International Journal of artificial intelligence research (IJAIR)
Not a member yet
271 research outputs found
Sort by
Performance of Deep Face Recognition Models under Adaptive Margin Loss: A Real-Time Evaluation
Full text linkReal-time face recognition systems encounter a critical trade-off between high-security demands and computational efficiency, particularly when deployed in unconstrained open-set environments. This study presents a comprehensive benchmarking of four distinct deep learning backbones ResNet100, GhostFaceNet, LAFS, and TransFace specifically trained using the Adaptive Margin Loss (AdaFace) function to handle image quality variations. The primary objective is to identify the optimal architecture for secure attendance systems operating on standard hardware with limited training data. The evaluation protocol employs a rigorous real-world open-set test to quantify performance using False Acceptance Rate (FAR) and False Rejection Rate (FRR). The experimental results demonstrate that ResNet100 establishes the highest security standard, achieving a 0.00% FAR at strict thresholds. Meanwhile, GhostFaceNet emerges as the most balanced solution for resource-constrained deployments, delivering competitive accuracy above 93% with significantly lower computational complexity. Conversely, the Vision Transformer (TransFace) fails to generalize in this low-data regime, resulting in unacceptable false acceptance rates. These findings definitively recommend GhostFaceNet for efficient edge-based implementations, while ResNet100 remains the superior choice for mission-critical security applications
ENHANCING COMMUNITY HEALTH CENTER PERFORMANCE THROUGH PUBLIC SERVICE INNOVATION AND DIGITAL TRANSFORMATION STRATEGIES
Full text linkThe performance of the Community Health Center (Puskesmas) has been in the spotlight of all parties concerned with its ability to adopt the right innovation strategy and implement appropriate digital transformation. This research aims to uncover (1) how public service innovation strategies impact digital transformation implementation; (2) their effect on Puskesmas performance; (3) the impact of digital transformation implementation on Puskesmas performance; and (4) the mediating role of digital transformation in this relationship within West Java, viewed from a strategic management perspective. This quantitative study was conducted across 285 from 1,098 Puskesmas in West Java Province, utilizing proportional random sampling. The sample consisted of Puskesmas distributed across cities and districts in the region, ensuring representation from each area. A questionnaire with a 5-point rating scale was the primary research instrument, validated for accuracy and reliability. Data analysis included descriptive categorization and inferential analysis using Partial Least Square (PLS) via SmartPLS. The study found that public service innovation strategies positively impacted digital transformation implementation and Puskesmas performance. Additionally, digital transformation implementation positively influenced Puskesmas performance, with digital transformation partially mediating the relationship between innovation strategies and performance
Factors Influencing Customer Satisfaction and Their Impact on Customer Loyalty
Full text linkWorkshop customers, known as unit entry by the automotive world, are one of the sources of revenue for automotive companies. Workshop unit entry is related to customer loyalty. The study aims to determine the relationship between perceived price on customer loyalty through customer satisfaction in auto repair shops customers. A quantitative approach is used as a research approach. Data were collected by questionnaire. The collected data were analyzed to determine the effect between variables, and the analysis technique was Partial Least Square (PLS). The research was conducted at Auto2000 West Sumatra. The research subjects are Auto2000 consumers who use workshop services. The object of research is a review of the influence of perceived price on customer loyalty with customer satisfaction as mediation. The sampling technique is purposive sampling. The results showed that Customer satisfaction can mediate the effect of perceived price on customer loyalty of consumers who use Auto2000 workshop services.
Omnichannel Customer Experience Model towards Customer Repurchase Intentions and Word of Mouth on Cellular Products Loyalty
No full textCustomer loyalty to mobile products in Indonesia is influenced by various factors, including download speed, upload internet speed, latency, streaming, and browsing. This study aims to analyze the effect of Omnichannel Customer Experience (OCX) on Cellular Product Loyalty (CPL) through the mediation role of Customer Repurchase Intentions (CRI) and Word of Mouth (WoM). This study uses an explanatory survey design involving 384 mobile product users throughout Indonesia, who were selected using proportional sampling techniques. Data analysis was carried out using descriptive and verification methods, using Structural Equation Modeling - Partial Least Square (SEM-PLS) to test the relationship between variables. The findings of the study indicate that OCX does not have a direct effect on CPL, but must go through the mediation of CRI and WoM. Between the two mediators, WoM has a stronger influence than CRI, indicating that a positive omnichannel experience encourages customers to recommend products, ultimately increasing customer loyalty. The novelty of this research lies in simultaneously testing the relationship between OCX, CRI, and WoM on CPL in the context of the cellular industry, as well as identifying the more dominant mediation pathways in increasing customer loyalty. The implications of this research provide insights for telecommunications companies to focus more on omnichannel strategies that can improve customer experience and encourage Word of Mouth, as well as for the government in designing policies that support the digital ecosystem for the cellular industry in Indonesia
GLCM Texture Feature Selection for Alzheimer's Detection: A Combination of Statistical Tests, Decision Tree, and Random Forest
