Deep Learning-Based Tag Mapping Automation of Ship Data Models with Natural Language Processing

Abstract

The current Maritime Autonomous Surface Ship (MASS) presents many challenges and complexities in terms of risk, stability and implementation. To overcome these obstacles and realize the MASS vision, it has become imperative to develop a shore-based data platform capable of monitoring and supporting ship conditions in real time. In the data platform, a tag mapping operation is required to collect ship data onshore. The traditional approach is usually for the designer to perform the mapping manually by looking at the description and performance of the ship’s I/O list. Such an approach usually requires a lot of design M/H and may lead to human errors, wasting a lot of time and resources. With the rapid development of deep learning, it becomes possible to realize the tag mapping automation task by converting it into a natural language classification task. Therefore, we propose a deep learning-based framework for tag mapping automation of ship data. Specifically, we consider the tag mapping task as a natural language classification task as follows: Classify the ship data into the corresponding platform data by a natural language classification model, and select the corresponding rules to achieve the tag mapping according to the classification result. Our proposed method reduces the manual involvement in the label mapping operation, minimizes the risk of manual errors, and saves resources. abstract environment.