Methods and systems for mining association rules

1. A computer implemented method executed by one or more computing devices for mining association rules, said method comprising:

sampling, by at least one of the one or more computing devices, a plurality of transactions from a transaction dataset according to a predetermined sampling method;

mining, by at least one of the one or more computing devices, the sampled plurality of transactions based at least in part on a minimum support threshold and a minimum confidence threshold, the mining identifying one or more association rules when the minimum support threshold and the minimum confidence threshold are satisfied;

determining, by at least one of the one or more computing devices, whether the mining identified at least one association rule; and

adjusting, by at least one of the one or more computing devices, the minimum support threshold and the minimum confidence threshold based at least in part on a determination that the mining step did not identify at least one association rule, wherein the adjusting increases the probability that further mining will identify at least one association rule by:

decrementing at least one of the minimum support threshold and the minimum confidence threshold by a first percentage based at least in part on a determination that at least one of the minimum support threshold and the minimum confidence threshold is greater than a first value; and

decrementing at least one of the minimum support threshold and the minimum confidence threshold by a second percentage based at least in part on a determination that at least one of the minimum support threshold and the minimum confidence threshold is less than or equal to the first value and greater than a second value.

2. The method of claim 1 , further comprising receiving, by at least one of the one or more computing devices, initial values of the minimum support threshold and the minimum confidence threshold.

3. The method of claim 2 , further comprising:

scanning, by at least one of the one or more computing devices, the transaction dataset to identify 1-itemsets;

computing, by at least one of the one or more computing devices, a statistic measure between the 1-itemsets; and

pruning, by at least one of the one or more computing devices, the 1-itemsets from future calculations based at least in part on a determination that the statistic measure is not greater than at least one of the minimum support threshold and the minimum confidence threshold.

4. The method of claim 1 , wherein initial values of the minimum support threshold and the minimum confidence threshold are heuristically generated.

5. The method of claim 1 , further comprising generating initial values of the minimum support threshold and the minimum confidence threshold, the generating comprising:

scanning, by at least one of the one or more computing devices, the transaction dataset for all 1-itemset transactions;

determining, by at least one of the one or more computing devices, a support count for at least one 1-itemset transaction; and

initializing, by at least one of the one or more computing devices, the minimum support threshold and the minimum confidence threshold based on the support count.

6. The method of claim 5 , wherein the minimum support threshold and the minimum confidence threshold is half of the sum of a minimum support count for a 1-itemset and a maximum support count for a 1-itemset.

7. The method of claim 1 , wherein the first value is five, the second value is zero, the first percentage is five percent, and the second percentage is one percent.

8. The method of claim 1 , further comprising preprocessing, by at least one of the one or more computing devices, the transaction dataset containing the plurality of transactions.

9. The method of claim 1 , wherein the sampling method is based at least in part on a selection of a sampling method from a user.

10. The method of claim 1 , wherein the sampling method comprises a stratified sampling method and wherein sampling comprises:

grouping, by at least one of the one or more computing devices, transactions in the transaction dataset into plural subgroups according to an attribute associated with a customer associated with each transaction; and

sampling, by at least one of the one or more computing devices, transactions from the subgroups according to one of a random sampling model and a systematic sampling model.

11. The method of claim 1 , wherein the sampling method comprises a systematic sampling method and wherein sampling comprises:

determining, by at least one of the one or more computing devices, an ordered frame size;

grouping, by at least one of the one or more computing devices, transactions into frames according to the ordered frame size; and

sampling, by at least one of the one or more computing devices, a transaction from each frame.

12. The method of claim 1 , wherein the sampling method comprises a cluster sampling method and wherein sampling comprises:

grouping, by at least one of the one or more computing devices, transactions in the transaction dataset into plural heterogeneous clusters;

selecting, by at least one of the one or more computing devices, one of the plural heterogeneous clusters; and

sampling, by at least one of the one or more computing devices, transactions from the selected heterogeneous cluster.

13. The method of claim 1 , wherein the sampling method comprises a classified data sampling method and wherein sampling comprises:

receiving, by at least one of the one or more computing devices, attributes of the plurality of transactions in the transaction dataset;

grouping, by at least one of the one or more computing devices, transactions in the transaction dataset into plural sets of classified transactions; and

sampling, by at least one of the one or more computing devices, transactions from at least one set of classified transactions in the plural sets of classified transactions.

14. A system for mining association rules, the system comprising:

one or more processors; and

one or more memories operatively coupled to at least one of the one or more processors and having instructions stored thereon that, when executed by at least one of the one or more processors, cause at least one of the one or more processors to:

sample a plurality of transactions from a transaction dataset according to a predetermined sampling method;

mine the sampled plurality of transactions based at least in part on a minimum support threshold and a minimum confidence threshold, the mining identifying one or more association rules when the minimum support threshold and the minimum confidence threshold are satisfied;

determine whether the mining identified at least one association rule; and

adjust the minimum support threshold and the minimum confidence threshold based at least in part on a determination that the mining step did not identify at least one association rule, wherein the adjusting increases the probability that further mining will identify at least one association rule by:

decrementing at least one of the minimum support threshold and the minimum confidence threshold by a first percentage based at least in part on a determination that at least one of the minimum support threshold and the minimum confidence threshold is greater than a first value; and

decrementing at least one of the minimum support threshold and the minimum confidence threshold by a second percentage based at least in part on a determination that at least one of the minimum support threshold and the minimum confidence threshold is less than or equal to the first value and greater than a second value.

15. The system of claim 14 , wherein at least one of the one or more memories has further instructions stored thereon that, when executed by at least one of the one or more processors, cause at least one of the one or more processors to:

scan the transaction dataset to identify 1-itemsets;

compute a statistic measure between the 1-itemsets; and

prune the 1-itemsets from future calculations based at least in part on a determination that the statistic measure is not greater than at least one of the minimum support threshold and the minimum confidence threshold.

16. The system of claim 14 , wherein initial values of the minimum support threshold and the minimum confidence threshold are heuristically generated.

17. At least one non-transitory computer-readable medium storing computer-readable instructions that, when executed by one or more computing devices, cause at least one of the one or more computing devices to:

sample a plurality of transactions from a transaction dataset according to a predetermined sampling method;

mine the sampled plurality of transactions based at least in part on a minimum support threshold and a minimum confidence threshold, the mining identifying one or more association rules when the minimum support threshold and the minimum confidence threshold are satisfied;

determine whether the mining identified at least one association rule; and

adjust the minimum support threshold and the minimum confidence threshold based at least in part on a determination that the mining step did not identify at least one association rule, wherein the adjusting increases the probability that further mining will identify at least one association rule by:

decrementing at least one of the minimum support threshold and the minimum confidence threshold by a first percentage based at least in part on a determination that at least one of the minimum support threshold and the minimum confidence threshold is greater than a first value; and

decrementing at least one of the minimum support threshold and the minimum confidence threshold by a second percentage based at least in part on a determination that at least one of the minimum support threshold and the minimum confidence threshold is less than or equal to the first value and greater than a second value.

18. The at least one non-transitory computer-readable medium of claim 17 , further storing computer-readable instructions that, when executed by at least one of the one or more computing devices, cause at least one of the one or more computing devices to:

scan the transaction dataset to identify 1-itemsets;

compute a statistic measure between the 1-itemsets; and

prune the 1-itemsets from future calculations based at least in part on a determination that the statistic measure is not greater than at least one of the minimum support threshold and the minimum confidence threshold.

19. The at least one non-transitory computer-readable medium of claim 17 , wherein initial values of the minimum support threshold and the minimum confidence threshold are heuristically generated.