SYSTEM, METHOD AND COMPUTER PROGRAM FOR MULTI-DIMENSIONAL TEMPORAL DATA MINING

• US 20120166484A1
• Filed: 07/22/2010
• Published: 06/28/2012
• Est. Priority Date: 07/22/2009
• Status: Active Grant

• Current Assignee: University Of Ontario Institute Of Technology
• Original Assignee: University Of Ontario Institute Of Technology
• Inventors: McGregor, Carlolyn Patricia
• Granted Patent: US 8,583,686 B2
• Time in Patent Office: Days
• US Class Current: 707/776
• CPC Class Codes: G06N 5/022 Knowledge engineering; Knowledge acquisition
• CPC Class Codes: G16H 50/20 for computer-aided diagnosis, e.g. based on medical expert systems
• CPC Class Codes: G16H 50/70 for mining of medical data, e.g. analysing previous cases of other patients
• CPC Class Codes: G16Z 99/00 Subject matter not provided for in other main groups of this subclass

Abstract

The present invention provides a system, method and computer program for multi-dimensional temporal abstraction and data mining. The invention comprises collecting and optionally cleaning multi-dimensional data, the multi-dimensional data including a plurality of data streams; temporally abstracting the multi-dimensional data; and relatively aligning the temporally abstracted multi-dimensional data based on a shared time point of interest.

Claims

1. A method for multi-dimensional temporal abstraction and data mining, the method characterized by:

collecting and optionally cleaning multi-dimensional data, the multi-dimensional data including a plurality of data streams;

temporally abstracting the multi-dimensional data; and

relatively aligning the temporally abstracted multi-dimensional data based on an at least one time point of interest.

2. The method of claim 1, wherein the collected, temporally abstracted and relatively aligned data is stored in a data store for subsequent retrieval.

3. The method of claim 1, further characterized by engaging in exploratory data mining of the relatively aligned temporally abstracted multi-dimensional data.

4. The method of claim 3, wherein the exploratory data mining step includes a mechanism for enabling a user to generate rules for use by an intelligent monitoring system.

5. The method of claim 1, further characterized by linking an at least one remote device to the temporally abstracted multi-dimensional data to enable further processing of the data by the at least one remote device.

6. The method of claim 1, wherein the temporal abstraction step includes the steps of retrieving an at least one temporal rule, and applying the at least one rule to the multi-dimensional data to create temporally abstracted multi-dimensional data.

7. The method of claim 1, wherein the temporally abstracted multi-dimensional data is characterized by a plurality of data values, each of the plurality of data values having an actual start time and an actual end time.

8. The method of claim 7, wherein the relative alignment step includes the step of calculating the difference between the actual times and the at least one time point of interest for each of the plurality of data values.

9. The method of claim 1, wherein the at least one time point of interest is an event.

10. The method of claim 1, wherein the relative alignment step includes the steps of retrieving the temporally abstracted multi-dimensional data from a data store based on selection specifications provided by a user, and applying transformations to the temporally abstracted multi-dimensional data.

11. A computer implemented data mining system characterized by:

an at least one data store;

a processor in communication with the at least one data store, the processor configured to;

collect and construct from the data store an at least one time dependent data set using selection criteria set by a user;

apply temporal abstractions to the at least one time dependent data set in accordance with pre-defined abstraction rules to produce an at least one temporally abstracted data set; and

re-align the at least one temporally abstracted data set relative to an at least one time point of interest to produce an at least one relatively aligned data set.

12. The system of claim 11, wherein the at least one time dependent data set, the at least one temporally abstracted data set, and the at least one relatively aligned data set are stored for subsequent retrieval.

13. The system of claim 11, wherein the processor is further configured to engage in exploratory data mining of the at least one relatively aligned data set.

14. The system of claim 13, wherein the exploratory data mining process includes a mechanism for enabling a user to generate rules for use by an intelligent monitoring system.

15. The system of claim 11, wherein the process is further configured to link an at least one remote device to the at least one temporally abstracted data set to enable further processing of the data set by the at least one remote device.

16. The system of claim 11, wherein the application of temporal abstractions includes the steps of retrieving an at least one temporal rule, and applying the at least one rule to the at least one time dependent data set to create an at least one temporally abstracted data set.

17. The system of claim 11, wherein the at least one temporally abstracted data set is characterized by a plurality of data values, each of the plurality of data values having an actual start time and an actual end time.

18. The system of claim 17, wherein re-alignment of the at least one temporally abstracted data set is characterized by calculating the difference between the actual times and the at least one time point of interest for each of the plurality of data values.

19. The system of claim 11, wherein the at least one time point of interest is an event.

20. The system of claim 11, wherein re-alignment of the at least one temporally abstracted data set is characterized by retrieving the temporally abstracted multi-dimensional data from a data store based on selection specifications provided by a user, and applying transformations to the temporally abstracted multi-dimensional data.

21. A computer readable medium having stored thereon a computer program for data mining, the computer program comprising a set of instructions for generating and storing a plurality of accessible information files when used with a processor, the set of instructions comprising a method characterized by:

constructing an at least one time dependent data set using selection criteria set by a user;

applying temporal abstractions to the at least one time dependent data set in accordance with pre-defined abstraction rules to produce an at least one temporally abstracted data set;

relatively aligning the at least one temporally abstracted data set relative to a time point of interest to create an at least one relatively aligned data set; and

storing each at least one data set for subsequent retrieval.