Scientific knowledge percolation process and social impact: A case study on the biotechnology and microbiology perceptions on Twitter

This article describes a methodology for analysing the diffusion of scientific information into the social sphere, termed the 'scientific knowledge percolation process'. The methodology was built using automated data collection and lexicon-based data mining techniques. We analysed literature from the scientific biotechnology community (158 journals in 2011; 29,892 articles generating 50,591 different keywords) and how it is perceived by users of the social media site Twitter (375,660 tweets with a subset classified by sentiment (positive, negative, and neutral) for a total of 33,900 tweets for 2012). We show that our method is able to provide data from which we can draw robust conclusions concerning the relationship between scientific and social media information. The study shows that the scientific production of our subset is socially perceived in a neutral manner although it is skewed towards the negative. Because sentiments are relevant for explaining the sharing behaviour of social media users, the results suggest that more attention needs to be paid towards the social perceptions of scientific research. We found that similar scientific concepts can be socially perceived in different ways, which may suggest that there is room for scientists to choose more 'socially friendly' descriptions.