This study examines research trends regarding the prediction of academic achievement using machine learning. Research in the field of academic achievement is currently continuing to develop, but has not been explored comprehensively in a bibliometric context. The visualization provided includes a map of publication development using machine learning methods based on country, analysis of bibliographic pairs and keywords used. To find out the visualization results, bibliographic analysis was used using VOSviewer. The data used in this analysis were 76 articles collected from the Scopus database from 2018-2023. From the results of the analysis, it is known that research related to academic achievement still shows a growing trend in publications in the field of discussion of factors or predictors that influence academic achievement as well as research that proposes or evaluates models for predicting academic achievement. The research results show that although machine learning techniques such as Random Forest and Support Vector Machine are often used in academic achievement prediction research. Future research could consider developing a more adaptive and comprehensive approach regarding the contribution of specific factors that influence the accuracy of more in-depth prediction models in this field.
Websites have become essential platforms in the Higher Education Institution (HEI) system, enabling flexible access to education and academic information. Therefore, HEI websites' Usability and User Experience (UX) have gained significant scholarly attention. This study aims to identify research trends evaluating usability and UX in HEI website design (2015- 2024) through bibliometric analysis of publications (n = 124) from Scopus, IEEE Xplore, and ScienceDirect. The scope of the analysis includes publication frequency, keyword networks, citation performance, journals, authors, institutional affiliations, and research clusters. The results showed that usability and UX evaluations are studied across multiple disciplines and are concentrated in five clusters: (1) E-learning website interfaces during the COVID-19 pandemic, (2) Library information systems, (3) Students' continuance intention on learning websites, (4) Marketing communication websites, and (5) Learning management systems. These results underscore that usability and UX evaluation on HEI website platforms leans towards student-centric studies encompassing some sites and service units. This paper also emphasizes the importance of user-centred evaluation throughout HEI websites’ design, development, and post-implementation phases. Given that HEIs must provide optimal services to stakeholders with diverse backgrounds, needs, and information access capabilities.
This study aims to analyse the factors influencing customer engagement in social commerce live streaming, specifically on the TikTok Shop platform, and their impact on purchase intentions and continuance usage intention. The significance of this research lies in the rapid development of social media applications within the social commerce sector, which presents new opportunities for brands to interact directly with users through live streaming features. A quantitative research approach was employed, utilizing survey techniques to collect data from respondents who are active users of the TikTok Shop platform in Indonesia and engage with its live streaming feature. The data were analysed using the Partial Least Squares Structural Equation Modelling (PLS-SEM) method. The results indicate that the hedonic, symbolic, and interactivity values of live streaming positively influence customer engagement, whereas utilitarian values have no significant effect. Moreover, customer engagement positively affects purchase intentions and continuance usage intentions. These findings underscore the critical role of emotional and interactive aspects of live streaming in fostering customer engagement, which subsequently enhances purchase intentions and sustained usage. This research offers valuable insights into the development of digital marketing and user interaction strategies in the context of social commerce.
Fake news detection has become a critical issue in the digital era, especially with the rapid growth of social media and online platforms. This research aims to enhance the accuracy of detecting fake news in Indonesian by developing a model using lexicon-based and Long Short-Term Memory (LSTM) approaches. The study integrates sentiment analysis with lexicon-based scoring to identify key features in news articles, while LSTM is employed to analyze sequential patterns in the data. The methods were tested on a dataset consisting of both hoax and non-hoax news collected from reliable sources. The results indicate that the hybrid model significantly improves the detection accuracy, achieving an impressive accuracy rate of 99%. This research demonstrates the potential of combining lexicon-based and LSTM approaches to overcome challenges in detecting fake news, especially in low-resource languages like Indonesian. The findings contribute to advancing the development of reliable and efficient systems for combating misinformation in the digital age.
