<![CDATA[This paper aimed to develop a Precision Document Transaction Type Classifier using machine learning to identify transaction types, aligning with the Ease of Doing Business Law (RA 11032), which aims to streamline government services and improve service delivery. With the use of existing government documents, a dataset was created and processed for the training and evaluation of models, including Naïve Bayes, Bidirectional Long Short-Term Memory (Bi-LSTM), and Bidirectional Encoder Representations from Transformer (BERT). The BERT Model was the most accurate, efficient, and precise among other models. For the development of the software application Agile Methodology was used to ensure iterative progress and adaptability during the development phase. For the software quality evaluation, it was assessed using ISO/IEC 25010:2011, achieving a general high score mean of 4.25 corresponding to a descriptive equivalent of Excellent covering various software quality metrics demonstrating reliability, efficiency and overall performance.]]>
Copyrights © 2025
