AUTOMATIC GENERATION OF NEW RULES FOR PROCESSING SYNTHETIC EVENTS USING COMPUTER-BASED LEARNING PROCESSES

• US 20090024553A1
• Filed: 10/01/2008
• Published: 01/22/2009
• Est. Priority Date: 10/03/2006
• Status: Active Grant

• Current Assignee: ServiceNow Incorporated
• Original Assignee: International Business Machines Corporation
• Inventors: Kraemer, James R., Angell, Robert Lee, Friedlander, Robert R.
• Granted Patent: US 8,055,603 B2
• Time in Patent Office: Days
• US Class Current: 706/47
• CPC Class Codes: G06N 5/025 Extracting rules from data
• CPC Class Codes: Y02A 90/10 Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation

Abstract

A computer implemented method. A first synthetic event is received. The first synthetic event is derived from a first cohort comprising a first set of data and a second cohort comprising a second set of data. The first synthetic event comprises a third set of data representing a result of a mathematical computation on the first and second cohorts. A first rule set is created, the first synthetic event being expected as a result of application of the first rule set to the first cohort and the second cohort. The first rule set is applied to the first cohort and the second cohort to achieve a first result. The first result comprises a second event that is compared to the first synthetic event. A comparison is formed, the comparison comprising additional data that can be used to describe a difference between the second event and the first synthetic event.

Claims

1. A computer implemented method comprising:

receiving a first synthetic event, wherein the first synthetic event is derived from a first cohort comprising a first set of data and a second cohort comprising a second set of data, wherein the first synthetic event comprises a third set of data representing a result of a mathematical computation defined by the operation S(p1)==>

F(p2), wherein S comprises a set of input facts with probability p1, wherein the set of input facts comprise the first cohort and the second cohort, wherein F comprises an inferred event with probability p2, wherein the term “

event”

means a particular set of data that represents, encodes, or records at least one of a thing or happening, and wherein each of the first set of data, the second set of data, the first cohort, the second cohort, the first synthetic event, and subcomponents thereof all comprise different events;

creating a first rule set, wherein the first synthetic event is expected as a result of application of the first rule set to the first cohort and the second cohort;

applying the first rule set to the first cohort and the second cohort, wherein a first result is achieved, and wherein the first result comprises a second event;

comparing the second event to the first synthetic event, wherein a comparison is formed, the comparison comprising additional data that can be used to describe a difference between the second event and the first synthetic event; and

storing the comparison.

2. The computer implemented method of claim 1 wherein the second event is different than the first synthetic event, and wherein the method further comprises:

after comparing, modifying the first rule set to form a second rule set, wherein modifying is based at least in part on the comparison;

applying the second rule set to the first cohort and the second cohort, wherein a second result is achieved, and wherein the second result comprises a third event;

comparing the third event to the first synthetic event, wherein a second comparison is formed, the second comparison comprising further data that can be used to describe a second difference between the third event and the first synthetic event; and

storing the second comparison.

3. The computer implemented method of claim 2 wherein modifying comprises deleting a first rule from the first rule set, wherein existence of the first rule reduces a probability of an approximate match between the second event and the first synthetic event.

4. The computer implemented method of claim 3 wherein selection of the first rule is accomplished by using one of a neural network and a genetic algorithm.

5. The computer implemented method of claim 2 wherein modifying comprises modifying a first rule from the first rule set, wherein modification of the first rule is calculated to increase a probability of an approximate match between the second event and the first synthetic event.

6. The computer implemented method of claim 2 wherein modifying comprises adding an additional rule to the first rule set, wherein addition of the additional rule is calculated to increase a probability of an approximate match between the second event and the first synthetic event.

7. The computer implemented method of claim 2 further comprising:

iteratively performing modifying the first rule set, applying the second rule set, comparing the third event, and storing the second comparison, wherein at each iteration a different rule set, a different event, and a different comparison are created, and wherein iteration continues until a last event is approximately equivalent to the first synthetic event.

8. The computer implemented method of claim 1 wherein creating the first rule set is accomplished by using a neural network to suggest rules that will allow creation of the first synthetic event from the first cohort and the second cohort.

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

analyzing the first rule set to determine a causal relationship between the first synthetic event and at least one datum of one of the first cohort and the second cohort.

10. The computer implemented method of claim 1 further comprising:

after comparing, receive a third cohort;

after comparing, modifying the first rule set to form a second rule set, wherein modifying is based at least in part on the comparison and the third cohort;

applying the second rule set to the first cohort, the second cohort, and the third cohort, wherein a second result is achieved, and wherein the second result comprises a third event;

comparing the third event to the first synthetic event, wherein a second comparison is formed, the second comparison comprising further data that can be used to describe a second difference between the third event and the first synthetic event; and

storing the second comparison.