Online Health Communities (OHCs) have become a key source of social support for individuals with health concerns. OHC members engage in communication and information exchange, with trust among members playing a crucial role in the acceptance of these platforms. This research aims to examine the determinants affecting OHC acceptance by employing trust transfer theory, social support, and self-efficacy as core variables. The proposed model was empirically tested using data from 100 members of the Indonesian Diabetes Forum on Facebook. This quantitative study employed a 5-point Likert scale to evaluate user perceptions. The findings indicate that OHC acceptance is significantly supported by both information support and emotional support, which foster trust among community members. Trust in members subsequently leads to trust in the broader community, culminating in the sustained use of the OHC. Furthermore, emotional support positively influences self-efficacy, encouraging users to join and actively participate in OHCs. However, information support does not have a significant effect on self-efficacy. This research offers significant understanding of the relationships among social support, self-efficacy, and trust in promoting the continued use of OHCs. The research model offers a framework that can be applied in other contexts with similar technological and community-based perspectives.
The management of building asset data has experienced many problems related to accessibility and storage issues. This causes difficulties for asset managers in managing and developing buildings after the construction phase. This research aims to develop a 7D Building Information Modeling (BIM) model of buildings to facilitate access to asset data in operational and maintenance activities. Autodesk Revit was used to perform BIM modeling and was integrated with COBIe and the cloud via QR code. The Faculty of Engineering Building Diponegoro University was used as a case study. Data was collected through observation, project documents, as-built drawings, technical specifications, etc. This 7D BIM model with COBIe plug-in is expected to address gaps in asset management, improve operational efficiency, reduce the risk of damage, improve the ability to classify assets in a systematic and integrated manner, and improve collaboration between stakeholders to increase the effectiveness of asset management significantly.
This disruptive era, tourist villages must adapt to technological advancements to innovate and drive the digitalization of these villages. Data processing, speed, and clarity of information can be utilized for the development of smart village tourism, where innovation and technology become the main drivers of transformation. The smart city concept can be adopted in the development of smart village tourism to enhance sustainability in tourist villages. A deep and ongoing study of the potential and local wisdom of rural communities is a key factor in the development of smart village tourism. This research serves as a preliminary study in the context of smart village tourism. The aim of this study is to formulate a development model for smart village tourism based on the identification and analysis of the barriers and facilitators in adopting the smart city perspective. The case studies involve two tourist villages in Boyolali Regency. The data analysis method uses SWOT analysis (Strengths, Weaknesses, Opportunities, Threats). SWOT analysis is useful for evaluating various aspects of tourist villages and identifying strategies that can be implemented for development and improvement. These findings represent an initial step towards formulating the development of smart village tourism by adopting a smart city perspective based on an ICT model, aligned with local potential and wisdom as key factors for the sustainability of tourist villages.
Increasing the accuracy value can be increased by using other algorithms. Increasing the accuracy value of a classification algorithm, the level of success of the algorithm's prediction is more precise and appropriate in providing its label. The purpose of the research is look performance of accuracy value for prediction with bagging algorithm. This research use random forest algorithm and bagging algorithm used for optimization. 12 data whose position is far from other data. 12 data deviate from the data pattern and are outliers. With z-score process, it will be processed to eliminate outlier data. After removing the outlier data, the data clean is 137 toddler data. After removing outliers and standardizing the data, the accuracy value obtained was 71% up to 100th accuracy with random forest algorithm. Optimization of a bagging algorithm to predict stunting in a dataset of toddlers that has been acquired and assessed its performance. This can be seen from the optimization of prediction results up to the 100th iteration, where the prediction accuracy results were 80.67%. Using the Random Forest algorithm and bagging techniques, the prediction of stunting in toddlers works well. Optimization of prediction results up to the 100th iteration, where the prediction accuracy results were 80.67%.
The bi-criteria objective scheduling is essential in the Jepara furniture industry due to its competitiveness. Scheduling that not only considers the company's profits but also takes into account the customer's perspective can add significant value to the company. Based on that, this paper proposed a mathematical model for the furniture finishing industry. Then it transformed into a Microsoft Excel Solver model. Cost calculation is also considered to choose the best model. The system's characteristic is flexible flow shop production, not identical at the last stage, and sequence-dependent set up time. The objective of scheduling is to minimize total maximum completion time and total weighted tardiness. There are 3 scenarios in this paper, company focused, customer, and bi-criteria objective. After running the model, scenario 3 is the best choice for completing priority orders on time, while scenario 1 is ideal when seeking efficiency in production with delays being less of a concern.
