This study conducts a bibliometric analysis of data mining publications in the Scopus database, examining the evolution of the field from 2015 to 2024. The study examines the bibliometric structure of data mining in management. Analyzing 2,942 publications, the research identifies significant growth in data mining studies. It reveals gaps in integrating data mining with decision-making, artificial intelligence, forecasting, and sentiment analysis. Despite a large number of publications, interdisciplinary applications of data mining are limited. The scientific publication on data mining and its relationship with decision-making, artificial intelligence, forecasting, and sentiment analysis is found to be weak, showing significant research gaps in these areas. China and the USA are prominent contributors, indicating geographical concentration. The study highlights the need for broader interdisciplinary exploration in data mining beyond traditional areas, urging global researchers to diversify contributions. The analysis focuses solely on publications indexed in Scopus, potentially excluding relevant studies from other databases or sources. This study provides insights into the evolution of data mining research and identifies areas for further interdisciplinary exploration, contributing to the advancement of the field's boundaries.

The website of higher educational institutes is considered a vital communication channel to provide main resources to their stakeholders. It plays an important role in disseminating information about an institute to a variety of visitors at a time. Thus, the quality of an academic website requires special attention to respond to the users’ demands. This study aims to explore the quality of the PSAU website based on data mining techniques. The first step: was collecting opinions about the PSAU website using a survey. After that, data mining processes were used as descriptive and predictive models. The descriptive model was applied to describe and extract the major indicators of website quality. Besides, the predictive model was applied to create models for predicting the website quality level. More than one classification algorithm was used. Naive Bayes and Support Vector Machine have given the best results in all performance indicators, and the achieved accuracy rate for both algorithms was 86% and 84% respectively. The results revealed that the overall quality level of the PSAU website is very good. The usability quality and content quality were very good. The service quality needs more attention. which indicated that the service level is inadequate and needs to be further enhanced. The results of the study should be useful to the deanship of Information Technology at PSAU, and website developers, in redesigning with high quality in terms of its usability, content, and service.

In the era of big data, clustering and data mining have become essential tools for uncovering patterns and insights from vast datasets. However, these processes often involve the use of sensitive data, raising significant concerns about privacy, security, and trustworthiness. This paper proposes N2P-CM, a novel privacy-preserving framework designed to protect sensitive information during the entire clustering and mining lifecycle. Unlike existing methods that focus on partial aspects of security or apply generic encryption techniques, N2P-CM integrates five innovative and synergistic modules: Sensitive Feature Obfuscation, Adaptive Trust Weight Aggregation, Compressed Secure Semantic Embedding, Differential Traceable Execution Engine, and Blockchain Auditable Ledger. Each module contributes a distinct layer of privacy and accountability, ranging from feature-level data transformation and federated trust scoring to secure semantic encoding and traceable execution logging with blockchain support. We provide formal definitions and algorithms for each module and demonstrate their integration in a unified architecture. Extensive simulations using real-world datasets validate the efficacy of N2P-CM, showing that it achieves strong privacy guarantees with minimal degradation in clustering accuracy. This research contributes a comprehensive and modular solution to the growing need for privacy-preserving analytics in sensitive domains such as healthcare, finance, and smart cities.

The use of technology in the financial industry has experienced sustained growth in recent years. However, in many emerging economies, a significant proportion of the population still does not utilize digital solutions for financial transactions. Promoting financial inclusion through digital environments is essential for driving social and economic development. This study aims to develop machine learning models to predict the adoption of digital payments in Latin America and the Caribbean using statistical data from the World Bank's Global Findex Database for 2021. The performance of the Random Forest, LightGBM, XGBoost, and CatBoost algorithms was compared, with the optimal hyperparameter combination identified through Bayesian optimization. The results show that LightGBM achieved the highest performance in predicting digital payments, with an F1-score of 90.25% and a more stable balance between precision and recall compared to the other models. These findings highlight the value of machine learning models in the financial sector, as they enable a more accurate identification of users adopting digital solutions, facilitating the design of strategies to strengthen financial inclusion in the region.

The primary objective of the research paper is to provide some insights on the concepts of intellec-tual capital and its important dimensions alongside to investigate the possible association between intellectual capital and attainment of competitive advantage. This study focuses on Iraqi hotel industry which is an under-research area in the context of intellectual capital and its relationship with competitive advantage. An adapted questionnaire was utilized to collect the responses from top and middle level managers of four and five-star hotels in Iraq. The reliability and validity of data collec-tion instrument were measured through Cronbach’s alpha, Composite Reliability and Average Var-iance Extracted respectively. The competitive advantage was then regressed against Human, Rela-tional and Structural capital by application of Partial Least Square methodology. Results of the study showed a positive and strong connotation among intellectual capital and attainment of competitive advantage. The human capital had the highest contribution for competitive advantage in hotel sector of Iraq. The top management of hotels should take interest to develop, maintain and retain human capital to attain competitive advantage over competitors.

