Method and system for data mining automation in domain-specific analytic applications

A system for automated data mining using a domain-specific analytic application for solving predefined problems, the system comprising:

• means for populating input data schema, wherein said means for populating comprises
• means for reading input data from a data store and
• means for writing the input data to input data schema,
• the input data schema having a format appropriate to solution of a predefined problem;
• means for production training a predefined data mining model to produce a trained data mining model, the predefined data mining model comprising a predefined data mining model definition, production training having
• an input comprising the input data stored in the input data schema, and
• an output comprising a knowledge base,
• means for production training further comprising means for executing
• a preselected data mining algorithm in production training mode,
• means for executing the data mining algorithm in production training mode comprising means for executing a software process within the analytic application,
• the trained data mining model comprising the predefined data mining model definition and the knowledge base;
• means for production scoring input data from the input data schema, means for production scoring further comprising
• means for applying the trained data mining model through use of means for executing the data mining algorithm in production scoring mode,
• wherein the means for executing the data mining algorithm in production scoring mode comprises a software process within the analytic application,
• wherein means for executing the data mining algorithm is capable of creating an output comprising production scored data;
• means for scheduling operation of the means for populating input data schema,
• means for production training, and means for production scoring, said
• means for scheduling further comprising means for storing in computer memory a schedule;
• wherein the analytic application comprises
• the predefined problems to be solved,
• the predefined problems having referents defined in a specific computational domain;
• predefined data mining algorithms capable of using input data read from predefined input data schema for solving the predefined problems;
• predefined data schema appropriate for solution of the predefined problems, the predefined data schema further comprising the input data schema and output data schema;
• and
• at least one predefined data mining model definition, the predefined data mining model definition dependent upon the predefined data schema.

A computer program product for automated data mining by use of a domain-specific analytic application for solving predefined problems, the product further comprising:

• a recording medium;
• means, recorded on the recording medium, for populating input data schema,
• wherein said means for populating comprises
• means for reading input data from a data store and
• means for writing the input data to input data schema,
• the input data schema having a format appropriate to solution of a predefined problem;
• means, recorded on the recording medium, for production training a predefined data mining model to produce a trained data mining model, the predefined data mining model comprising a predefined data mining model definition, means for production training having:
• an input comprising the input data stored in the input data schema, and
• an output comprising a knowledge base,
• means for production training further comprising means for executing
• a preselected data mining algorithm in production training mode,
• the means for executing a preselected data mining algorithm in production training mode comprising means for executing a software program within the analytic application,
• the trained data mining model comprising the predefined data mining model definition and the knowledge base;
• means, recorded on the recording medium, for production scoring input data from the input data schema, means for production scoring further comprising
• means for applying the trained data mining model by use of means for executing the data mining algorithm in production scoring mode,
• wherein the means for executing a data mining algorithm in production scoring mode comprises a software program within the analytic application,
• wherein executing the data mining algorithm has an output comprising production scored data;
• means, recorded on the recording medium, for scheduling the operation of the means for populating input data schema, means for production training, and means for production scoring, said means for scheduling further comprising means for storing in computer memory a schedule;
• wherein the analytic application comprises
• the predefined problems to be solved,
• the predefined problems having referents defined in a specific computational domain;
• predefined data mining algorithms capable of
• using input data read from predefined input data schema for solving the predefined problems;
• predefined data schema appropriate for solution of the predefined problems, the predefined data schema further comprising the input data schema and output data schema;
• and
• at least one predefined data mining model definition, the predefined data mining model definition dependent upon the predefined data schema.