Bi-Level Decision Tree Approach for Web Quality Assessment

The rapid expansion of smart mobile technology is revolutionizing how governments deliver information and services to the citizens. Conspicuously, the usability and accessibility of government websites are crucial factors that determine the quality and accessibility of mobile e-government. This research aims to analyze the accessibility and usability of public sector websites. This study assesses how well websites adhere to usability standards and accessibility. Specifically, the study presents an automated framework that uses the Probabilistic Bayesian Technique (PBT) to analyze data from websites. The data is used to determine the Probability of Website Quality (PWQ) and evaluate the cumulative temporal aspect in terms of Website Quality Measure (WQM). Additionally, bi-level decision-tree modeling is presented for decision services focused on website quality. For performance validation, the evaluation of websites was carried over challenging dataset with nearly 60,625 instances. Statistical results demonstrate that the provided model surpasses state-of-the-art decision-making techniques of Support Vector Machines, Artificial Neural Networks, and K Nearest Neighbor in terms of Data Assessment Effectiveness (95.93 seconds), Decision-making efficiency (96.4%), Classification Analysis (Specificity (96.32%), Precision (94.45%), and Sensitivity (95.33%)), Stability (0.73), and Reliability (91.58 %).