The increasing amount of household waste presents a major environmental challenge, worsened by inefficient and outdated waste management practices. Traditional systems lack real-time monitoring and responsiveness, creating a gap in timely waste management. This research introduces a creative solution through the development of an Android-based Household Waste Monitoring System, integrating Internet of Things (IoT) technology to provide real-time data on waste bin capacities and immediate notifications. Unlike conventional approaches, this system creatively bridges the gap by enabling proactive waste management through instant alerts and real-time tracking, allowing users to act before issues escalate. The system development follows an Agile/Scrum framework, fostering rapid iteration and user-driven enhancements. Through the innovative application of IoT and Agile methodologies with JIRA Software, this solution effectively addresses the inefficiencies of current waste management systems, as evidenced by an 80% success rate across five testing activities. This creative approach not only improves development efficiency but also accelerates adaptability in response to evolving waste management needs.
Educational Data Mining provides an effective approach to tackle numerous issues within the education sector, including the capacity to perform predictive analyses regarding student attrition based on academic information. In this research, data from the Open University Learning Analytics dataset (OULAD), which is publicly accessible, has been employed, which encompasses student information collected during online learning. We apply various Machine Learning models, including Decision Trees, Naïve Bayes, Logistic Regression, and ensemble approaches like Random Forest and AdaBoost. Among the models tested, Random Forest (RF) achieved the highest accuracy of 89.37%, along with a precision of 89.57% and a recall of 93.86%, using the data splitting approach. When employing an alternative evaluation model, specifically K-Fold Cross Validation, the maximum F1 score achieved was 9.45%. In summary, the ensemble machine learning algorithm, specifically Random Forest (RF), exhibited strong performance in predicting student academic achievement quality.
The Ministry of Religious Affairs of Lubuklinggau City organizes government affairs in Hajj and Umrah services. This research aims to design an enterprise architecture that can align the implementation of information systems with ongoing business activities, to improve the quality of hajj and umrah services at the Ministry of Religious Affairs of Lubuklinggau City. The method used in this study is the TOGAF Architecture Development Method (ADM), which consists of several phases: introduction, architectural vision, business architecture, data architecture, application architecture, technology architecture, and opportunities and solutions. The results of this study are in the form of an enterprise architecture blueprint that includes artifacts in the form of diagrams, catalogs, and matrices to describe existing conditions and proposed target conditions. In addition, this study produces a roadmap as a reference for implementing the architectural design that has been made. In conclusion, designing an enterprise architecture using TOGAF ADM can support the integration of information systems and business activities, thus potentially increasing the efficiency and effectiveness of hajj and umrah services at the Ministry of Religious Affairs of Lubuklinggau City.
Sleep is crucial indicator for an individual. Poor sleep quality has serious implication for health. This condition is often triggered by high work pressure and imbalance between work and rest time. While previous research with similar topic has been conducted, it has not comprehensively elucidated the key factors influencing sleep disorders. Therefore, this study conducts more in-depth analysis of factors contributing to sleep disorders including; gender, age, occupation, sleep duration, quality of sleep, physical activity level, stress level, BMI, heart rate, and daily steps. Subsequently, we employ Machine Learning (ML) techniques to investigate further sleep disorders. The ML models include: Naïve Bayes (NB), Support Vector Machine (SVM), K-Nearest Neighbor (KNN), Logistic Regression (LR), Convolutional Neural Network (CNN), dan Long Short-Term Memory Network (LSTM). The objective is to assess the effectiveness of ML model implementation based on information from data and the significance of specific factors in predicting sleep disturbances. The results of this study indicate that the combination of the LR model with Chi-Square achieved the highest average F1 score, which was 84.75%, in sleep disorder classification. The research comprises several stages: (1) Data collection, (2) Pre-processing of the collected data, and (3) Training models capable of processing data for evaluation to understand the contribution of indicators to sleep disorder predictions. The findings of this study provide insights into the effectiveness of the constructed models in predicting sleep disorders
PT Cottonink Duo Kreasindo utilizes SAP Business One technology to assist business operation used production, warehouse, MD Sales and operational as the moduls. In practice, deficiencies are still found that hamper work and operational productivity to improve information system support to improve and complete requirement. To ensure that PT Cottonink Duo Kreasindo require a system that able to facilitate users and run optimally, a paradigm requires function is to plan, design and develop the system; enterprise architecture. Enterprise architecture expectantly able to provide solutions for PT Cottonink Duo Kreasindo to increase the productivity and further align business strategy with the company's information systems. The TOGAF ADM method applied in this research. The blueprint resulting by enterprise architecture planning using the TOGAF ADM method includes detailed architectural designs, including business architecture, data architecture, application architecture, information system architecture, and technology architecture. There are two modules that need to be removed and six modules that need to be added to the business architecture; three modules that need to be removed and two modules that need to be added to the data architecture; as well as four modules that need to be added to application architecture and technology architecture.
