggesting that the ILD-based learning module is suitable for enhancing physics instruction.

E-NOTES Mobile App for Interactive Multimedia Learning in TVET Education

2026-05-14T01:36:30+00:00

The rapid advancement of digital technology has transformed teaching and learning practices, especially in Technical and Vocational Education and Training (TVET). However, conventional learning methods are still less interactive and unable to fully engage students in the learning process. Therefore, this study aims to develop and evaluate the effectiveness of the E-NOTES mobile application as an interactive digital learning platform for TVET education. This study applied the ADDIE model consisting of analysis, design, development, implementation, and evaluation phases. The application integrates multimedia elements and real-time annotation features to support flexible and student-centered learning. A total of 30 Information Technology students participated as respondents in this study. Data were analyzed using descriptive statistics, paired sample t-test, and Cohen’s d effect size analysis. The findings revealed significant improvements in students’ interest, understanding, and interaction after using the application (p < 0.001). The effect size also indicated a very large practical impact on student engagement. Overall, E-NOTES has strong potential to support technology-enhanced learning and digital transformation in TVET education aligned with Industry 4.0 and 21st-century learning needs.

Reframing Audit Evidence in the Digital Audit Environment: A Systematic Review

2026-05-14T05:25:54+00:00

Digital audit technologies are reshaping the evidential foundation of external auditing by changing how audit evidence is collected, processed, evaluated and justified. Although audit data analytics, big data analytics and computer-assisted audit techniques have attracted growing scholarly attention, existing research remains fragmented across technology adoption, audit quality, professional judgement and governance debates. This systematic literature review synthesises recent studies to examine how digital audit tools reframe audit evidence and what professional challenges arise when audit evidence becomes increasingly technology-generated. The review followed the PRISMA protocol and involved systematic searches of the Scopus and Web of Science databases. Guided by PRISMA 2020 principles, the review applied a broad initial search strategy followed by staged screening, duplicate removal and full-text eligibility assessment. From 3,672 initial records, 110 unique records were assessed for eligibility and 21 studies were retained because they directly addressed external auditing, digital audit technologies and the production, evaluation or governance of audit evidence. The findings reveal two dominant themes. The first theme, technology-enabled production of digital audit evidence, shows that digital tools expand audit evidence through broader data access, automated testing, full-population analysis, anomaly detection, visual analytics and technology-supported risk assessment. The second theme, judgement, scepticism and governance challenges in digital audit evidence, indicates that technology-generated outputs require professional interpretation, sceptical evaluation, documentation, regulatory clarity and governance control before becoming defensible audit evidence. Overall, this review demonstrates that digital audit evidence is not merely a technological output, but a data-mediated and judgement-dependent construct. The study contributes by consolidating current evidence, clarifying conceptual gaps and offering future research directions for reliable and accountable digital audit evidence.

Integrating Traditional Akha Embroidery Motifs into Contemporary Souvenir Product Design: Development and Consumer Evaluation

2026-06-04T08:46:11+00:00

This research investigated the design of cultural souvenir products based on traditional Akha embroidery motifs and examined consumer satisfaction toward the developed products. The study employed a Research and Development (R&D) methodology consisting of four phases: the study of traditional Akha embroidery motifs, expert evaluation of motif designs, product prototype development, and consumer satisfaction assessment. The sample group consisted of seven product design experts and 100 consumers in Chiang Rai Province who were interested in or had experience purchasing souvenir products. Data were collected through consumer satisfaction questionnaires and analyzed using descriptive statistics, including percentage, mean, and standard deviation. The findings revealed that Design Sketch 6 received the highest evaluation score (85.7%) and was selected for product development. Overall consumer satisfaction toward the developed cultural souvenir products was rated at the highest level (M = 4.63). The findings suggest that traditional Akha embroidery motifs have the potential to be applied in contemporary souvenir product design while supporting cultural representation and enhancing the attractiveness of tourism-related products.