11. A computer program product stored on a recordable-type medium, the computer program product comprising:

instructions for receiving a first synthetic event, wherein the first synthetic event is derived from a first cohort comprising a first set of data and a second cohort comprising a second set of data, wherein the first synthetic event comprises a third set of data representing a result of a mathematical computation defined by the operation S(p1)==>

F(p2), wherein S comprises a set of input facts with probability p1, wherein the set of input facts comprise the first cohort and the second cohort, wherein F comprises an inferred event with probability p2, wherein the term “

event”

means a particular set of data that represents, encodes, or records at least one of a thing or happening, and wherein each of the first set of data, the second set of data, the first cohort, the second cohort, the first synthetic event, and subcomponents thereof all comprise different events;

instructions for creating a first rule set, wherein the first synthetic event is expected as a result of application of the first rule set to the first cohort and the second cohort;

instructions for applying the first rule set to the first cohort and the second cohort, wherein a first result is achieved, and wherein the first result comprises a second event;

instructions for comparing the second event to the first synthetic event, wherein a comparison is formed, the comparison comprising additional data that can be used to describe a difference between the second event and the first synthetic event; and

instructions for storing the comparison.

12. The computer program product of claim 11 wherein the second event is different than the first synthetic event, and wherein the computer program product further comprises:

instructions for, after comparing, modifying the first rule set to form a second rule set, wherein modifying is based at least in part on the comparison;

instructions for applying the second rule set to the first cohort and the second cohort, wherein a second result is achieved, and wherein the second result comprises a third event;

instructions for comparing the third event to the first synthetic event, wherein a second comparison is formed, the second comparison comprising further data that can be used to describe a second difference between the third event and the first synthetic event; and

instructions for storing the second comparison.

13. The computer program product of claim 12 further comprising:

instructions for iteratively performing modifying the first rule set, applying the second rule set, comparing the third event, and storing the second comparison, wherein at each iteration a different rule set, a different event, and a different comparison are created, and wherein iteration continues until a last event is approximately equivalent to the first synthetic event.

14. The computer program product of claim 11 further comprising:

instructions for analyzing the first rule set to determine a causal relationship between the first synthetic event and at least one datum of one of the first cohort and the second cohort.

15. The computer program product of claim 11 further comprising:

instructions for, after comparing, receive a third cohort;

instructions for after comparing, modifying the first rule set to form a second rule set, wherein modifying is based at least in part on the comparison and the third cohort;

instructions for applying the second rule set to the first cohort, the second cohort, and the third cohort, wherein a second result is achieved, and wherein the second result comprises a third event;

instructions for comparing the third event to the first synthetic event, wherein a second comparison is formed, the second comparison comprising further data that can be used to describe a second difference between the third event and the first synthetic event; and

instructions for storing the second comparison.

16. A data processing system comprising:

a bus;

a processor connected to the bus;

a memory connected to the bus, wherein the memory contains a set of instructions, and wherein the processor can execute the set of instructions to;

receive a first synthetic event, wherein the first synthetic event is derived from a first cohort comprising a first set of data and a second cohort comprising a second set of data, wherein the first synthetic event comprises a third set of data representing a result of a mathematical computation defined by the operation S(p1)==>

F(p2), wherein S comprises a set of input facts with probability p1, wherein the set of input facts comprise the first cohort and the second cohort, wherein F comprises an inferred event with probability p2, wherein the term “

event”

means a particular set of data that represents, encodes, or records at least one of a thing or happening, and wherein each of the first set of data, the second set of data, the first cohort, the second cohort, the first synthetic event, and subcomponents thereof all comprise different events;

create a first rule set, wherein the first synthetic event is expected as a result of application of the first rule set to the first cohort and the second cohort;

apply the first rule set to the first cohort and the second cohort, wherein a first result is achieved, and wherein the first result comprises a second event;

compare the second event to the first synthetic event, wherein a comparison is formed, the comparison comprising additional data that can be used to describe a difference between the second event and the first synthetic event; and

store the comparison.

17. The data processing system of claim 16 wherein the second event is different than the first synthetic event, and wherein the processor can further execute the set of instructions to:

after comparing, modify the first rule set to form a second rule set, wherein modifying is based at least in part on the comparison;

apply the second rule set to the first cohort and the second cohort, wherein a second result is achieved, and wherein the second result comprises a third event;

compare the third event to the first synthetic event, wherein a second comparison is formed, the second comparison comprising further data that can be used to describe a second difference between the third event and the first synthetic event; and

store the second comparison.

18. The data processing system of claim 17, wherein the processor can further execute the set of instructions to:

iteratively perform modifying the first rule set, applying the second rule set, comparing the third event, and storing the second comparison, wherein at each iteration a different rule set, a different event, and a different comparison are created, and wherein iteration continues until a last event is approximately equivalent to the first synthetic event.

19. The data processing system of claim 16, wherein the processor can further execute the set of instructions to:

analyze the first rule set to determine a causal relationship between the first synthetic event and at least one datum of one of the first cohort and the second cohort.

20. The data processing system of claim 16, wherein the processor can further execute the set of instructions to:

after comparing, receive a third cohort;

after comparing, modify the first rule set to form a second rule set, wherein modifying is based at least in part on the comparison and the third cohort;

apply the second rule set to the first cohort, the second cohort, and the third cohort, wherein a second result is achieved, and wherein the second result comprises a third event;

compare the third event to the first synthetic event, wherein a second comparison is formed, the second comparison comprising further data that can be used to describe a second difference between the third event and the first synthetic event; and

store the second comparison.