Classification model for NAVTEX navigational warning messages based on adaptive weighted TF-IDF
Pengbo Sun, Yi Zuo, Yudi Wang · 2023
NAVTEX is a crucial marine safety information broadcasting system for ensuring the safe navigation of ships, which plays a significant role in ship safety. However, the current manual reading and subject classification of NAVTEX suffer from low efficiency and accuracy. To enhance the processing efficiency of maritime safety information (MSI) and promote information and communication, achieving automated MSI classification with high confidence in the accuracy of the results becomes imperative. In the context of machine learning, this study proposes an adaptive weight TFIDF method to address the aforementioned challenges. The primary objective is to optimize the weights of keywords with prominent classification features in NAVTEX. Experimental results demonstrate that the adaptive weight-based TFIDF algorithm significantly improves the classification outcomes for NAVTEX. By enhancing the accuracy and efficiency of MSI classification, this approach facilitates the automation of NAVTEX analysis and promotes the reliability of the generated classification results.