Clustering of scientific publications on marine autonomous surface navigation to identify clusters of keywords

Abstract

An approach to structuring scientific publications is proposed using the example of scientific works on marine autonomous surface navigation, which allows for data mining with clustering and visualization of the main keywords and phrases related to the designated topic. The resource of the Russian scientific electronic library was used as a source for obtaining data eLIBRARY.RU. Information has been collected on 534 relevant scientific publications for the period from 2013 to the beginning of 2024, containing the title of the work, the year of publication, the list of authors and their affiliation, keywords and phrases, etc. A database for intellectual analysis has been formed, containing 1304 keywords or phrases indicated by the authors of scientific publications related to the selected topic. Data mining was performed using an open source software package for visualization and analysis of related data Gephi. The processing results are presented in the form of undirected graphs. The undirected graphs obtained are presented, and the characteristics of certain graphs for various kernel parameters are indicated. Clusters of keywords or phrases and connections between them have been identified, demonstrating current trends in the development of the chosen scientific field. With the advent of new scientific publications on marine autonomous surface navigation, the graphs will change, fixing new connections between the authors of scientific publications, key words and phrases. Such an approach to data processing may be useful for students, graduate students and researchers to identify a pool of keywords and phrases that need to be paid attention to when studying this topic.