Systems and methods that facilitate data mining

• US 7,398,268 B2
• Filed: 02/02/2005
• Issued: 07/08/2008
• Est. Priority Date: 07/09/2004
• Status: Expired due to Fees

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Crivat, Ioan Bogdan, MacLennan, C. James, Gorbach, Irina G., Iyer, Raman S., Kim, Pyungchul, Tang, ZhaoHui
• Primary Examiner(s): Alam; Hosain T
• Assistant Examiner(s): CHEEMA, AZAM M
• Application Number:
  US11/049,031
• Publication Number:
  US 20060010110A1
• Time in Patent Office: 1,252 Days
• Field of Search: None
• US Class Current: 1/1
• CPC Class Codes: G06F 16/2465 Query processing support for facilitating data mining operations in structured databases
• CPC Class Codes: G06F 16/284 Relational databases
• CPC Class Codes: G06Q 40/00 Finance; Insurance; Tax strategies; Processing of corporate or income taxes
• CPC Class Codes: Y10S 707/99933 Query processing, i.e. searching

Abstract

A system that facilitates data mining comprises a reception component that receives command(s) in a declarative language that relate to utilizing an output of a first data mining model as an input to a second data mining model. An implementation component analyzes the received command(s) and implements the command(s) with respect to the first and second data mining models. In another aspect of the subject invention, the reception component can receive further command(s) in a declarative language with respect to causing one or more of the first and second data mining models to output a prediction, the prediction desirably generated without prediction input, the implementation component causes the one or more of the first and second data mining models to output the prediction.

Claims

1. A system that facilitates data mining comprising a computer processor for executing the following software components, the system is recorded on a computer-readable storage medium and capable of execution by a computer, comprising:

a reception component that receives command(s) in a declarative language that relate to communicatively coupling together two disparate data mining models via utilizing an output of a first data mining model as an input to a second data mining model; and

an implementation component that analyzes the received command(s) and implements the command(s) with respect to the first and second data mining models;

wherein the reception component receives further command(s) in a declarative language with respect to causing one or more of the first and second data mining models to output a prediction, the prediction generated without prediction input, the implementation component causes the one or more of the first and second data mining models to output the prediction utilizing training data and global statistics associated therewith; and

wherein the declarative language is Data Mining Extensions (DMX).

2. The system of claim 1, the reception component receives further command(s) in a declarative language with respect to causing a prediction to be output from one or more of the first and second data mining models, the command(s) are a singular command that can cause output of a plurality of prediction types, the implementation component analyzes the one or more of the first and second data mining models and causes a prediction to be generated with a prediction type that corresponds to the analysis of the one or more of the first and second data mining models.

3. The system of claim 2, the prediction type is related to one or more of a time-series prediction, a sequential prediction, an associative prediction, and a regression prediction.

4. The system of claim 1, the received command(s) are generated by a user.

5. The system of claim 1, the received command(s) are generated by a computer component.

6. The system of claim 1, the implementation component implements the command(s) upon a server, a result of the implementation returned to a client.

7. The system of claim 1, the command(s) received and implemented by way of a data mining object model.

8. A method that facilitates data mining, comprising:

receiving a first data mining model;

receiving command(s) in a declarative language that relate to communicatively coupling together two disparate data mining models via utilizing an output of the first data mining model as an input to a second data mining model;

analyzing the received command(s) and implementing the command(s) with respect to the first and second data mining models;

receiving further command(s) in a declarative language with respect to causing one or more of the first and second data mining models to output a prediction, the prediction generated without prediction input, wherein the command(s) enable the first and second data mining models to utilize only training data and global statistics associated therewith in connection with generating the prediction, and wherein the declarative language is Data Mining Extensions (DMX); and

generating the prediction upon receipt of the command(s).

9. The method of claim 8, the declarative language is based at least in part upon Structured Query Language (SQL).

10. The method of claim 8, the generated prediction is a time-series prediction.

11. The method of claim 8, the generated prediction is an associative prediction.

12. The method of claim 8, further comprising:

analyzing parameters associated with the data mining model; and

determining a prediction type as a function of parameters associated with the data mining model.

13. The method of claim 12, further comprising analyzing data type associated with the data mining model.

14. A system that facilitates data mining comprising a computer processor for executing the following software components, the system is recorded on a computer-readable storage medium and capable of execution by a computer, comprising:

means for reviewing content of a data mining model by way of a data mining object model;

means for performing operations upon the content of the data mining model at a sewer that stores the data mining model;

means for receiving command(s) in a declarative language that relate to communicatively coupling together two disparate data mining models via utilizing an output of a first data mining model as an input to a second data mining model;

means for analyzing the received command(s) and implementing the command(s) with respect to the first and second data mining models;

means for receiving further command(s) in a declarative language that cause the first and second data mining models to output a prediction, the prediction generated without prediction input, wherein the command(s) enable the first and second data mining models to utilize training data and global statistics associated therewith in connection with generating the prediction, and wherein the declarative language is Data Mining Extensions (DMX); and

means for generating the prediction upon receipt of the command(s).

15. The system of claim 14, further comprising means for delivering results of the operation from the server to a client application.

16. The system of claim 14, further comprising means for generating a graphical rendering of the operation at a client computer.