International Journal of Data Science and Advanced Analytics

Early Adoption of Online and Mobile Banking: An Analysis of Consumer Attitudes

Isaac Ngugi — 2025-07-30

Consumer behaviour is influenced by a variety of factors, among which consumer attitudes are particularly significant. Attitudes play a pivotal role in shaping individuals’ purchasing and consumption decisions. This study conducts a literature review focusing on consumer attitudes toward online and mobile banking during the early stages of their adoption, specifically between 2002 and 2011. As digital banking channels gained popularity, this behavioural shift contributed to a gradual decline in the relevance of physical bank branches. The review is guided by three primary objectives: (1) to explore general consumer attitudes toward online and mobile banking, (2) to identify the factors associated with of these digital banking platforms that positively or negatively influenced consumer perceptions, and (3) to examine demographic variations in consumer attitudes. The findings indicate a generally favourable consumer disposition toward online and mobile banking during the study period, with increased adoption reflecting growing acceptance. Positive influences on consumer attitudes included time efficiency, ease of use, convenience, and enhanced autonomy through 24/7 account access and control. Conversely, concerns related to security and the loss of in-person interactions emerged as key negative factors. Demographic analysis revealed that younger consumers were more inclined to embrace digital banking compared to older age groups. These insights offer practical implications for financial institutions, suggesting strategic areas for improving customer engagement and service delivery in the digital banking landscape.

Enhancing Judicial Efficiency through Artificial Intelligence: Analyzing Federal Justice Systems from an Organizational Behavior Perspective- A Data-Driven Study

Saleh MANSOUR — 2025-07-26

Artificial Intelligence (AI) has brought transformation prospects in many fields, and that covers the U.S. Federal Justice System. This study specifically identifies AI-powered risk assessment algorithms, predictive analytics, and automated case management systems as holding potential to minimize judicial backlogs, foster more consistent decision-making, and reduce recidivism rates. With this integration of AI, on the other hand, comes some new challenges: it risks reinforcing historical biases; there are some ethical concerns about transparency; and a few related to public trust. This paper discusses the impacts of organizational behavior using both quantitative data and qualitative insights, assesses ethical risks, and presents recommendations for the responsible deployment of AI. Results have indicated a considerable gain in efficiency after the integration of AI, but at the same time pointed out that continuous refinement of AI tools is necessary in the course of upholding the principles of fairness and justice. The paper concludes by discussing some policy suggestions, laying out future directions for research to better improve AI's role at the judiciary.

Balancing Accuracy and Efficiency: A Comparative Analysis of Collaborative Filtering Algorithms for Job Recommendation Systems

Arash Khosravi — 2024-08-13

Recommender systems are commonly used to suggest relevant items to users, like movies or products. The digital transformation of the business sector has led to a surge in online job opportunities. This shift necessitates effective job recommendation systems to connect qualified candidates with relevant positions. This study evaluates the performance of four collaborative filtering algorithms for a job recommender system: Singular Value Decomposition (SVD), SVD++ (SVDPP), co-clustering, and Non-Negative Matrix Factorization (NMF). We employ error rate, training time, and cross-validation performance as key evaluation metrics. Our findings reveal a trade-off between accuracy and efficiency. The co-clustering approach achieves the lowest error rates, indicating its effectiveness in recommending relevant jobs. However, this benefit potentially comes at the cost of increased training time compared to other methods. Conversely, the NMF-based model demonstrates significantly faster training times, making it computationally efficient.

Leveraging Educational Data Mining: XGBoost and Random Forest for Predicting Student Achievement

Arash Khosravi — 2025-07-26

Universities and educational institutions are accumulating and storing substantial amounts of data that include the personal and educational information of students. There is an ongoing debate regarding the most crucial factors for predicting students' academic achievement, as well as determining the most suitable algorithm to employ. Furthermore, if these results are achieved, administrators need to develop better planning strategies. Educational Data Mining (EDM) is a technique used to extract specific data types from an educational system, aiding in a comprehensive understanding of students and the system itself. EDM involves transforming raw data obtained from training systems into valuable data that can facilitate data-driven decision-making. In comparison to other fields, the development of data mining and analysis in education has been relatively slow. However, mining educational data on the web presents unique challenges due to specific characteristics of the data. Although various data types possess sequential aspects, the distribution of training data over time exhibits remarkable properties. In this research, we want to find out whether alternative machine learning models, in addition to random forest, can perform comparable or even better in predicting students' academic achievement, therefore, we propose a method that utilizes XGboost and Random Forest algorithms to identify the significant factors influencing prediction accuracy.

Predicting Crude Oil Price Using Time Series Statistical Modelling Techniques

Sunny Raj Pradeep Kumar Gupta — 2024-06-21

One of the important topics in Global Economy is crude oil prediction. Nonetheless, crude oil prediction is related to many factors that play a role in it and include geological factors, production, technical etc…Accurate prediction of crude oil is significant considering all the previous factors. However, these factors are volatile and vary under different conditions. Therefore, this study proposes using machine learning models for time series predictions of crude oil. More specifically, three types of convolutional neural networks were used being autoregressive integrated moving average (ARIMA), seasonal autoregressive moving average (SARIMA) and long short term memory (LSTM). The models were applied to three datasets of crude oil including Crude Oil Daily Price Data, Crude Oil Weekly Inventory Data and Crude Oil Monthly Production Data. Data analysis was conducted using Python. The outcomes of the predictions showed that LSTM had higher accuracy and prediction than ARIMA and SARIMA. The LSTM daily model achieved a MAPE score of as low as 0.087. Future research involve more reliable production data and include more reliable input parameters to improve accuracy.