-
Notifications
You must be signed in to change notification settings - Fork 2
Text Mining Provider
The Text Mining Provider provides multiple KPs to the Translator ecosystem that are derived from information mined from the available biomedical literature. The underlying natural language processing pipelines identify mentions of biomedical concepts in scientific text as well as assertions of relations between concept. Assertions of the relations are served by the Text-mined assertion KP. The concept mentions are also used to compute various literature cooccurrence metrics, which are made available through the Literature cooccurrence KP.
Bill Baumgartner <william . baumgartner AT cuanschutz . edu>
Text-mined assertion KP: Provides text-mined assertions between pairs of entities grounded to biomedical ontologies/terminologies. Node types include chemicals/drugs, diseases/phenotypes, and genes/proteins. Predicates are modeled using Biolink qualifiers and serve to represent the following relationships: treats, causes_or_contributes_to, positively_regulates, and negatively_regulates.
Literature cooccurrence KP: Provides links between pairs of biomedical entities that have been mentioned together in the biomedical literature. Six different cooccurrence metrics are computed on four different document levels, e.g., abstract, paragraph, sentence, title.