A Modeling Approach for the Extraction of Semantic Information from a Maritime Corpus

Abstract

This paper introduces an algorithm for retrieving semantic information from a maritime corpus. The method is based on Natural Language Processing (NPL) and combines a segmentation of large documents with principles of a conceptual vector model (CVM) and synsets of words. This research is applied to the context of intelligent transport systems and maritime navigation. Based on documents regulating maritime traffic, this approach proposes an aid for navigational decision-making while significantly reducing the number of entities and relations required in the modeling process.