Publication Search

In the era of data-driven digital commerce, Artificial Intelligence (AI)-based product personalization has become a key strategy to enhance user experience and foster customer loyalty. However, in the Indonesian e-commerce landscape, there remains a lack of empirical understanding of how personalization systems influence long-term user engagement. This study investigates the impact of AI-driven product personalization on customer loyalty among Indonesian e-commerce users. Employing a mixed-methods approach, quantitative data were collected through an online survey of 150 active users, and qualitative insights were obtained from in-depth interviews with six informants. Statistical analysis using simple linear regression revealed that personalization significantly influences customer loyalty, with a beta coefficient of 0.653 (t = 8.241, p < 0.001) and an R² value of 0.567. Qualitative findings highlight user concerns over recommendation accuracy, interface overload, and repetitive suggestions, which affect emotional satisfaction and platform attachment. This research contributes to the growing body of knowledge on AI adoption in e-commerce by integrating behavioral and technological dimensions of loyalty formation. It also offers practical implications for designing more context-sensitive personalization systems that prioritize not only algorithmic precision but also user control and experience quality.

Dengue Fever (DF) continues to be a major public health threat in Indonesia, especially in urban areas with high population density, such as Purwokerto City. This study aims to develop a predictive model to identify high-risk areas for DF outbreaks by integrating Machine Learning (ML) algorithms and Geographic Information Systems (GIS). The research utilizes historical dengue case data, meteorological parameters (rainfall, temperature, humidity), and population density as predictive variables. Three ML classification algorithms—Naïve Bayes, Logistic Regression, and Support Vector Machine (SVM)—were implemented to develop risk prediction models. Extensive data preprocessing, feature selection, and spatial integration were applied to ensure model robustness. The results show that the SVM model outperformed other methods, achieving the highest accuracy, precision, recall, and F1-score in classifying dengue risk zones. Risk maps generated through GIS visualization successfully identify priority areas for targeted interventions. The novelty of this research lies in the combination of local epidemiological data, multi-algorithm comparison, and geospatial mapping to improve early warning systems for DF in Purwokerto. This integrated approach is expected to support more effective prevention strategies and enhance public health preparedness.

TikTok has transformed into an effective digital marketing platform, not only for introducing products but also for shaping impulsive shopping behavior. This study aims to analyze the impulsive shopping behavior of Generation Z on TikTok Shop. The research was conducted using a qualitative descriptive approach, with data obtained from digital searches and observations. The findings indicate that Generation Z has a strong interest in utilizing the TikTok Shop feature within the TikTok application, making it one of their preferred alternatives for online shopping transactions. TikTok significantly influences Generation Z’s shopping behavior, as they tend to purchase products after seeing them on the platform. Businesses leverage TikTok as an efficient marketing tool, particularly through creative content, short videos, and influencer recommendations. By utilizing an algorithm that tailors content to users’ preferences, TikTok Shop creates offers that appear attractive to consumers. Generation Z is often influenced to buy items they do not actually need. Sudden purchases driven by emotions—such as low prices or limited-time offers—often lead them to overlook product quality. As a generation living in the digital era, Gen Z needs to better regulate their impulsive consumer behavior. Wisely utilizing technology and understanding the marketing strategies used by platforms like TikTok Shop can help them avoid excessive impulsive spending. By prioritizing needs over wants and carefully considering the value and benefits of a product before purchasing, Gen Z can become smarter and more responsible consumers.

Learners are continually faced with choosing appropriate courses or making career choices due to increased educational opportunities. The emergence of machine learning-based course and career recommender systems has the potential to address this issue, offering personalized course recommendations tailored to individual learning pathways, preferences, and learning history. The optimization and feature engineering techniques and practical deployment environments have not been collectively examined in the previous research, despite the significant advancements in this area of research. Furthermore, previous research has rarely synthesized how these technical components help students choose appropriate courses and careers. This systematic review was carried out to investigate the current state of machine learning-based course and career recommender systems, focusing on key elements, such as primary data sources, feature engineering methods, algorithms, optimization techniques, evaluation metrics, and the environments where the existing course recommendation models are deployed. The PRISMA method for conducting a systematic review was used to choose studies that met the requirements for inclusion and exclusion. The study findings show significant reliance on interpretable and traditional machine learning algorithms, such as K-Nearest Neighbor and Random Forest, to develop recommender models. Feature engineering remains basic, as most studies rely on normalization, while optimization processes are often underreported. Also, evaluation metrics varied widely, impeding comparability, while most of the recommender models are deployed in an e-learning environment, leaving the traditional learning environment underrepresented. Furthermore, the study findings identified issues including data sparsity and diversity, data security and privacy, and changes in learner preferences that may have an impact on the performance of recommender systems while recommending further studies to make use of standardized optimization methods, and automated domain-informed feature engineering frameworks, benchmark and annotated datasets in developing models the gives priority to learners’ success and educational relevance.

