Research on Natural Language Processing and Aspose Technology in the Automatic Generation of Ocean Weather Public Report

Abstract

Combining with the requirements of the production of the ocean weather public reports, based on the analysis of a large number of historical documents issued daily by National Meteorological Center (NMC), it proposes an automatic generation method of ocean weather public reports by using Natural Language Processing technology and Aspose technology. Furthermore, it focuses on three key issues for further discussion, such as historical text content analysis and feature extraction, text description rules and their natural language processing, report template customization and automatic generation of report document from the template by Aspose technology. Meanwhile, it proposes the implementation in detail for these issues. Compared with the content between forecasters’ manuscript and the computer report document, the results show that the automatically generated forecast text is correct in content, semantics and syntax, smooth in sentences, and the text style conforms to the habits of forecasters. The automatically generated report document effectively avoids the common errors of weather elements and time interval selection, illustration sorting and text style in the process of manually editing documents. It can effectively improve the efficiency of forecasters.