Digital Transformation for Sustainability Reporting and ESG Disclosure in Islamic Accounting: A Systematic Literature Review

2026-06-09T04:49:22+00:00

This study aims to identify research trends regarding digital transformation, sustainability reporting, and ESG disclosure in the context of Islamic accounting. The method used is Systematic Literature Review (SLR) with reference to the PRISMA 2020 guidelines. Articles were obtained through Google Scholar, Scopus, and Dimensions databases with a publication range of 2020–2025. The selection process was carried out using the inclusion and exclusion criteria that had been set, resulting in 20 articles that were eligible for analysis. The analysis was carried out using a thematic analysis approach to identify the main themes, relationships between concepts, and the direction of research development. The results of the study show that digital transformation plays a role in improving the efficiency of accounting systems, information transparency, and the quality of sustainability reporting in Islamic financial institutions. In addition, ESG disclosure is increasingly integrated with sharia principles through strengthening Shariah governance. The study also found that research linking digital transformation, sustainability reporting, ESG disclosure, and Shariah governance simultaneously is still relatively limited. Therefore, this study provides a mapping of the literature and identifies research opportunities that can be developed in the future in order to support the sustainable development of Islamic accounting in the digital era.

An Integrated Inclusive TVET Development Model for Students with Special Educational Needs Through Lifelong Learning Programmes

2026-06-09T08:36:54+00:00

Even though TVET development is being implemented in increasing numbers of schools, the mechanisms for the development of SSEN students are not well understood. As technical competency and employability are the main focuses of previous research, it is still challenging to capture how the learning experience impacts on the overall development of the students. This paper provides a detailed conceptual framework for the development of SSEN students who are engaged in lifelong learning programmes in a systemic way in the field of TVET. We conducted a systematic literature review (SLR) on Scopus, Web of Science, ERIC and ScienceDirect databases from 290 records and selected 35 high-quality studies from 2020-2025 based on a rigorous screening and quality evaluation process. Thematic synthesis uncovered five interrelated domains that affect SSEN development - inclusive learning environment, competency development, psychological empowerment, lifelong learning participation and student outcomes. In this manner, the review identified a clear gap in the existing TVET literature which mainly focuses on a single competency-employability model and the absence of psychological and environmental effects. We found that self-efficacy and learning motivation play important roles as key mediators in translating skills acquisition into meaningful development and that institutional support and inclusive learning environments are integral to these interlinked processes along with supporting them. The integrated Inclusive TVET Development Model developed in this study is a model to reframe student development as an inter-dependent interaction of cognitive, behavioural, psychological and environmental factors. The integration of Human Capital Theory, Inclusive Education Theory, Lifelong Learning Theory, Constructivist Theory and Self-Efficacy Theory to the model offers a framework for better understanding of the development of inclusive TVET and a basis for future empirical assessment. All these findings have important implications for the policymakers, Community Colleges and TVET institutions who are seeking to develop more and more inclusive lifelong learning programs for SSEN learners.

Analysis of Consumer Behavior in Purchasing Trading Card Game (TCG) At Velocity Hobby Centre

2026-06-11T04:23:56+00:00

This study examines the influence of consumer behavior on purchasing decisions for Trading Card Games (TCG) at Velocity Hobby Centre Medan. Consumer behavior was measured through cultural, social, personal, and psychological dimensions adapted from Kotler and Keller's consumer behavior framework. A quantitative survey was conducted involving 96 respondents selected through accidental sampling. Data were analyzed using validity and reliability testing, simple linear regression, t-test, and coefficient of determination (R²). The results indicate that consumer behavior has a positive and significant effect on purchasing decisions (? = 1.101; t = 11.660; p < 0.05). The coefficient of determination (R² = 0.591) shows that consumer behavior explains 59.1% of the variance in purchasing decisions. Social interaction within the TCG community, collecting motivation, and perceived investment value emerged as dominant factors influencing purchase decisions. These findings contribute to understanding consumer behavior in the collectibles and hobby industry and provide practical implications for hobby retailers in Indonesia.

