Cognitive claims processing
In some embodiments, a processor may receive a claim, where the claim relates to an event. The processor may identify characteristics of the claim. The processor may determine, using an AI model, an alert reason based on a similarity of the characteristics of the claim to characteristics of one or more other claims. The processor may generate an alert notification, and the processor may output the alert notification to a user.
1. A computer-implemented method, the method comprising:
receiving, by a processor, a claim, wherein the claim is related to an event;
identifying characteristics of the claim;
determining, using a plurality of artificial intelligence (AI) models trained on a plurality of previous claims, respective ones of the previous claims having a set of characteristics, an alert reason based on a similarity of the characteristics of the claim to characteristics of one or more previous claims of the plurality of previous claims, wherein the plurality of AI models comprises a first AI model and a second AI model that are each configured to generate a classification for a different aspect of the claim, and wherein the plurality of AI models further comprises an ensemble model configured to aggregate classifications from the first and second AI models;
generating an alert notification;
outputting the alert notification to a user;
receiving feedback from the user confirming the alert notification; and
updating the first AI model from the plurality of AI models with the feedback,
wherein the feedback increases an accuracy of subsequent alert notifications.
2. The method of claim 1 , wherein the plurality of AI models are trained to generate the alert notification based on a similarity of the claim to other claims based on a combination of the alert reason and the characteristics of the claim.
3. The method of claim 1 , wherein the alert notification output includes a similarity score, wherein the similarity score measures the similarity of the claim to one or more of the previous claims .
4. The method of claim 3 , wherein the alert notification output includes a claim characteristics identifier and an alert reason identifier, wherein the claim characteristics identifier indicates a category of claim characteristics that led to generating the alert notification and the alert reason identifier indicates a category of alert reasons that led to generating the alert notification.
5. The method of claim 3 , wherein generating the alert notification includes identifying that the similarity score exceeds a similarity threshold.
6. The method of claim 1 , wherein the plurality of AI models includes:
an AI model configured to classify a first risk of the claim based on an individual associated with the claim within the claim characteristics;
an AI model configured to classify a second risk of the claim based on text within a textual explanation of the claim within the claim characteristics;
an AI model configured to classify a third risk of the claim based on details of a vehicle associated with the claim within the claim characteristics; and
the ensemble model configured to aggregate the first, second, and third classified risks.
7. The method of claim 6 , wherein the ensemble model calculates a risk score for the claim based on aggregating the first, second, and third classified risks.
8. The method of claim 1 , wherein the alert reason is output according to a narrative template that replaces variables with model outputs.
9. A system comprising:
a memory; and
a processor in communication with the memory, the processor being configured to perform operations comprising:
receiving a claim, wherein the claim is related to an event;
identifying characteristics of the claim;
determining, using a plurality of artificial intelligence (AI) models trained on a plurality of previous claims , respective ones of the previous claims having a set of characteristics, an alert reason based on a similarity of the characteristics of the claim to characteristics of one or more previous claims of the plurality of previous claims , wherein the plurality of AI models comprises a first AI model and a second AI model that are each configured to generate a classification for a different aspect of the claim, and wherein the plurality of AI models further comprises an ensemble model configured to aggregate classifications from the first and second AI models;
generating an alert notification;
outputting the alert notification to a user;
receiving feedback from the user confirming the alert notification; and
updating the first AI model from the plurality of AI models with the feedback,
wherein the feedback increases an accuracy of subsequent alert notifications.
10. The system of claim 9 , wherein the plurality of AI models are trained to generate the alert notification based on a similarity of the claim to other claims based on a combination of the alert reason and the characteristics of the claim.
11. The system of claim 9 , wherein the alert notification output includes a similarity score, wherein the similarity score measures the similarity of the claim to one or more of the previous claims .
12. The system of claim 11 , wherein the alert notification output includes a claim characteristics identifier and an alert reason identifier, wherein the claim characteristics identifier indicates a category of claim characteristics that led to generating the alert notification and the alert reason identifier indicates a category of alert reasons that led to generating the alert notification.
13. The system of claim 11 , wherein generating the alert notification includes identifying that the similarity score exceeds a similarity threshold.
14. A computer program product comprising a computer readable storage medium having program instructions embodied therewith, the program instructions executable by a processor to cause the processor to perform operations, the operations comprising:
receiving a claim, wherein the claim is related to an event;
identifying characteristics of the claim;
determining, using a plurality of artificial intelligence (AI) models trained on a plurality of previous claims , respective ones of the previous claims having a set of characteristics, an alert reason based on a similarity of the characteristics of the claim to characteristics of one or more previous claims of the plurality of previous claims , wherein the plurality of AI models comprises a first AI model and a second AI model that are each configured to generate a classification for a different aspect of the claim, and wherein the plurality of AI models further comprises an ensemble model configured to aggregate classifications from the first and second AI models;
generating an alert notification;
outputting the alert notification to a user;
receiving feedback from the user confirming the alert notification; and
updating the first AI model from the plurality of AI models with the feedback,
wherein the feedback increases an accuracy of subsequent alert notifications.
15. The computer program product of claim 14 , wherein the plurality of AI models are trained to generate the alert notification based on a similarity of the claim to other claims based on a combination of the alert reason and the characteristics of the claim.
16. The computer program product of claim 14 , wherein the alert notification output includes a similarity score, wherein the similarity score measures the similarity of the claim to one or more of the previous claims .
17. The computer program product of claim 16 , wherein the alert notification output includes a claim characteristics identifier and an alert reason identifier, wherein the claim characteristics identifier indicates a category of claim characteristics that led to generating the alert notification and the alert reason identifier indicates a category of alert reasons that led to generating the alert notification.
18. The computer program product of claim 16 , wherein generating the alert notification includes identifying that the similarity score exceeds a similarity threshold.