Creation and aggregation of predicted data

• US 7,702,615 B1
• Filed: 11/04/2005
• Issued: 04/20/2010
• Est. Priority Date: 11/04/2005
• Status: Active Grant

• Current Assignee: X Corp. (f/k/a Twitter, Inc.) (X Holdings Corp.)
• Original Assignee: M-Factor, Inc. (International Business Machines Corporation)
• Inventors: Babcock, Brian Frederick, Shwe, Michael A., Whipkey, Chad William, Hillion, Steven John Peter, Delurgio, Phillip Dennis
• Primary Examiner(s): Pham; Khanh B
• Assistant Examiner(s): CHEEMA, AZAM M
• Application Number:
  US11/267,613
• Time in Patent Office: 1,628 Days
• Field of Search: None
• US Class Current: 707/781
• CPC Class Codes: G06F 16/283 Multi-dimensional databases or data warehouses, e.g. MOLAP or ROLAP
• CPC Class Codes: G06Q 10/04 Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem" market predictions or forecasting for commercial activities G06Q30/0202

Abstract

Methods and apparatuses for predicting set of multi-dimensional dependent data and non-measurable data from a set of multi-dimensional historical dependent and causal data are described. In one embodiment, the method comprises receiving input data that comprises multi-dimensional historical dependent data and causal data and anticipated activity data, determining a set of multi-dimensional predicted dependent data using a predictive model and the input data, creating non-measurable data based on the set of multi-dimensional predicted dependent data and the input data.

Claims

1. A computer implemented scenario analysis method of calculating multi-dimensional predicted historical or future dependent data that corresponds to a predicted or input scenario comprising:

receiving input data comprising multi-dimensional actual historical dependent data, multi-dimensional historical causal data, and a plurality of scenarios consisting of at least one of user input, predicted, and system generated activity data, wherein the actual historical dependent data includes unaggregated marketing data for a plurality of products, the multi-dimensional historical causal data comprises data affecting the multi-dimensional actual historical dependent data and may comprise data affecting the activity data, the activity data includes a plurality of predicted business drivers for each of the plurality of products and each scenario in the plurality of scenarios represents a scenario different from a scenario represented by the multi-dimensional actual historical dependent data;

calculating the set of multi-dimensional predicted historical or future dependent data using a predictive model and the input data, wherein the set of multi-dimensional predicted historical or future dependent data has the same granularity as the unaggregated marketing data included in the multi-dimensional actual historical dependent data, the activity data affects the set of multi-dimensional predicted historical or future dependent data, and each of the plurality of predicted business drivers contains at least one of price, merchandizing, advertising, distribution and competitive activity; and

calculating business metrics based on the set of multi-dimensional predicted historical or future dependent data and the input data for the plurality of scenarios, wherein said business metrics consist of at least one of business outcome measures, business efficiency measures, and business sensitivity measures.

2. The computer implemented method of claim 1, wherein the input data is associated with a product.

3. The computer implemented method of claim 1, further comprising:

generating an analytical report from the predicted historical multi-dimensional dependent data.

4. The computer implemented method of the claim 3, wherein the analytical report is selected from the group of due-to report, volume decomposition, scenario and financial information.

5. The computer implemented method of claim 4, wherein generating the analytical report comprises:

generating a due-to report by,

calculating a contribution for each of the business drivers in the set of scenario input data by varying that business driver to determine a change in the calculated set of multi-dimensional predicted dependent data.

6. The computer implemented method of claim 4, wherein the calculating a contribution for a first business driver of the business drivers comprises:

manipulating a second business driver in addition to manipulating the first business driver if the first business driver depends on the second business driver.

7. The computer implemented method of claim 4, wherein generating the analytical report comprises:

generating a volume decomposition report by,

for each of the business drivers,

determining a point where that business driver has zero contribution, and

varying that business driver to determine a volume contribution to the set of calculated multi-dimensional predicted dependent data from that business driver.

8. The computer implemented method of claim 1, further comprising:

validating the predictive model by comparing the predicted multi-dimensional historical dependent data with the multi-dimensional actual historical dependent data.

9. The computer implemented method of claim 1, further comprising:

validating the predictive model by accruing additional multi-dimensional historical dependent data and comparing the additional multi-dimensional historical dependent data with the multi-dimensional predicted dependent data.

10. The computer implemented method of claim 1, wherein the unaggregated marketing data includes sale, time, and location data for each of the plurality of products.

11. The computer implemented method of claim 1, further comprising:

calculating the activity data from the multi-dimensional actual historical dependent data and the predictive model.