No full textAlzheimer’s disease is a progressive neurodegenerative disorder that leads to a gradual decline in memory and cognitive function, most commonly affecting individuals over the age of 65. Early detection is essential to enable timely interventions, slow disease progression, and improve quality of life. This study aimed to identify the most dominant texture features from brain MRI images using the Gray Level Co-occurrence Matrix (GLCM) for feature extraction. The extracted features were analyzed through non-parametric statistical tests and machine learning algorithms, including Decision Tree and Random Forest, and validated with cross-validation procedures to ensure robustness. The findings revealed that contrast at 90° consistently emerged as the most significant feature, capturing vertical texture variations associated with brain atrophy, while correlation at 135° provided additional discriminatory power by representing disrupted pixel intensity relationships. In combination, these features enhanced the accuracy of classification models, outperforming other GLCM parameters. The results emphasize that careful selection of texture features improves both accuracy and stability in distinguishing between Alzheimer’s and non-Alzheimer’s brains. This study demonstrates that image-based machine learning frameworks can serve as reliable tools to support early detection of Alzheimer’s disease, offering valuable implications for clinical practice and guiding future research on efficient, non-invasive diagnostic approaches
Performance Improvement Analysis of Design and Build Construction Project Managers of State Builidngs
Full text linkProject managers of planning and construction of government buildings are experts of the implementing contractor who determine the timeliness of the implementation of design and build construction. In order for timely implementation, project managers not only have higher education and long experience, but must have a good work culture and work behavior as well. Therefore, it is necessary to examine project performance based on the work performance of project manager
Auditor Behavior Toward Artificial Intelligence (AI) Development: A Literature Review
No full textThe integration of Artificial Intelligence (AI) has fundamentally reshaped the auditing profession by enhancing efficiency, analytical precision, and audit quality. This study aims to explore auditors’ behavioral responses toward AI adoption, emphasizing individual, organizational, and ethical dimensions. Using a systematic literature review approach of scholarly articles published between 2020 and 2025, the paper synthesizes findings from contemporary research focusing on the drivers, impacts, and governance challenges of AI implementation in auditing. The results show that willingness to learn, performance expectancy, and AI readiness are the most influential individual factors determining auditors’ intention to adopt AI, while top management support and technological infrastructure play key organizational roles. AI-based systems significantly improve anomaly detection accuracy and operational efficiency; however, algorithmic bias, limited transparency, and accountability gaps remain critical ethical concerns. The study also highlights that AI cannot replace human professional judgment—ethical reasoning, contextual interpretation, and moral accountability must remain central to the audit process. Future auditors are expected to evolve from compliance examiners to strategic advisors equipped with multidisciplinary competencies in data analytics, digital governance, and ethics. Therefore, the synergy between human insight and AI-driven analytics is essential for ensuring trustworthy, transparent, and sustainable audit practices in the digital era
Governance-Based SPBE Model Framework Design in Multi-Level Governmentl
No full textThis research aims to develop a governance-based framework model for the implementation of the Electronic-Based Government System (ESG) in Indonesia. Although ESG is a national strategic agenda, its implementation still faces various challenges, such as institutional fragmentation, low interoperability, minimal public participation, and weak adaptation to technological change. Through a qualitative approach with a literature review and conceptual construction, this research formulates four interrelated governance models: the Transparent Digital Governance Framework, the Collaborative ESG Ecosystem Model (CSEM), the Participatory E-Governance Model (PEGM), and the Adaptive ESG Governance Framework (ASGF). Each model is built on the governance principles of transparency, collaboration, participation, and adaptability as a response to structural weaknesses in current ESG practices. The results show that effective and sustainable digital governance requires the synergy of these four principles in an integrated and contextual manner, particularly to bridge the gap between national policy and local implementation
E-GOVERNMENT IN INNOVATION AND PUBLIC COMMUNICATION
No full textPublic communication has undergone major transformation since the early stages of digitalization of services as part of innovation. One of the main digital transformations in public communication, especially in government systems, is e-government. This study aims to discuss what e-government really is, what research is related to e-government, and what the practice of e-government applications is. This research is a library study research by taking documentation data about innovation and public communication from various sources. The data presented is in the form of the latest studies regarding innovation, especially in public communication. The data collected in this research will then be analyzed using narrative analysis. Narrative analysis refers to a set of methods for interpreting texts that take the form of exposition. The conclusions in this study show that globally, the implementation of e-government throughout the world still has many challenges, especially in Indonesia. Therefore, many improvements still need to be made and this also requires further study about what and how to improve innovation in e-governmen
Factors Influencing Sharia Property Purchase Decisions: Tasnim Property Case Study
No full textStudy This aiming For analyze influence product, price,location, promotion, word of mouth, style life, technology and information to interest buy and decision purchase Sharia property at Tasnim Property, as well as role interest buy as intervening variable. Research done use approach quantitative with voluntary sampling technique. A total of 195 respondents who had do survey to Tasnim Property office fills online questionnaire. Data was analyzed using SEM-PLS version 4. Research results show that variable product, price, promotion,word of mouth, style life and technology influential significant to interest buy. However, the location and information No influential significant. Product, price,promotion, and word of mouth also have an influence. significant to decision purchase, while location, style life,technology and information no. Interest in buying own influence significant to decision purchase. In No direct,variable product, price, promotion, word of mouth, style life,and technology also influences decision purchase through interest buy. Implications managerial from study This that is the need for segmentation, targeting, and positioning (STP) strategies and optimization influential variables significant to decision purchase, namely products, prices, promotions, and word of mout