Aim: The research work aims to improve the accuracy of the information extraction and digitalization of human handwritten character documents using Decision Trees with machine learning algorithms. Materials and Methods: The categorizing is performed by adopting a sample size of N=20 in the Decision Tree and sample size of N=20 in Random Forest algorithms. Results: Novel Decision Tree delivered significant results with 95.00% accuracy, compared to Random Forest 93.33% accuracy. Novel Decision Tree and Random Forest statistical significance is p = 0.57 (p < 0.05). The Independent sample T-test value states that the results in the study are significantly not achieved with a 95% confidence level. Conclusion: Information extraction and digitalization of human handwritten character documents with the Novel Decision Tree provides better accuracy than the Random Forest algorithm.