| CPC G16H 10/60 (2018.01) [G06F 16/904 (2019.01); G06F 16/906 (2019.01); G06F 18/24 (2023.01); G06N 20/00 (2019.01)] | 18 Claims |
|
1. A computer-implemented method comprising:
receiving, by one or more processors, a first claim for a provider;
generating, by the one or more processors and for the first claim, a plurality of predicted taxonomy-based classification scores by inputting features extracted from the first claim into one or more machine learning models and obtaining the plurality of predicted taxonomy-based classification scores from the one or more machine learning models, wherein (a) the plurality of predicted taxonomy-based classification scores comprises a predicted taxonomy-based classification score for each classification of a plurality of classifications in a taxonomy and (b) the one or more machine learning models are (i) trained with training data comprising (A) relevant features extracted from a plurality of historical claims and (B) target variables comprising the plurality of classifications in the taxonomy, (ii) trained to determine correlations among the features of the first claim, and (iii) re-trained based at least in part on a validation dataset to refine the one or more machine learning models;
identifying, by the one or more processors, a predicted taxonomy-based classification for the first claim based at least in part on the plurality of predicted taxonomy-based classification scores;
comparing, by the one or more processors, the predicted taxonomy-based classification for the provider with an assigned taxonomy-based classification for the provider, wherein the assigned taxonomy-based classification for the provider is one of (i) unknown, (ii) equivalent to the predicted taxonomy-based classification, or (iii) not equivalent to the predicted taxonomy-based classification;
automatically assigning, by the one or more processors and based at least in part on the comparing of the predicted taxonomy-based classification with the assigned taxonomy-based classification, a condition to the first claim, wherein the condition is selected from a group consisting of (i) an unknown condition assigned to any claim whose assigned taxonomy-based classification is at least one of null, empty, unknown, or not populated, (ii) a match condition assigned to any claim whose assigned taxonomy-based classification and predicted taxonomy-based classification are equivalent, and (iii) a not matched condition assigned to any claim whose assigned taxonomy-based classification and predicted taxonomy-based classification are not equivalent;
responsive to assigning the match condition to the first claim, automatically updating, by the one or more processors, a record for the provider comprising the assigned taxonomy-based classification;
automatically removing the first claim for the provider currently assigned to a review queue and reassigning the first claim to a standard queue for normal claims processing; and
providing, by the one or more processors and to a computing entity, access to the record via a user interface comprising one or more selectable user interface elements configured for facilitating display of the record.
|