Commentary - Journal of Neuroinformatics and Neuroimaging (2021) Volume 6, Issue 3
A text mining pipeline for curating information in computational neuroscience using active and deep learning.
Curation of neuroscience things is critical to current work in neuroinformatics and computational neuroscience, such as those used in large-scale brain modelling projects. Manually combing through hundreds of articles in search of fresh information on modelled entities, on the other hand, is a time-consuming and low-reward activity. Text mining can aid a curator in extracting important data from this material in a methodical manner. The use of text mining algorithms to the neuroscience literature is proposed. Two computational neuroscientists used active learning techniques to annotate a corpus of items relevant to neuroscience, allowing for quick, focused annotation.Author(s): Shawn Kruger