TOPIC ANALYSIS OF SCIENTIFIC ARTICLES FROM FOUR ROMANIAN UNIVERSITIES, BASED ON NATURAL LANGUAGE PROCESSING
TOPIC ANALYSIS OF SCIENTIFIC ARTICLES FROM FOUR ROMANIAN UNIVERSITIES, BASED ON NATURAL LANGUAGE PROCESSING

Summary/Abstract: In this paper we used Natural Language Processing techniques to perform a thematic analysis of scientific articles from four Romanian Universities. Our main objective is to find important word combinations in the text data and to highlight the most frequent of these. The data used in this research refers to higher education institutions in Romania, covering a period of twelve years and focuses on text data represented by abstracts of scientific articles. We also limited our analysis to a number of four universities that are structured in a similar way. We performed an analysis of the most frequent words for the four universities studied and presented the results in word clouds graphs and also created a ranking of the most common bi-grams, tri-grams and tetra-grams according to their frequency of occurrence.

• Details
• Contents