Unsupervised approaches for mining author contributions in scientific manuscripts

Abstract

Measuring scientific collaboration is of utmost importance To quantify the credits among collaborating authors is yet more challenging Authorship attribution only conveys who may be the writer of a scientific article It does not quantify the contribution of an individual author As hyper authorship is increasing methodologies to quantify author credits using newlinebibliometric author level metrics fail This thesis attempts to arrive at a solution for this question using machine learning algorithms by semantically analysing the authors contributions newline newline