A network approach for mapping and classifying shared terminologies between disparate literatures in the social sciences

Abstract

Methodologies for literature-based discovery in their closed modality, aim to connect two disjoint literatures A and C by providing a list of B-terms from where hypotheses can be drawn. In this paper, we propose a method for representing B-terms as a network. Such representation further allows us to obtain shared themes and ease the exploration of common terms. This is done by firstly dividing the A and C literatures into narrower topics and extracting intersecting lists of B-terms between those. The lists are then used to compute a term cooccurrence network. The method can be applied to fields where curated or standardized thesaurus does not exist, like in the social sciences. We illustrate the method by linking the disparate literatures on poverty alleviation and the Internet of Things. One being the first and most pressing social issue targeted by the Sustainable Development Goals and the other a rapidly growing concept that encompasses technological solutions. We expect our method contributes as an exploratory tool to navigate literature-based discovery outputs easily.