12. A machine-readable storage medium having executable instructions to cause a processor to perform to method comprising:

receiving input data comprising multi-dimensional historical actual dependent data, multi-dimensional historical causal data, and a plurality of scenarios consisting of at least one of user input, predicted, and system generated activity data, wherein the historical actual dependent data includes unaggregated marketing data for a plurality of products, the multi-dimensional historical causal data comprises data affecting the multi-dimensional actual historical dependent data and may comprise data affecting the activity data, and the activity data that includes a plurality of predicted business drivers for each of the plurality of products and each scenario in the plurality of scenarios represents a scenario different from a scenario represented by the multi-dimensional actual historical dependent data;

calculating a set of multi-dimensional historical predicted dependent data using a predictive model and the input data, wherein the set of multi-dimensional predicted historical or future dependent data has the same granularity as the unaggregated marketing data, the predicted activity data affects the set of multi-dimensional predicted historical or future dependent data, and each of the plurality of business drivers contains at least one of price, merchandizing, advertising, distribution and competitive activity; and

calculating business metrics based on the set of multi-dimensional predicted historical or future dependent data and the input data for the plurality of scenarios, wherein said business metrics consist of at least one of business outcome measures, business efficiency measures, and business sensitivity measures.

13. The machine-readable storage medium of claim 12, wherein the method further comprises:

generating analytical reports from the set of predicted historical multi-dimensional dependent data.

14. The machine-readable storage medium of claim 12, wherein the method further comprises:

validating the predictive model comparing the predicted multi-dimensional historical dependent data with the multi-dimensional historical dependent data.

15. An apparatus comprising:

a memory;

a data collection module to receive input data that comprises multi-dimensional actual historical dependent data, multi-dimensional historical causal data, and a plurality of scenarios consisting of at least one of user input, predicted, and system generated activity data, wherein the actual historical dependent data includes unaggregated marketing data for a plurality of products, the multi-dimensional historical causal data comprises data affecting the multi-dimensional actual historical dependent data and may comprise data affecting the activity data, and the activity data that includes a plurality of predicted business drivers for each of the plurality of products and each scenario in the plurality of scenarios represents a scenario different from a scenario represented by the multi-dimensional actual historical dependent data;

a predictive model module to receive a predictive model;

a predictive dependent data module to calculate a set of multi-dimensional predicted historical or future dependent data using the predictive model and the input data, wherein the set of multi-dimensional predicted historical or future dependent data has the same granularity as the unaggregated marketing data, the activity data affects the set of multi-dimensional predicted historical or future dependent data, and each of the plurality of predicted business drivers contains at least one of price, merchandizing, advertising, distribution and competitive activity; and

a non-measurable data model module to calculate business metrics based on the set of multi-dimensional predicted historical or future dependent data and the input data for the plurality of scenarios, wherein said business metrics consist of at least one of business outcome measures, business efficiency measures, and business sensitivity measures.

16. The apparatus of claim 15, further comprising:

an analysis module to generate analytical reports from the predicted historical or future dependent data.

17. The apparatus of claim 15, further comprising:

a model validation module to validate the predictive model.

18. The apparatus of claim 17, wherein the validation of the predictive model is selected from the group of accruing additional multi-dimensional historical dependent data and comparing the additional multi-dimensional historical dependent data with the predicted dependent and generating predicted multi-dimensional historical dependent data from the predictive model and comparing the predicted multi-dimensional historical dependent data with the multi-dimensional historical dependent data.

19. A system comprising:

a processor;

a memory coupled to the processor though a bus; and

a process executed from the memory by the processor to cause the processor to,

receive input data comprising multi-dimensional actual historical dependent data, multi-dimensional historical causal data, and a plurality of scenarios consisting of at least one of user input, predicted, and system generated activity data, wherein the actual historical dependent data includes unaggregated marketing data for a plurality of products, the multi-dimensional historical causal data comprises data affecting the multi-dimensional actual historical dependent data and may comprise data affecting the activity data, and the activity data that includes a plurality of predicted business drivers for each of the plurality of products and each scenario in the plurality of scenarios represents a scenario different from a scenario represented by the multi-dimensional actual historical dependent data;

calculate a set of multi-dimensional predicted historical or future dependent data using a predictive model and the input data, wherein the set of multi-dimensional predicted historical or future dependent data has the same granularity as the unaggregated marketing data, the activity data affects the set of multi-dimensional predicted historical or future dependent data, and each of the plurality of predicted business drivers contains at least one of price, merchandizing, advertising, distribution and competitive activity; and

calculate business metrics based on the set of multi-dimensional predicted historical or future dependent data and the input data for the plurality of scenarios, wherein said business metrics consist of at least one of business outcome measures, business efficiency measures, and business sensitivity measures.

20. The system of claim 19, wherein the process further causes the processor to generate analytical reports from the set of multi-dimensional predicted dependent data.

21. The system of claim 19, wherein the process further causes the processor to validate the predictive model by generating predicted set of multi-dimensional historical dependent from the predictive model and comparing the predicted set of multi-dimensional historical dependent data with the set of multi-dimensional historical dependent data.

22. The system of claim 19, wherein the process further causes the processor to validate the predictive model by accruing additional set of multi-dimensional historical dependent data and comparing the additional set of multi-dimensional historical dependent data with the set of multi-dimensional predicted dependent data.