cc-IFF : a cascading citations impact factor framework for the automatic ranking of research publications

Abstract

A new framework is proposed for the calculation of impact factor ratings of research publications. Given a collection of research articles, the corresponding citations graph is constructed in the form of a relational table. The impact value is considered at the article level, and is calculated by considering not only the citations made directly to an article, but also citations made to the corresponding citing article(s). In this respect, an improved algorithm is utilized, namely one that traverses all the threads in the citations graph, in an attempt to improve the degree of fairness in assigning credit for the impact value of each one article. When two articles have an equal number of (direct) citations, the one that has triggered more research activity (i.e. its citing articles attract a larger number of citations at subsequent levels in the citations graph) is assigned a higher impact value and, consequently, is ranked to be better