Method, apparatus and computer program product for generating encounter vectors and client vectors using natural language processing models

1. An apparatus comprising at least one processor and at least one non-transitory memory comprising a computer program code, the at least one non-transitory memory and the computer program code configured to, with the at least one processor, cause the apparatus to:

retrieve one or more record data elements associated with a client identifier, wherein each of the one or more record data elements is associated with an encounter identifier of one or more encounter identifiers, wherein the one or more encounter identifiers are associated with the client identifier;

generate one or more encounter vectors based at least in part on the one or more record data elements, wherein the one or more encounter vectors are associated with the one or more encounter identifiers;

generate a client vector based at least in part on the one or more encounter vectors, wherein the client vector is associated with the client identifier, wherein a first natural language processing model is associated with at least one of the one or more encounter vectors or the client vector;

generate a prediction data element based at least in part on the client vector and a machine learning model, wherein the prediction data element is associated with the client identifier; and

perform at least one data operation based at least in part on the prediction data element.

2. The apparatus of claim 1 , wherein, when generating the one or more encounter vectors, the at least one non-transitory memory and the computer program code are configured to, with the at least one processor, cause the apparatus to:

associate the one or more record data elements with one or more record category identifiers;

associate the one or more record data elements with one or more record text descriptors; and

generate one or more record text vectors based at least in part on the one or more record text descriptors and a second natural language processing model.

3. The apparatus of claim 2 , wherein the at least one non-transitory memory and the computer program code are configured to, with the at least one processor, cause the apparatus to generate one or more record category vectors associated with the one or more record category identifiers, wherein, when generating a record category vector of the one or more record category vectors that is associated with a record category identifier of the one or more record category identifiers, the at least one non-transitory memory and the computer program code are configured to, with the at least one processor, cause the apparatus to:

identify at least one record text vector, from the one or more record text vectors, being associated with the record category identifier of the one or more record category identifiers; and

generate the record category vector associated with the record category identifier based at least in part on the at least one record text vector and a third natural language processing model.

4. The apparatus of claim 3 , wherein the at least one non-transitory memory and the computer program code are configured to, with the at least one processor, cause the apparatus to:

identify at least one record category identifier, from the one or more record category identifiers, being associated with an encounter identifier of the one or more encounter identifiers based at least in part on the one or more record data elements;

identify at least one record category vector, from the one or more record category vectors, being associated with the at least one record category identifier; and

generate at least one encounter vector associated with the encounter identifier based at least in part on the at least one record category vector.

5. The apparatus of claim 4 , wherein the at least one non-transitory memory and the computer program code are configured to, with the at least one processor, cause the apparatus to:

determine whether more than one record category vector is associated with the encounter identifier;

in response to determining that only one record category vector is associated with the at least one record category identifier, assign the only one record category vector as the at least one encounter vector; and

in response to determining that only one record category vector is associated with the encounter identifier, assign the only one record category vector as the at least one encounter vector; and

in response to determining that a plurality of record category vectors are associated with the at least one record category identifier, generate the at least one encounter vector based at least in part on the plurality of record category vectors.

6. The apparatus of claim 5 , wherein, when generating the at least one encounter vector based at least in part on the plurality of record category vectors, the at least one non-transitory memory and the computer program code are configured to, with the at least one processor, cause the apparatus to combine the plurality of record category vectors based at least in part on one or more of summation, multiplication, or one or more artificial neural networks.

7. The apparatus of claim 3 , wherein at least one of the first natural language processing model, the second natural language processing model, or the third natural language processing model is a transformer.

8. The apparatus of claim 7 , wherein at least one of the first natural language processing model, the second natural language processing model, or the third natural language processing model is a Bi-directional Encoder Representations from Transformer (BERT).

9. The apparatus of claim 3 , wherein, when generating the record category vector, the at least one non-transitory memory and the computer program code are configured to, with the at least one processor, cause the apparatus to:

select the third natural language processing model, from a plurality of natural language processing models, based at least in part on the record category identifier.

10. The apparatus of claim 2 , wherein the one or more record category identifiers comprise at least one of: a diagnosis identifier, a procedure identifier, a pharmaceutical identifier, a lab test identifier, or a lab test result identifier.

11. The apparatus of claim 2 , wherein, when associating the one or more record data elements with the one or more record text descriptors, the at least one non-transitory memory and the computer program code are configured to, with the at least one processor, cause the apparatus to:

query a network database based at least in part on the one or more record data elements.