The rapid advancement and integration of Artificial Intelligence (AI) into diverse sectors of society have generated complex ethical and human rights challenges. Technologies involving surveillance, data collection, algorithmic decision-making, and facial recognition pose significant risks to privacy, equality, and freedom of expression. This study examines the intersection of AI and human rights through a comparative analysis of regulatory frameworks in the European Union (EU), the United States (US), and Asia. Employing a comparative legal approach, the research analyzes international and national regulatory instruments, including the EU AI Act, the General Data Protection Regulation (GDPR), and China’s Personal Information Protection Law (PIPL). Case studies of AI-related human rights violations, such as algorithmic bias and discrimination, are incorporated to illustrate real-world implications. Findings reveal substantial differences in governance approaches: the EU emphasizes a risk-based model prioritizing human rights protections, while the US and Asia adopt more fragmented or centralized strategies. The study underscores the urgent need for global regulatory harmonization to safeguard fundamental rights and promote ethical AI development. By highlighting both strengths and limitations of existing frameworks, the research contributes to ongoing debates on balancing innovation with accountability, transparency, and human rights protection in the digital era.

This study investigates how artificial intelligence (AI) is transforming Islamic journalism and digital da’wah by analyzing the ethical challenges posed by algorithmically mediated media. Employing a critical-interpretive approach, the research examines how algorithmic logic impacts the production, visibility, and ethical dimensions of Islamic content on platforms such as YouTube and TikTok. The findings reveal a growing tension between Islamic communication principles—sidq (truth), amanah (trust), and hikmah (wisdom)—and the operational values of AI systems that prioritize virality, personalization, and engagement metrics. This paper contributes to global discourse by proposing a hybrid ethical framework that bridges Islamic communication ethics with digital media accountability. It argues for an integrative model where spiritual integrity, transparency, and justice guide the development and use of AI in Islamic media contexts.  

The rapid adoption of algorithmic systems in public governance has transformed decision-making and service delivery, offering promises of efficiency and transparency. Yet, these technologies raise pressing concerns regarding fairness, bias, and social justice. This study investigates the intersection of digital governance, algorithmic decision-making, and social justice, with particular emphasis on emerging democracies. Employing a qualitative socio-legal approach, the research combines normative analysis of governance regulations, case studies of algorithmic applications in public administration, and interviews with policymakers and technology law experts. Comparative analysis across emerging democracies highlights diverse strategies for addressing equity concerns in algorithmic systems. Findings reveal that while algorithmic systems enhance efficiency, they often reinforce existing inequalities due to insufficient safeguards against bias and discrimination. Moreover, regulatory frameworks remain fragmented and inadequate to ensure fairness and accountability. The study proposes the development of adaptive legal frameworks that integrate transparency, accountability, and citizen engagement into AI governance. By embedding social justice principles into algorithmic regulation, governments can foster inclusive policy design and equitable outcomes. This research contributes to ongoing debates on balancing technological innovation with democratic values, emphasizing the need for governance models that prioritize fairness alongside efficiency.

This research uses a qualitative method to examine the influence of emotional copywriting on user engagement in algorithm-curated social media platforms such as Instagram Reels and TikTok’s For You Page (FYP). The increasing consumption of short-form, personalized content driven by user behavior highlights the need for more targeted communication strategies. Emotional copywriting—writing that emphasizes affective elements such as empathy, joy, fear, or nostalgia—is considered effective in creating emotional bonds between content creators and audiences, thus encouraging interactions like likes, comments, and shares. This study was conducted through content analysis of 25 viral videos with high engagement, along with in-depth interviews with five active content creators. The findings reveal that using emotionally resonant phrases aligned with users’ social values and experiences significantly boosts content performance within algorithmic systems. Emotionally charged content tends to receive broader exposure as algorithms prioritize highly interactive posts. Therefore, emotional copywriting is not only a creative strategy but also an effective tactic for optimizing content visibility in the age of social media algorithms.   Keywords: , , , , , Reels, ,  social platforms.