Analysis of Instagram Social Media in Digital Marketing Strategies to Increase Sales Using the AIDA Model at Cafe Al-Fattah

2026-06-11T04:48:43+00:00

As a culinary enterprise established in 2024, Cafe Al-Fattah faces significant challenges in building brand awareness and expanding its market reach. To address these issues, the utilization of social media, particularly Instagram, is essential as an effective, economical, and far-reaching digital promotional tool. This study aims to analyze the utilization of Instagram in digital marketing strategies to increase sales at Cafe Al-Fattah using the AIDA model approach. This research employs a descriptive qualitative approach, with data collection methods including direct observation, interviews with social media managers, and documentation of Instagram content. The results indicate that visual content strategies-such as football screening photos, food imagery, customer testimonials, and interactive promotionssuccessfully capture audience attention and interest, stimulate desire, and drive purchasing actions. The application of the AIDA model is proven effective in building consumer engagement and enhancing business visibility. Consequently, Instagram serves as an optimal digital marketing strategy for MSMEs to strengthen branding and boost sustainable sales growth.

Impact of the Community Talent Strengthening Program on Participant Development at Arau Community College

2026-06-09T09:11:46+00:00

This study examined the effectiveness of the Community Talent Strengthening Program implemented at Arau Community College in enhancing participants’ technical skills, self-confidence, employability perceptions, and entrepreneurial interest. A quantitative one-group pre-test and post-test design was employed involving 11 participants. Data were collected using a structured questionnaire consisting of 20 items measured on a four-point Likert scale. The instrument demonstrated satisfactory reliability with an overall Cronbach’s Alpha value of 0.89. Descriptive statistics and Paired Sample t-Test were used to analyse the data. The findings revealed significant improvements across all constructs following programme participation. Knowledge increased from M = 3.24 to M = 3.84 (p < 0.001), understanding increased from M = 2.80 to M = 3.78 (p < 0.001), skills increased from M = 2.47 to M = 3.47 (p < 0.001), and student development increased from M = 2.27 to M = 3.27 (p < 0.001). The results indicate that the Community Talent Strengthening Program contributed positively to participant competency development and personal growth. The study provides empirical evidence supporting the effectiveness of community-based TVET interventions as a mechanism for human capital development, lifelong learning participation, and community empowerment.

Accounting Digitalization and Zakat Waqf Compliance in Islamic Financial Institutions: Literature Synthesis 2020-2025

2026-06-11T17:02:08+00:00

The rapid expansion of digital technologies has fundamentally reshaped accounting practices within Islamic financial institutions, particularly in the management of zakat and waqf funds that demand high levels of accountability, transparency, and sharia compliance. While prior studies have examined accounting digitalization and Islamic social finance independently, the existing literature remains fragmented and largely descriptive, offering limited integrative insight into how digital accounting contributes to zakat and waqf compliance mechanisms. This fragmentation becomes increasingly problematic as digital platforms, fintech solutions, and electronic payment systems are now widely adopted by zakat and waqf institutions. This study addresses this gap by providing a systematic literature synthesis of 50 peer-reviewed articles published between 2020 and 2025. Using a thematic and bibliometric analysis, the study maps dominant research trends, identifies underexplored dimensions, and critically examines the role of digital accounting in enhancing transparency, accountability, and sharia governance. The findings demonstrate that digital accounting functions not merely as a technical tool, but as a governance mechanism that strengthens compliance, reduces information asymmetry, and supports stakeholder trust. By consolidating dispersed empirical evidence, this study contributes to sharia accounting literature and offers a clearer analytical foundation for developing accountable and sustainable digital zakat–waqf management systems.