This study aims to answer these questions by examining the effectiveness of green tourism marketing strategies by the Indonesian government, utilizing social media platforms Instagram and X (formerly Twitter), and learning how the public perceives its initiative. We collected comments on government posts promoting green infrastructure and tourism and analyzed them using a mixed-method approach. Sentiments were divided into three categories: positive, neutral, and negative, using Natural Language Processing (NLP) techniques. The results reveal a general negative reception of AI-authored advertising messages, amounting to 64.18% disapproving comments for Instagram and 81.93% for X, both expressing suspicion toward the overuse of AI and lack of authenticity calling into question transparency as well as genuine intentions behind sustainability goals. While a few responders gave positive reviews, the fact that so many responses were bordering or fully negative indicates that there needs to be clearer and more genuine communication strategies. Our research helps illuminate how the public perceives aspects of green tourism marketing, thus underscoring the significance of authenticity in promotional practices designed for sustainable development.
Rice plants that are processed into rice are the staple food of the Indonesian people, and the lack of rice production will have an impact on weakening national food security. Efforts that can be made are to process harvest data in national rice barn areas such as Indramayu Regency properly. So far, there are still many errors and differences in harvest data both by agencies and original data in the field. Differences in data cause inaccurate harvest data to be used as a reference for policies or to see the potential of rice in Indramayu. This study aims to build a website-based data processing information system so that it can be accessed and managed by agricultural officers in all sub-districts in Indramayu, and the agricultural service as admin, so that the data produced is accurate data and provides predictions of harvest results, and makes predictions of future harvests based on harvest data, land area and rainfall that affect the rice harvest in Indramayu using fuzzy tsukamoto. From the predictions made, there are 16 sub-districts that have the potential to experience a decrease in harvest from 31 sub-districts in Indramayu. This information system also displays harvest data and graphs based on year and sub-district in Indramayu so that the increase or decrease in harvest in previous years can be seen compared to predictions for the coming year.
Sentiment analysis, also known as opinion mining, is an important task in natural language processing and data mining. It involves extracting and analyzing subjective information from textual data to determine the sentiment or opinion expressed by the author. With the advancement of technology and the widespread use of social media and online review platforms, it is increasingly important to understand users' opinions and sentiments regarding a particular product, service or issue. The purpose of this research is to present a comprehensive literature review on sentiment analysis techniques. This research utilizes the systematic literature review method. This method involves systematic steps in searching, evaluating, and analyzing relevant literature in the field of sentiment analysis. The literature search was conducted through scientific databases and other reliable sources. Relevant articles were then selected based on pre-determined inclusion and exclusion criteria. The data from the selected articles were then comprehensively analyzed to identify the sentiment analysis techniques used and the key findings in the research. The results show that there are various techniques and approaches that have been developed and tested in sentiment analysis, some of the commonly used techniques include rule-based methods, classification-based methods, and machine learning-based methods.