This research analyzes user sentiment reviews of the KAI Access application from Google Play Store to improve customer service at PT Kereta Api Indonesia. The study uses a Natural Language Processing (NLP) approach with the Latent Dirichlet Allocation (LDA) algorithm to extract main topics from 10,000 reviews collected from April 2024 to April 2025. Analysis results show 40.7% positive sentiment reviews and 49.3% negative. After data preprocessing through case folding, normalization, tokenization, stopword removal, and stemming, seven optimum topics were found from negative sentiment with a coherence score of 0.508343 and two optimum topics from positive sentiment with a coherence score of 0.511673. Analysis based on five service quality dimensions (tangibles, reliability, responsiveness, assurance, and empathy) reveals that the reliability dimension becomes the main issue, including system instability, transaction failures, login difficulties, and data inaccuracy. The responsiveness dimension is the second priority, with users expecting fast and responsive service to complaints. The results of this study provide recommendations for PT KAI to prioritize improvements in system reliability and responsiveness aspects to enhance the overall user experience, which will ultimately impact customer satisfaction and loyalty.    

The evolution of social media and its algorithms has transformed the dissemination of Pancasila as Indonesia’s national ideology. This study aims to analyze the representation of Pancasila in algorithmically curated digital discourse and its impact on public understanding. A qualitative approach, combining digital discourse analysis and netnography, was applied to examine content on Twitter, Facebook, Instagram, and YouTube. Findings reveal the dual role of social media algorithms: while facilitating educational content on Pancasila (e.g., religious tolerance, human rights campaigns, and nationalism), they also amplify polarizing content, hoaxes, and hate speech that contradict Pancasila values. Analysis of Pancasila’s five principles shows that conflict-driven content (e.g., ethnic-religious issues) gains higher virality due to algorithms’ prioritization of engagement metrics. The filter bubble and echo chamber phenomena exacerbate discourse fragmentation, hindering inclusive dialogue about national ideology. The study concludes that Pancasila-based digital literacy, algorithmic transparency, and multistakeholder collaboration are critical to optimizing social media’s role in strengthening national identity. Policy recommendations include integrating "social cohesion" parameters into algorithm design and regulating content to uphold diversity. 

This research investigates the implementation of Random Forest algorithms combined with Synthetic Minority Over-sampling Technique (SMOTE) to predict elementary school dropout rates in Indonesia, supporting the Indonesia Emas 2045 vision. A significant gap was identified in previous studies, which, despite utilizing artificial intelligence for dropout interventions, had not integrated temporal dimensions into data analysis. A temporal data-based classification model was developed using Indonesian Ministry of Education data from 2021-2023, incorporating lag features, delta calculations, and rolling statistics. Two models were implemented: one with SMOTE achieving 99% accuracy with perfect recall for high-risk regions, while the non-SMOTE model reached 100% accuracy. Temporal features were identified as crucial predictors, reflecting external fluctuations and annual changes impacting dropout decisions. This approach enables educational institutions to allocate resources more efficiently by prioritizing operational assistance for high-risk schools. The model's capacity to identify high-risk regions with 100% recall represents a strategic investment in strengthening Indonesia's human resource sustainability. To address the limitations of provincial aggregate data, expansion to include individual-level variables and model validation at district or school scales is recommended for future research.

The increasingly rapid development of the culinary sector makes commercial competition in this field increasingly fierce. Food stalls serve a variety of menus and drinks, but stall owners need to strive to create product innovations in order to provide satisfactory service to customers. Under these conditions, a computer technician is needed to find out recommendations for food stall menus. The analysis method used is a data mining technique with the Apriori algorithm, where this algorithm is used to identify the data sets that appear most frequently (frequent itemset). The research results show that the highest support and confidence values ​​are Ayam Penyet and Fried Rice with 50% support and confidence values 76%. This can be a combination of menus suggested from data that has been collected and applied according to an a priori algorithm which is expected to be used to evaluate service and possibly increase guest satisfaction so that the Food Stall can develop more quickly.

