<![CDATA[This research investigates the integration of Latent Dirichlet Allocation (LDA) for topic modeling with the performance evaluation of various classification algorithms—specifically, k-nearest Neighbors (k-NN), Support Vector Machines (SVM), Naive Bayes Classifier (NBC), and Decision Trees (DT)—within the Digital Content Reviews and Analysis Framework. The framework systematically processes and analyzes digital content, including data cleaning, extraction, evaluation, and visualization techniques, to enhance machine learning models' interpretability and predictive accuracy. The study demonstrates that combining LDA with these classification algorithms significantly improves data interpretation and model performance, particularly in handling large-scale textual datasets. Notably, the Decision Tree algorithm achieved a 98.86% accuracy post-SMOTE. At the same time, the Support Vector Machine reached a near-perfect AUC of 1.000, highlighting the efficacy of these methods in managing imbalanced datasets. The findings provide valuable insights for optimizing model selection and developing more robust and adaptive machine-learning models across various applications. This research contributes to advancing the field of artificial intelligence by proposing a comprehensive framework that effectively addresses complex data-driven challenges, encouraging further exploration of more flexible and scalable models to accommodate evolving data environments.]]>
Copyrights © 2024
