ABSTRACT
One methodology to the above problem is literature-based discovery (LBD) which directly addresses the problems of knowledge overspecialization. LBD strives to nd connections that are novel and that have not been previously explicitly published. In 1986, Don Swanson presented his rst literature-based hypothesis that sh oil may have benecial eects in patients with Raynaud’s syndrome. Literature search identied 2000 articles on Raynaud’s syndrome and around 1000 articles on dietary sh oil. e two groups of articles appear to be isolated but have signi- cant common attributes related to blood viscosity, platelet aggregability, and vascular reactivity. e possibility of linking fragmented literature through intermediate or shared attributes has commonly been described as Swanson’s ABC model (Figure 14.1). e model can be implemented as two discovery processes. In the rst, open discovery, concepts A, B, and C are known and the relationship between A and B and B and C are known,
but the relationship between A and C has not been identied. e goal is to discover that relationship. Discovery is facilitated by using particular B concepts to draw attention to a connection between A and C that has not been previously noticed. Starting with concept A, all the B concepts related to A are retrieved and then all C concepts related to B are also retrieved. If there are no prior reported relations between A and C, then a hypothesis of association between A and C can be formulated which can be conrmed or rejected through human judgment, laboratory methods, or clinical investigations. e second paradigm, closed discovery, is used to explain an observed phenomenon. A relationship between, two entities provided by user, A (e.g., ecainide) and C (e.g., heart failure), is assumed but undened. Proving the existence of the relationship is facilitated by identifying B (e.g., kcnip2 gene) concepts that provide mechanistic links between A and C.
