Resilient classifier for rule-based system
View Patent ↗A resilient classifier for using with a rule-based system is provided. A system for classifying data for a rule-based system, may include: a system(s) for generating two training data sets, one data set is generated from input data while the second data set is generated from disturbed data; a system for merging the two training data sets; and a system for training a data classifier with the merged training data sets. As a result, the classification of data becomes more accurate, including when disturbed data is encountered.
1. A method for classifying data for a rule-based system, comprising:
generating a first training data set from input data;
generating a second training data set from disturbed data;
merging the first training data set and the second training data set; and
training a data classifier with the merged first training data set and second training data set.
2. The method of claim 1 , further comprising:
receiving a data set; and
using the trained data classifier to classify the received data set.
3. The method of claim 1 , wherein the data comprises text data.
4. The method of claim 1 , wherein generating the second data set comprises adding noise or a disturbance to the first data set.
5. The method of claim 4 , the adding comprising at least one of:
replacing a term in the first data set with a synonym;
replacing a term in the first data set with a misspelled term; and
changing a voice of a phrase in the first data set.
6. The method of claim 1 , wherein the data classifier is an electronic mail classifier.
7. The method of claim 1 , wherein the generating the first training data set from input data is done with a rule-based system.
8. A system for classifying data for a rule-based system, comprising:
a system for generating a first training data set from input data;
a system for generating a second training data set from disturbed data;
a system for merging the first training data set and the second training data set; and
a system for training a data classifier with the merged first training data set and second training data set.
9. The system of claim 8 , further comprising:
a system for receiving a data set; and
a system for using the trained data classifier to classify the received data set.
10. The system of claim 8 wherein the data comprises text data.
11. The system of claim 8 , wherein the system for generating the second data set comprises a system for adding noise or a disturbance to the first data set.
12. The system of claim 11 , the system for adding comprising at least one of:
a system for replacing a term in the first data set with a synonym;
a system for replacing a term in the first data set with a misspelled term; and
a system for changing a voice of a phrase in the first data set.
13. The system of claim 8 , wherein the data classifier is an electronic mail classifier.
14. The system of claim 8 , wherein the system for generating the first training data set from input data comprises a rule-based system.
15. A computer program comprising program code stored on a computer-readable medium, which when executed, enables a computer system to implement a method for classifying data for a rule-based system, the method comprising:
generating a first training data set from input data;
generating a second training data set from disturbed data;
merging the first training data set and the second training data set; and
training a data classifier with the merged first training data set and second training data set.
16. The computer program of claim 15 , the method further comprising:
receiving a data set; and
using the trained data classifier to classify the received data set.
17. The computer program of claim 15 , wherein the data comprises text data.
18. The computer program of claim 15 , wherein the generating the second data set comprises adding noise or a disturbance to the first data set.
19. The computer program of claim 18 , the adding comprising at least one of:
replacing a term in the first data set with a synonym;
replacing a term in the first data set with a misspelled term; and
changing a voice of a phrase in the first data set.
20. The computer program of claim 15 , wherein the data classifier is an electronic mail classifier.
21. The computer program of claim 15 , wherein the generating the first training data set from input data is done with a rule-based system.
22. A data processing system for classifying data for a rule-based system, comprising:
a processing unit;
a bus coupled to the processing unit; and
a memory medium coupled to the bus comprising program code, which when executed by the processing unit causes the data processing system to:
generate a first training data set from input data;
generate a second training data set from disturbed data;
merge the first training data set and the second training data set; and
train a data classifier with the merged first training data set and second training data set.
23. A method of generating a system for classifying data for a rule-based system, the method comprising:
providing a computer system operable to:
generate a first training data set from input data;
generate a second training data set from disturbed data;
merge the first training data set and the second training data set; and
train a data classifier with the merged first training data set and second training data set.