| CPC G06F 40/247 (2020.01) [G06F 40/40 (2020.01); G06N 5/022 (2013.01)] | 7 Claims |
|
1. A computer-implemented method, comprising:
identifying a plurality of initial concept groups from medical records or medical literature, wherein each initial concept group comprises clinical concepts and potential meanings of the clinical concepts;
constructing a knowledge set, using machine learning, the plurality of initial concept groups, wherein the constructing comprises:
(i) evaluating associations between the clinical concepts and the potential meanings of the clinical concepts, the associations comprising:
(a) a strength between each clinical concept and corresponding potential meanings weighted based on a token distance between occurrences of the associated concept and each potential meaning,
(b) a frequency of occurrence of each potential meaning of the clinical concepts, and
(c) directionality between associated clinical concepts, wherein the directionality indicates whether the occurrence of one clinical concept more likely leads to the occurrence of the other clinical concept than the reverse; and
(ii) filtering the plurality of initial concept groups based on the evaluation; and
storing the knowledge set into a database for deployment in disambiguating terms in an information source.
|