12. The apparatus of claim 1 , wherein, when generating the one or more encounter vectors, the at least one non-transitory memory and the computer program code are configured to, with the at least one processor, cause the apparatus to:

associate the one or more record data elements with one or more record text descriptors;

generate one or more record text vectors based at least in part on the one or more record text descriptors and a second natural language processing model; and

generate the one or more encounter vectors based at least in part on the one or more record text vectors and a third natural language processing model.

13. The apparatus of claim 1 , wherein, when generating the one or more encounter vectors, the at least one non-transitory memory and the computer program code are configured to, with the at least one processor, cause the apparatus to:

associate the one or more record data elements with one or more record category identifiers; and

generate one or more record category vectors associated with the one or more record category identifiers based at least in part on the one or more record data elements.

14. The apparatus of claim 13 , wherein, when generating a record category vector of the one or more record category vectors that is associated with a record category identifier of the one or more record category identifiers, the at least one non-transitory memory and the computer program code are configured to, with the at least one processor, cause the apparatus to:

identify at least one record data element, from the one or more record data elements, being associated with the record category identifier of the one or more record category identifiers; and

generate the record category vector associated with the record category identifier based at least in part on the at least one record data element and a second natural language processing model.

15. The apparatus of claim 13 , wherein the at least one non-transitory memory and the computer program code are configured to, with the at least one processor, cause the apparatus to:

identify at least one record category identifier, from the one or more record category identifiers, being associated with an encounter identifier of the one or more encounter identifiers based at least in part on the one or more record data elements;

identify at least one record category vector, from the one or more record category vectors, being associated with the at least one record category identifier; and

generate an encounter vector associated with the encounter identifier based at least in part on the at least one record category vector.

16. The apparatus of claim 15 , wherein the at least one non-transitory memory and the computer program code are configured to, with the at least one processor, cause the apparatus to:

determine whether more than one record category vector is associated with the encounter identifier;

in response to determining that only one record category vector is associated with the encounter identifier, assign the only one record category vector as the encounter vector; and

in response to determining that a plurality of record category vectors are associated with the at least one record category identifier, generate the encounter vector based at least in part on combining the plurality of record category vectors.

17. The apparatus of claim 1 , wherein, when generating the client vector, the at least one non-transitory memory and the computer program code are configured to, with the at least one processor, cause the apparatus to:

generate the client vector based at least in part on the one or more encounter identifiers and the one or more encounter vectors using the first natural language processing model.

18. The apparatus of claim 17 , wherein the at least one non-transitory memory and the computer program code are configured to, with the at least one processor, cause the apparatus to:

calculate one or more encounter date metadata based at least in part on the one or more encounter identifiers and a base date.

19. A computer-implemented method comprising:

retrieving one or more record data elements associated with a client identifier, wherein each of the one or more record data elements is associated with an encounter identifier of one or more encounter identifiers, wherein the one or more encounter identifiers are associated with the client identifier;

generating one or more encounter vectors based at least in part on the one or more record data elements, wherein the one or more encounter vectors are associated with the one or more encounter identifiers;

generating a client vector based at least in part on the one or more encounter vectors, wherein the client vector is associated with the client identifier, wherein a first natural language processing model is associated with at least one of the one or more encounter vectors or the client vector;

generating a prediction data element based at least in part on the client vector and a machine learning model, wherein the prediction data element is associated with the client identifier; and

performing at least one data operation based at least in part on the prediction data element.

20. A computer program product comprising at least one non-transitory computer-readable storage medium having computer-readable program code portions stored therein, the computer-readable program code portions comprising an executable portion configured to:

retrieve one or more record data elements associated with a client identifier, wherein each of the one or more record data elements is associated with an encounter identifier of one or more encounter identifiers, wherein the one or more encounter identifiers are associated with the client identifier;

generate one or more encounter vectors based at least in part on the one or more record data elements, wherein the one or more encounter vectors are associated with the one or more encounter identifiers;

generate a client vector based at least in part on the one or more encounter vectors, wherein the client vector is associated with the client identifier, wherein a first natural language processing model is associated with at least one of the one or more encounter vectors or the client vector;

generate a prediction data element based at least in part on the client vector and a machine learning model, wherein the prediction data element is associated with the client identifier; and

perform at least one data operation based at least in part on the prediction data element.