Internal audit is a bulwark ensuring the integrity of financial statements, a linchpin for stakeholder trust and informed corporate decision-making. With the proliferation of complex financial transactions, audit teams face mounting challenges in deciphering voluminous transactional data to safeguard financial reporting quality. Machine learning has the potential to identify signifiers of financial reporting quality. Within the Design Science Methodology framework, we apply the Random Forest Classifier technique to metrics such as the error rate to enhance financial reporting. We find that the Random Forest Classifier identifies that certain parameters are critical to error detection, which enhance account receivable accuracy, lower receivable account control risk. This research advances the argument that technologically-enhanced internal audit procedures can play a pivotal role in ensuring that financial reporting mirrors the economic reality of the company.

The aim of this study was to evaluate the impact of social media, big data, and data mining on the development of organizational behavior within the telecommunications industry in Jordan. The main objective of this study was to investigate the effects of technological components on the alteration of organizational behavior in the communications sector of Jordan. To accomplish this objective, a thorough empirical investigation was undertaken, encompassing the collecting of data from key stakeholders within the telecommunications sector in Jordan. A sample size of 412 participants, encompassing people from diverse roles within the communications sector, was chosen for the purpose of this study. The participants' replies and perspectives were gathered via the administration of surveys and conducting interviews, resulting in a comprehensive data set suitable for analysis. This study investigated the intricate relationship between the utilization of social media, the application of big data analytics, and the implementation of data mining techniques in influencing the dynamics of organizational behavior. The study's results underscored the substantial impact that social media platforms have on communication patterns and collaboration within telecommunication firms. Furthermore, the utilization of big data analysis has emerged as a significant catalyst for the enhancement of informed decision-making processes, exerting influence on diverse facets of organizational behavior, including strategic planning, employee engagement, and customer interactions. Data mining techniques have been identified as having a crucial function in extracting significant patterns and trends from extensive datasets, hence helping to the improvement of organizational learning and adaptation. The research findings indicated that the incorporation of social media, big data, and data mining technologies had a beneficial effect on the development of organizational behavior within the telecommunications industry in Jordan. The findings underscore the importance for enterprises to proactively utilize these technologies to cultivate a work environment that is characterized by increased agility, responsiveness, and collaboration. This study provides significant contributions to the subject of organizational behavior by examining the impact of social media, big data, and data mining within the specific context of the telecommunication sector in Jordan. The research sheds light on the transformative consequences of these technological advancements. The consequences of these findings have broad relevance for organizational leaders, politicians, and researchers, serving as a basis for further investigations in the dynamic realm of technology-driven organizational behavior.

AI is tanking different functions of businesses, and marketing is no exception. Digital marketing is gaining pace with the advancement in technology and the internet. The research aims to find the answer to the research question that marketers can benefit from AI in content creation for the digital market. The study also finds the relevance and use of AI in content creation and develops an AI infrastructure adoption model for content creators in digital marketing. The findings of this study were compiled using a systematic literature review that adhered to the Preferred Reporting Items for Systematic Reviews (PRISMA) statement and the criteria established by Meta-Analyses. The findings revealed that using AI in content creation provides personalized data, which helps the creators make relevant, targeted, and specific content. The research also finds that AI alone is not mature enough to carry out the whole content creation procedure as there is some limitation attached, especially regarding ethical implications. That’s why human surveillance of AI systems involved in content creation for the digital market is needed.

Plastic bags are used by many people because they are inexpensive, lightweight, durable, and waterproof. Plastic bags, on the other hand, do not break down and can pollute the environment if not handled properly. Indonesia produces a lot of plastic waste and is one of the top ten countries that has a problem with plastic waste. In this study, we used three months of data of real transactions from a grocery store. This study shows how the decision tree can identify patterns on plastic bag usage at a small grocery store by using demography and products purchase. The attribute weights showed that in the hometown, the total of several products bought were the factors that affected the use of plastic bags.

A time series is a stochastic process which is arranged by time simultaneously. In this article, a time series model is used in accordance with Box-Jenkins' procedure. The Box-Jenkins procedure consists in identifying the model, estimating the parameters and diagnostic checking. The time series model is differentiated according to the number of variables, i.e. univariate and multivariate. The univariate method for the time series model that is often used is the Autoregressive Integrated Moving Average (ARIMA) model and the multivariate time series model is the Vector Autoregressive Integrated Moving Average (VARIMA) model. In this research, we studied the ARIMA model which is studied with a non-constant error variance. In this case, the Autoregressive Conditional Heteroscedasticity (ARCH) model is applied to outgrow the non-constant error variance. Selection of the best model by examining the minimum AIC for each model. The ARIMA-ARCH model is implemented on rainfall data in Bandung city with Knowledge Discovery in Database (KDD) in Data Mining. The methodology in the KDD process, including pre-processing, data mining process, and post-processing. Based on the results of model fitting, the best model is the ARIMA (2,1,4)-ARCH (1) model. The result of forecasting rainfall in Bandung shows a MAPE value is 11%, which has a similar pattern with actual data for short time 2-4 days. From these results, we conclude that the ARIMA-ARCH model is a good model for forecasting the rainfall in Bandung city.