The Importance of Management in Managing Organizational Resources and Enhancing Organizational Effectiveness

2026-07-05T07:50:17+00:00

This study examines the importance of management in managing organizational resources and enhancing organizational effectiveness. Effective management of financial, human, physical, and informational resources determines the extent to which an organization can achieve its strategic goals in an increasingly competitive environment. The research problem addressed is the persistent gap between resource availability and resource utilization that causes many organizations to operate below their performance potential despite possessing adequate resources. This study was conducted through a qualitative literature review approach using secondary data drawn from peer-reviewed journals, proceedings, and authoritative books published within the last five years. Data were analyzed using a descriptive-analytical technique that synthesizes findings across planning, organizing, leading, and controlling functions in relation to resource-based theory. The results indicate that management functions act as the primary mechanism converting available resources into valuable, rare, and difficult-to-imitate capabilities that drive competitive advantage. Organizations that apply structured planning, clear organizing, participative leadership, and continuous controlling demonstrate significantly higher levels of productivity, employee engagement, and adaptability to environmental change than organizations with weak managerial practices. The study also finds that resource management and organizational effectiveness are mutually reinforcing, since effective management improves resource allocation while well-managed resources strengthen managerial capacity to respond to future challenges. It is concluded that management is not merely an administrative function but a strategic capability that must be continuously developed through training, information systems, and evidence-based decision making. These findings provide practical implications for managers seeking to strengthen organizational performance and offer a foundation for future empirical research using quantitative methods across diverse organizational contexts.

Feature Selection-Based Support Vector Machine Optimization for Heart Disease Risk Classification

2026-06-27T07:35:12+00:00

Heart disease is a serious health issue that requires rapid and accurate detection and risk classification. This study aims to build a heart disease risk classification model using a Support Vector Machine optimized with Chi-Square Feature Selection. The dataset used is the EarlyMed Heart Disease Risk Dataset, which consists of 70,000 initial data points, 18 input features, and one classification target, namely Heart_Risk. The research stages include data preprocessing, removal of duplicate data, splitting the data into training and testing sets with an 80:20 ratio, building a Standard SVM model, and building an Optimized SVM model using MinMaxScaler, Chi-Square SelectKBest, and GridSearchCV. The results show that the Standard SVM model achieved an accuracy, precision, recall, and F1-score of 99.14% using all 18 features. Meanwhile, the Optimized SVM model achieved an accuracy, precision, recall, and F1-score of 98.17% using 11 selected features. Although the performance of the optimized model is slightly lower, it reduces the number of features by 38.89% while maintaining high classification performance. Thus, Chi-Square Feature Selection can be used to produce a more compact and efficient model for classifying heart disease risk.

Classification Analysis of Chronic Kidney Disease Using the Naive Bayes Algorithm to Support Early Detection

2026-06-27T07:31:18+00:00

Chronic Kidney Disease (CKD) is one of the non-communicable diseases with an increasing prevalence and has become a serious problem in the healthcare sector due to delayed diagnosis and lack of early detection. This study aims to analyze the performance of the Naive Bayes algorithm in classifying chronic kidney disease to support early detection systems. The research used a Chronic Kidney Disease dataset consisting of 400 patient records with 26 medical attributes, including blood pressure, albumin, hemoglobin, blood glucose, and other health indicators. Data processing was carried out using Python with several stages, including data cleaning, missing value checking, categorical data encoding using LabelEncoder, and data normalization using StandardScaler. The dataset was divided into training data and testing data with a ratio of 80:20, resulting in 320 training data and 80 testing data. The classification process was performed using the Naive Bayes algorithm and evaluated using accuracy, precision, recall, f1-score, and confusion matrix. The experimental results showed that the proposed model achieved an accuracy of 97.5%, indicating that the Naive Bayes algorithm is capable of classifying chronic kidney disease effectively. The findings of this study demonstrate that machine learning techniques can support early detection systems for chronic kidney disease and assist healthcare services in improving diagnostic accuracy.

