Bisociative Knowledge Discovery for Cross-domain Literature Mining

Abstract

Given its immense growth, the scientific literature can be explored to reveal new discoveries, based on as yet undiscovered relations between knowledge from different, relatively isolated fields of specialization. This chapter presents an approach to creative knowledge discovery through the mechanism of bisociation. Bisociative reasoning is at the heart of creative, accidental discovery, i.e., serendipity. Bisociative knowledge discovery is focused on finding unexpected links by crossing between different contexts. In this work, bisociative knowledge discovery is explored in the framework of text mining, addressing cross-domain literature-based discovery. Two approaches are briefly outlined: the CrossBee approach to cross-domain bridgingterm detection, and the OntoGen approach to bridging-term detection through outlier document exploration.