has been developed to identify articles as relevant citations in the SenseLab databases CellPropDB and NeuronDB by scanning the Natural
Language text of neuroscience articles. The program uses
a knowledgebase consistent with the SenseLab architecture, contextual and lexical scanners,
and two independent evolution methods to score articles based on relevance of deposition. NeuroText uses a two step
approach. The first step is aimed at specificity; the second at sensitivity.
The Journal of Neuroscience was used in this pilot study.