Hyperparameter Optimization of Logistic Regression Using GridSearchCV for Customer Churn Prediction

2026-06-27T07:40:05+00:00

Customer churn refers to a condition in which customers discontinue using a company's services, potentially resulting in decreased revenue and customer loyalty. Customer churn prediction on the Telco Customer Churn dataset was performed using the Logistic Regression algorithm optimized through GridSearchCV. The research process consisted of data preprocessing, categorical attribute transformation, feature scaling, splitting the dataset into training and testing sets, baseline model development, hyperparameter optimization, and model evaluation using accuracy, precision, recall, F1-score, confusion matrix, ROC curve, and AUC score metrics. The evaluation results showed that the baseline model achieved an accuracy of 81.61%, precision of 68.25%, recall of 55.13%, F1-score of 60.99%, and an AUC score of 86.00%. Meanwhile, the optimized model using GridSearchCV achieved an accuracy of 75.59%, precision of 52.05%, recall of 81.41%, F1-score of 63.50%, and an AUC score of 86.00%. The increase in recall from 55.13% to 81.41% and F1-score from 60.99% to 63.50% indicates that hyperparameter optimization improved the model's ability to identify customers who are likely to churn. However, these improvements were accompanied by decreases in accuracy and precision, indicating a trade-off between churn detection capability and positive prediction accuracy. Overall, the optimized model is considered more suitable for customer churn prediction tasks that prioritize the identification of customers at risk of churning

Explainable Machine Learning Approach for Burnout Risk Classification in Technology Workers

2026-06-27T07:50:17+00:00

Burnout among technology workers has become a significant occupational health concern due to increasing workload demands, rapid technological changes, and high-performance expectations. This study proposes an explainable machine learning framework for burnout risk classification using mental health and workplace indicators. The dataset consists of 3,000 samples derived from a larger mental health dataset, incorporating psychological variables such as stress, anxiety (GAD-7), and depression (PHQ-9), as well as workplace-related factors including work hours, sleep duration, work-life balance, social support, and job satisfaction. A balanced random sampling approach was applied by selecting 750 samples from each burnout class. The data were split into training and testing sets using an 80:20 stratified approach. Three machine learning algorithms Logistic Regression, Random Forest, and Extreme Gradient Boosting (XGBoost) were evaluated using accuracy, F1-score, and ROC-AUC with default hyperparameter settings. Feature importance analysis was applied to enhance model interpretability. The results show that Logistic Regression achieved the highest accuracy of 88.2% with an F1-score of 0.88 and the best discriminative performance with an AUC of 0.96, consistently outperforming Random Forest and XGBoost across all evaluation metrics. The findings indicate that psychological factors, particularly stress, anxiety, and depression, are the most influential predictors of burnout risk. Overall, the proposed framework demonstrates strong predictive performance and interpretability, making it suitable for early burnout detection in technology work environments.

Comparative Study of Ensemble Learning Models for Blood Cell Anomaly Detection

2026-06-27T07:45:29+00:00

Blood cell anomaly detection plays a crucial role in the early diagnosis of hematological disorders, including anemia, leukemia, and other blood-related diseases. Accurate identification of abnormal blood cells can support timely clinical decision-making and improve patient outcomes. However, selecting the most effective machine learning model for blood cell anomaly detection remains challenging due to variations in data characteristics and model performance. This study aims to compare the effectiveness of three ensemble learning models, namely Random Forest, Extreme Gradient Boosting (XGBoost), and Light Gradient Boosting Machine (LightGBM), for blood cell anomaly detection. The study utilized a publicly available blood cell anomaly dataset consisting of 5,880 blood cell records with morphological, staining, and hematological features. Data preprocessing included feature selection, categorical encoding, and train-test splitting using an 80:20 ratio. Model performance was evaluated using Accuracy, Precision, Recall, F1-Score, Area Under the Curve (AUC), and Matthews Correlation Coefficient (MCC). Experimental results showed that LightGBM achieved the best performance with an Accuracy of 98.13%, Precision of 98.10%, Recall of 96.01%, F1-Score of 97.04%, AUC of 99.83%, and MCC of 0.9569. Feature importance analysis further revealed that granularity score, cell diameter, and cell area were the most influential features. These findings indicate that LightGBM is a highly effective approach for blood cell anomaly detection and can support the development of intelligent hematology diagnostic systems.