Text Mining Approach for Identifying Research Trends

With the increase of unstructured data, the issues connected with automatic text processing, the categorization of documents and the discovery of topics have become objects of growing interest. In order to improve the process of grouping and processing research publications, we would like to propose a method based upon natural language processing. It is based on text mining technologies which aim to identify key tendencies in documents. It processes the content of publications by clustering and identifies the topics of each identified group. This analysis helps by identifying key tendencies as well as discovering emerging new areas of research. Publications from the research literature database, Scopus, were used to test the approach. The topic of the publications is “the application of digital technologies in the logistics business”. The experiments were completed using the RapidMiner Studio software.