The rapid development of multimedia technology has significantly advanced 3D animation techniques, enabling the production of high-quality visual content across industries such as film, gaming, architecture, and product visualization. Rendering, as the final stage of the 3D production pipeline, plays a crucial role in determining both visual realism and production efficiency. This study compares the performance of three rendering engines—Eevee, Cycles, and Radeon ProRender—by evaluating rendering speed, visual quality, and memory efficiency in Blender. The objective is to provide practical insights for designers and digital content creators in selecting the most suitable rendering engine based on project requirements. In this research, three identical 3D scenes were rendered using each of the three rendering engines under controlled experimental conditions. The comparison was conducted based on several parameters, including rendering time, output file size, shadow accuracy, lighting effects, and overall visual realism. Quantitative measurements were used to evaluate render speed and memory consumption, while qualitative analysis assessed differences in shadow detail, global illumination behavior, reflection accuracy, and material realism. The results indicate that Eevee outperforms the other engines in terms of rendering speed, making it highly suitable for real-time applications and projects requiring fast previews. Cycles produces the highest level of visual realism due to its physically based path-tracing algorithm, although it requires longer rendering time and higher computational resources. Meanwhile, Radeon ProRender demonstrates competitive performance, particularly in shadow quality and lighting effects, offering a balanced alternative between realism and efficiency. Based on the findings, Blender remains a flexible and effective platform. The choice of rendering engine should depend on whether speed, graphic quality, or memory optimization is prioritized.

Live streaming on social media has changed the way people share information and entertainment, providing greater convenience and access than conventional broadcasting. However, this change also brings challenges, especially regarding ethics and its impact on young audiences. Content on social media is often poorly monitored, making it easy to find sensational or inappropriate content, such as violence, hoaxes and unethical behavior. This is exacerbated by platform algorithms that prioritize content that triggers emotional reactions to increase popularity. Conventional broadcasting has strict rules to maintain the quality and ethics of content, while live streaming on social media often relies solely on platform policies which are not always effective. As a result, many content creators try to attract attention in extreme ways, which can influence the behavior and mindset of young audiences. This research discusses the differences between conventional broadcasting and live streaming on social media, as well as the importance of the responsibilities of platforms and content creators in maintaining ethics. Apart from that, digital literacy for young audiences is needed so that they can be more critical in choosing content. In conclusion, cooperation between content creators, social media platforms and society is needed to ensure social media becomes a safer and more beneficial place for all.

The application of the theory of truth in philosophy can make an important contribution to the strategy for eradicating hoaxes in the digital era. Hoaxes, as misinformation and often deliberately spread for certain purposes, have a broad negative impact, especially in the formation of public opinion and social stability. In this context, theories of truth such as correspondence, coherence, and pragmatism can be used to evaluate and identify the truth of information spread in cyberspace. Correspondence theory, for example, can be used to confirm whether information corresponds to objective facts. Coherence theory, on the other hand, prioritizes the consistency of information with a wider system of knowledge, while pragmatism theory assesses truth based on the practical impact of receiving information. The application of these theories in a hoax eradication strategy involves media education, increasing digital literacy, and developing information verification algorithms that prioritize the principles of philosophical truth. With this approach, it is hoped that the public will be smarter in filtering information, so that the spread of hoaxes can be minimized and public trust in digital media can be restored.

