Earning Code Classification

1 . A computer-implemented method for classifying and applying human resources data, the method comprising:

aggregating, by a number of processors, employee transaction data for an organization;

evaluating, by a number of processors, a number of human resources-related attributes across heterogeneous transaction data;

classifying, by a number of processors via statistical machine learning, the employee transaction data into a number of normalized codes according to the human resources-related attributes; and

presenting, by a display, a user interface to adjust a number of organizational operating procedures according to the normalized codes.

2 . The method of claim 1 , wherein the machine learning further comprises:

labelling the compensation data by independent first and second labelers according to code descriptions, wherein label agreements between the first and second labelers form a first labeled dataset;

resolving label disagreements between the first and second labelers by a third labeler to form a second labeled dataset;

labelling the compensation data with a semantic matcher to form a third labeled dataset;

combining the first, second, and third labeled datasets into a final labeled dataset; and

applying the final labeled dataset to a number of machine learning algorithms.

3 . The method of claim 2 , wherein machine learning algorithms comprise at least one of:

naïve Bayes;

logistic regression;

fully connected neural network;

distributed random forest;

gradient boosting machine; or

XG boost.

4 . The method of claim 1 , wherein the transaction data is labeled according to code descriptions.

5 . The method of claim 1 , further comprising deriving, by a number of processors, a number of transaction patterns relative to the normalized codes.

6 . The method of claim 5 , wherein the transaction patterns comprise at least one of:

employee ratio;

compensation frequency;

earning amount;

job category;

compensation rate type;

employee seniority;

working hours; and

employee age.

7 . The method of claim 1 , wherein the normalized codes comprise at one of the following:

pay codes;

benefits codes;

health care costs; or

deduction codes.

8 . The method of claim 1 , wherein adjusting organizational operating procedures according to the normalized codes comprises at least one of:

adjusting resource allocation;

employee compensation setup.

9 . The method of claim 1 , further comprising benchmarking the organization according to the normalized codes and sector.

10 . A system for classifying and applying human resources data, the system comprising:

a bus system;

a storage device connected to the bus system, wherein the storage device stores program instructions; and

a number of processors connected to the bus system, wherein the number of processors execute the program instructions to:

aggregate employee transaction data for an organization;

evaluate a number of human resources-related attributes across heterogeneous transaction data;

classify, via statistical machine learning, the employee transaction data into a number of normalized codes according to the human resources-related attributes; and

present a user interface to adjust a number of organizational operating procedures according to the normalized codes.

11 . The system of claim 10 , wherein the machine learning further comprises:

labelling the compensation data by independent first and second labelers according to code descriptions, wherein label agreements between the first and second labelers form a first labeled dataset;

resolving label disagreements between the first and second labelers by a third labeler to form a second labeled dataset;

labelling the compensation data with a semantic matcher to form a third labeled dataset;

combining the first, second, and third labeled datasets into a final labeled dataset; and

applying the final labeled dataset to a number of machine learning algorithms.

12 . The system of claim 11 , wherein machine learning algorithms comprise at least one of:

naïve Bayes;

logistic regression;

fully connected neural network;

distributed random forest;

gradient boosting machine; or

XG boost.

13 . The system of claim 10 , wherein the transaction data is labeled according to code descriptions.

14 . The system of claim 10 , wherein the number of processors further execute program instructions to derive a number of transaction patterns relative to the normalized codes.

15 . The system of claim 14 , wherein the transaction patterns comprise at least one of:

employee ratio;

compensation frequency;

earning amount;

job category;

compensation rate type;

employee seniority;

working hours; and

employee age.

16 . The system of claim 10 , wherein the normalized codes comprise at one of the following:

pay codes;

benefits codes;

health care costs; or

deduction codes.

17 . The system of claim 10 , wherein adjusting organizational operating procedures according to the normalized codes comprises at least one of:

adjusting resource allocation;

employee compensation setup.

18 . The system of claim 10 , wherein the number of processors further execute program instructions to benchmark the organization according to the normalized codes and sector.

19 . A computer program product for classifying and applying human resources data, the computer program product comprising:

a non-volatile computer readable storage medium having program instructions embodied therewith, the program instructions executable by a number of processors to cause the computer to perform the steps of:

aggregating employee transaction data for an organization;

evaluating a number of human resources-related attributes across heterogeneous transaction data;

classifying, via statistical machine learning, the employee transaction data into a number of normalized codes according to the human resources-related attributes; and

presenting a user interface to adjust a number of organizational operating procedures according to the normalized codes.

20 . The computer program product according to claim 19 , wherein the machine learning further comprises:

labelling the compensation data by independent first and second labelers according to code descriptions, wherein label agreements between the first and second labelers form a first labeled dataset;

resolving label disagreements between the first and second labelers by a third labeler to form a second labeled dataset;

labelling the compensation data with a semantic matcher to form a third labeled dataset;

combining the first, second, and third labeled datasets into a final labeled dataset; and

applying the final labeled dataset to a number of machine learning algorithms.

21 . The computer program product according to claim 20 , wherein machine learning algorithms comprise at least one of:

naïve Bayes;

logistic regression;

fully connected neural network;

distributed random forest;

gradient boosting machine; or

XG boost.

22 . The computer program product according to claim 19 , wherein the transaction data is labeled according to code descriptions.

23 . The computer program product according to claim 19 , further comprising deriving, by a number of processors, a number of transaction patterns relative to the normalized codes.

24 . The computer program product according to claim 23 , wherein the transaction patterns comprise at least one of:

employee ratio;

compensation frequency;

earning amount;

job category;

compensation rate type;

employee seniority;

working hours; and

employee age.

25 . The computer program product according to claim 19 , wherein the normalized codes comprise at one of the following:

pay codes;

benefits codes;

health care costs; or

deduction codes.

26 . The computer program product according to claim 19 , wherein adjusting organizational operating procedures according to the normalized codes comprises at least one of:

adjusting resource allocation;

employee compensation setup.

27 . The computer program product according to claim 19 , further comprising benchmarking the organization according to the normalized codes and sector.