In the telecommunications industry, predicting customer churn is crucial for maintaining business sustainability. High churn rates can negatively impact profitability, necessitating effective retention strategies. This research aims to enhance the accuracy of telecommunications customer churn prediction by optimizing the C4.5 classification algorithm through feature selection and hyperparameter tuning. The methods used include Information Gain for feature selection and hyperparameter tuning with Random Search and Grid Search. This study utilizes the Telco Customer Churn dataset from Kaggle, split into an 80:20 ratio for training and testing data. Six approaches are applied: (1) the basic C4.5 algorithm, (2) C4.5 with Information Gain, (3) C4.5 with Random Search, (4) C4.5 with Grid Search, (5) C4.5 with a combination of Information Gain and Random Search, and (6) C4.5 with a combination of Information Gain and Grid Search. The results indicate that the C4.5 algorithm alone achieves an accuracy of 74.09%, while applying Information Gain increases accuracy to 78.42%. Hyperparameter tuning with Random Search achieves the highest accuracy of 80.05%, whereas Grid Search reaches 77.71%. Combining Information Gain with Random Search results in an accuracy of 78.99%, while combining Information Gain with Grid Search yields an accuracy of 78.85%. These findings suggest that hyperparameter tuning using Random Search significantly improves accuracy compared to other methods, while Information Gain feature selection does not have a significant impact on performance in this context.
Information technology is currently developing rapidly and covers many aspects of people's lives. The development of information technology also includes in the financial sector. The development of information technology in the financial sector is referred to as Financial Technology. The purpose of this study was to test and analyze The Utilization of Electronic Payment moderated by Financial Technology Innovation on Financial Technology Payment in Indonesia. The theories used in this study are the Technology Acceptance Model (TAM) and the Unified Theory of Acceptance and Use of Technology (UTAUT), Electronic Payment, Financial Technology, Diffusion of Innovation Theory, and Financial Technology Payment. Diffusion of Innovation (DOI) Theory in 1962, making it one of the earliest social science theories. A theory known as "diffusion of innovations" aims to explain how, why, and how quickly new concepts and technologies proliferate. Everett Rogers made the theory more widely known in his 1962 book Diffusion of Innovations. The research method used in this research is quantitative methods with primary data obtained from distributing questionnaires using Google Form. The data were collected from 316 respondents. The questionnaire was structured using a Likert Scale of 1-5 (Strongly Disagree - Strongly Agree). The research data were analyzed using the Structural Equation Model (SEM) with WarpPLS7. The results showed that The Utilization of Electonic Payment accepted with P Value < 0.01 to Financial Technology Payment in Indonesia. The Information Technology Innovation variable can act as a moderating variable between The Utilization Electronic Payment and Financial Technology Payment in Indonesia with P Value = 0.01
Taxpayer supervision at the Directorate General of Taxes (DGT) faces new regulations requiring comprehensive oversight. The existing core tax system is deemed inadequate, prompting account representative (AR) officers to seek alternatives. The innovation in the form of an end user computing (EUC) applications used in supervision procedure has proven beneficial despite the lack of official support. This study aims to investigate the innovation characteristics that influence innovation adoption within the AR of DGT, drawn from the diffusion of innovation theory (DOI) and combining it with moderating variables of the unified theory of acceptance and use of technology (UTAUT). The study involved 224 AR officers at the DGT Bali regional office, selected through convenience sampling. Hypothesis testing was conducted using the Partial Least Squares Structural Equation Modeling (PLS-SEM) method. The results indicate that the characteristics of observability, relative advantage, and compatibility significantly influence AR’s intention to adopt innovation, while complexity and trialability proved insignificant. Furthermore, age, gender, and experience did not significantly moderate the influence of innovation characteristics. In conclusion, this integrated model successfully examined the innovation characteristic factors that influence the adoption of EUC in supervision at DGT. The study has theoretical implications by providing empirical evidence from TDI characteristics combined with the UTAUT model. This study has limitations in collecting AR research sample data only in the Bali Regional Tax Office work unit and data collection at one point in time that is not continuous, so the data is only cross-sectional.
