Measuring of "Idea-Based" Influence of Scientific Papers

Traditional bibliometrics such as citation counts and Science Citation Index are often used to find most influential and quality scientific articles. Modern methods of analysis of citation graph take into account the content of the articles. Topical influence measures a tendency of an article to influence the ideas of citing articles. The presented method estimates the information content and ideas of scientific articles based on linguistic approach. The method can measure "idea-based" influence and how ideas propagate between articles. Idea is presented very precisely as a set of phrases with similar meaning. Phrases representing the ideas in the texts of the articles form together a system of implicit references/links. Implicit link is a reference to the author or his ideas in the text. Bibliometric indicators which in addition to bibliographic references also take into account the implicit links can more accurately measure scientific impact and the quality of the articles.