Literature‐based discovery: Beyond the ABCs

Abstract

Literature-based discovery (LBD) refers to a particular type of text mining that seeks to identify nontrivial assertions that are implicit, and not explicitly stated, and that are detected by juxtaposing (generally a large body of) documents. In this review, I will provide a brief overview of LBD, both past and present, and will propose some new directions for the next decade. The prevalent ABC model is not "wrong"; however, it is only one of several different types of models that can contribute to the development of the next generation of LBD tools. Perhaps the most urgent need is to develop a series of objective literaturebased interestingness measures, which can customize the output of LBD systems for different types of scientific investigations.

Improving Literature-Based Discovery with Advanced Text Mining

“…genes, drugs and diseases and it has resulted in new scientific discoveries, e.g. identification of candidate genes and treatments for illnesses [6,21].…”

“…concept chains [8], network analysis [16], and logical reasoning [24] (see [21] for a survey of such extensions). For a concept pair A and C, these models identify the most obvious B and return a ranking of pairs using measures such as average minimum weight, linking term count and and literature cohesiveness [26].…”

“…These tools have been used to generate new scientific discoveries (e.g. candidate genes for Parkinson's disease, a link between hypogonadism and diminished sleep quality); see [6] and [21] for recent reviews. However, confirmation of such discoveries via actual laboratory experiments remains rare.…”

“…Just one example method is link prediction in complex networks [14] which has been applied successfully to to related problems in social network analysis [11] and web mining [2]. In comparison with most current LBD which is based on extensions of Swanson's ABC model [21] and considers only the most obvious intermediate concepts, such enhanced techniques could provide improved estimate of links between concepts. They could also generate data needed for calculating the likelihood of different concept pairs using statistical tests.…”

Mineração de textos biomédicos: uma revisão bibliométrica

“…Os relacionamentos implícitos são interesse de uma grande parte dos trabalhos que focam na extração de relacionamentos (TSURUOKA et al, 2011;MIYANISHI;SEKI;UEHARA, 2010;SWANSON;SMALHEISER;TORVIK, 2006). A ideia, por trás disso, é formular hipóteses potenciais para novas descobertas que ainda não foram comprovadas e publicadas.…”