Modeling sequence and time series data in predictive analytics

• US 20060010142A1
• Filed: 04/28/2005
• Published: 01/12/2006
• Est. Priority Date: 07/09/2004
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: MacLennan, C. James, Kim, Pyungchul, Tang, ZhaoHui
• Granted Patent: US 7,747,641 B2
• Time in Patent Office: Days
• 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 16/2474 Sequence data queries, e.g. querying versioned data

Abstract

The subject invention relates to systems and methods to extend the capabilities of declarative data modeling languages. In one aspect, a declarative data modeling language system is provided. The system includes a data modeling language component that generates one or more data mining models to extract predictive information from local or remote databases. A language extension component facilitates modeling capability in the data modeling language by providing a data sequence model or a time series model within the data modeling language to support various data mining applications.

Claims

1. A declarative data modeling language system, comprising:

a data modeling language component that generates one or more data mining models to extract predictive information from at least one database; and

a language extension component that facilitates modeling capability in the data modeling language, the language extension component provides at least one of a data sequence model and a time series model in the data modeling language.

2. The system of claim 1, the data modeling language is associated with a relational database language.

3. The system of claim 2, the relational database language is a structured query language.

4. The system of claim 1, the data mining models are employed with at least one automated application that queries a database for sequence model predictions or time series model predictions.

5. The system of claim 1, the data sequence model is associated with a data sequence that includes at least one of a DNA sequence, a web sequence, a weather sequence, or a buying sequence.

6. The system of claim 5, the data sequence model is associated with a sequence clustering algorithm.

7. The system of claim 6, the sequence algorithm is employed to cluster user navigation patterns with respect to a web site.

8. The system of claim 1, the time series model is associated with continuous variables.

9. The system of claim 1, the time series model is associated with an Online Analytical Processing (OLAP) cube or a relational data structure.

10. The system of claim 1, the time series model is associated with a data component that is associated with a time parameter.

11. The system of claim 1, the data sequence model is associated with a predict sequence function.

12. The system of claim 1, the time series model is associated with a predict time series function.

13. The system of claim 1, the data sequence model or the time series model are associated with a nested table or a flat table.

14. The system of claim 13, the nested table is associated with a probability.

15. The system of claim 14, the nested table is associated with a state transition schema.

16. A computer readable medium having computer readable instructions stored thereon for implementing the components of claim 1.

17. A method for generating data mining models, comprising:

providing language extensions to a database modeling language;

automatically generating data mining models from the language extensions;

generating a query for a database; and

automatically generating sequence predictions or time series predictions from the database based on the query and the data mining models.

18. The method of claim 17, further comprising generating a session identifier key, a sequence identifier or a sequence table in response to the query.

19. The method of claim 17, further comprising generating a probability prediction base at least in part on continuous variable data, a flat rowset, or a nested rowset.

20. A system to facilitate data mining operations, comprising:

means for querying a relational database;

means for generating a data mining model to determine predictive information form the database;

means for adapting the data mining model to a data time series or a data sequence; and

means for generating probabilities from the database in view of the data time series or the data sequence.