PT Astra International BMW Semarang is a company operating in the automotive sector with 3 supporting pillars, namely Sales, Aftersales and Spare Parts for BMW car units. The availability of spare parts is one of the determining factors for consumer satisfaction with the company because if the spare parts stock is empty it will cause consumer disappointment with the company. By using spare parts sales transaction data for the period January 2019 – June 2023, totaling 52,162, it will be utilized using data mining association techniques with the a priori algorithm and the eclat algorithm. The problem in this research is how to find out consumer purchasing patterns so that there is no shortage or empty stock of spare parts in the warehouse. This research aims to determine the association of spare parts purchasing patterns in sales transactions so that partman get recommendations in making decisions about providing priority types of spare parts. This research methodology uses CRISP-DM (Cross-Industry Standard Process for Data Mining) and is implemented with the R programming language with R studio software. In 3 trials using the Apriori algorithm and 3 trials with the Eclat algorithm, The result with the highest confidence appears in a combination of 3 itemsets with minimum support 0.01 and confidence 0.9, namely if a customer buys B11.42.8.593.186 (Set oil-filter Mx) and B83.12.5.A1A.683 (Washer Cleaner) then they will also buy Z99000000333 ( BMW Engine Oil) with confidence 1.00 or 100%. From the results of this association's analysis, it can be used as advice for the management of PT Astra International BMW Semarang in managing spare parts stock.

This research aims to address the issue of uneven bandwidth distribution in large organizational networks by implementing Quality of Service (QoS) using FIFO and the Hierarchical Token Bucket (HTB) algorithm on Mikrotik routers. Uneven bandwidth distribution can disrupt productivity and operational efficiency. This study creates a fair and efficient traffic management system, allowing bandwidth allocation according to user needs. The methodology involves detailed configuration of Mikrotik RouterOS to optimize QoS with adjusted HTB settings. Testing was conducted using IPerf3 to measure bandwidth variations received by clients in different conditions, including scenarios with two and three active clients. The results indicate that the HTB method provides more stable and consistent bandwidth distribution compared to FIFO. In the two-active client scenario, the unused bandwidth by the third client is allocated to higher priority clients, demonstrating HTB's effectiveness in managing traffic priorities. This research is expected to enhance user satisfaction by providing a network that is both stable and responsive to the needs of various operational applications, and contribute significantly to the development of best practices for bandwidth management in complex organizational environments.

Police Sector (Polsek) is one of the agencies that provide protection, order and ensure public safety in the sunggal area. The number of cases of criminal acts that occur makes residents feel unsafe and always feel threatened in certain areas in the Sunggal sub-district, the pattern of criminal acts that often occur due to several factors, one of which is due to the lack of security in the area so that many criminal acts occur as well as behaviour that has been planned by the perpetrator to achieve their goals by planning, preparing, implementing, disposing of evidence, even hiding or escaping depending on the type of crime committed based on the characteristics of the perpetrator, and the situation or context in which the crime occurred. Therefore, it is necessary to analyse techniques from existing criminal data using the a priori algorithm method to find patterns of relationships between variables that can assist agencies in taking action for public safety. Based on the research conducted, the above case is tested with a minimum support = 10%, confidence = 100% so that the results of the rule that meets the support and confidence values are obtained: ‘If the criminal act is theft then the job is self-employed’, then giving value is successful with 15% support, 100% confidence. And ‘If the age of 17-25 years, the criminal act is Theft then the job is unemployed’, then giving value is successful with 10% support, 100% confidence.

Student behavior is the actions of students which are influenced by their attitudes and responsibilities at school. Student behavior is a very important factor in determining student achievement in learning. Students who have personalities that improve skills, knowledge, attitudes, habits, understanding, skills, thinking power and other abilities will more easily increase their concentration in learning, and it will be easier to achieve the students' goals. At SDN 053960 MARYKE there are still students who do not know that student behavior greatly influences their level of achievement. Therefore, it is necessary to educate students from an early age so that students can be more responsible for the rules given by teachers at school, and students must understand that the attitude they carry out at school is assessed in improving their achievement, as well as their presence is very influential. his level of achievement at school. Therefore, there is a need for a solution to overcome the problems that exist at SDN 053960 MARYKE by utilizing data mining to collect data and then it will be processed using the a priori method with variables contained in the correlation between student behavior and student achievement levels. The a priori algorithm is able to determine min support and confidence in these variables will later show the relationship between student behavior and student achievement levels, so that researchers will get the best best rules and be able to produce the latest information..

Zakat fitrah is an obligation for every capable Muslim to purify oneself and one's wealth. Determining the priority of zakat fitrah recipients often becomes a challenge due to various factors that must be considered. This research aims to use the Fuzzy Simple Additive Weighting (F-SAW) algorithm as a decision support tool in determining the priority for distributing zakat fitrah. By using the F-SAW method, it is hoped that the prioritization process can be more objective and efficient..