Identification of clusters of elements causing network performance degradation or outage

1. A computer-implemented method comprising:

obtaining a tabular data set with columns that respectively characterize items relating to behavior of components of a communication network, wherein records within the tabular data set represent events in the communication network;

constructing a frequent-pattern tree, wherein each respective node is associated with: (i) an item-name for representing an item in the tabular data set, (ii) a count of transactions from a root node of the frequent-pattern tree to the respective node, and (iii) node-links that refer to other nodes in the frequent-pattern tree that represent items having the same item-name;

traversing the frequent-pattern tree in a bottom-up fashion to identify, as part of frequent patterns, a set of nodes with counts greater than a predefined support threshold;

selecting a target item of the items;

generating, from the set of nodes, association-rules that are based on antecedent items associated with the target item; and

reducing the association-rules by combining two or more of the association-rules that have antecedent items that are at least partially disjoint and conditionally dependent.

2. The computer-implemented method for claim 1 , further comprising:

determining, from the association-rules, respective confidence values each representing a likelihood that the antecedent items of a given association-rule are associated with the target item of the given association-rule.

3. The computer-implemented method of claim 2 , further comprising:

organizing numerical values within the tabular data into bins; and

reducing the association-rules by combining groups of the association-rules with adjacent bins and confidence values that are within a predetermined range.

4. The computer-implemented method of claim 2 , further comprising reducing the association-rules by removing the association-rules in which the antecedent items thereof are a superset or subset of the antecedent items of a further association-rule, wherein removing the association-rules comprises determining that the association-rules to remove have confidence values that are within a predetermined range of that of the further association-rule.

5. The computer-implemented method of claim 2 , wherein combining two or more of the association-rules comprises determining that the two or more of the association-rules have confidence values that are within a predetermined range.

6. The computer-implemented method of claim 1 , wherein each of the antecedent items and the target item respectively appear in the columns of the tabular data set, and wherein each of the antecedent items and the target item specify information about an event or state in the communication network.

7. The computer-implemented method of claim 1 , wherein combining two or more of the association-rules that have antecedent items that are at least partially disjoint and conditionally dependent comprises determining whether two or more of the association-rules have antecedent items that are at least partially disjoint and conditionally dependent by:

determining a first value, wherein the first value corresponds to an absolute value of a difference between a first confidence value and a second confidence value, and wherein the first confidence value is associated with a first association-rule and the second confidence value is associated with a second association-rule;

determining a second value, wherein the second value corresponds to a quotient between: (i) an absolute value of a difference between a first cover value and a second cover value and (ii) the second cover value, wherein the first cover value is a cover value of a union between the antecedent item of the first association-rule and the antecedent item of the second association-rule, and wherein the second cover value is a cover value of the antecedent item of the first association-rule;

determining a third value, wherein the third value corresponds to a quotient between: (i) an absolute value of a difference between the first cover value and a third cover value and (ii) the second cover value, and wherein the third cover value is a cover value of the antecedent item of the second association-rule;

comparing the first value to a confidence threshold;

comparing the second value to a support threshold; and

comparing the third value to the support threshold.

8. The computer-implemented method of claim 1 , wherein generating the association-rules comprises storing, in a memory, the association-rules.

9. The computer-implemented method of claim 8 , wherein reducing the association-rules comprises deleting, from the memory, any of the association-rules that are redundant due to the combining.

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

providing for display the association-rules as reduced, wherein the antecedent items thereof are identified as possible causes of the events or states within the communication network involving the target item.

11. The computer-implemented method of claim 10 , wherein providing for display the association-rules as reduced comprises transmitting a representation of the association-rules as reduced to a client device.

12. An article of manufacture including a non-transitory, computer-readable medium, having stored thereon program instructions that, upon execution by a computing system, cause the computing system to perform operations comprising:

obtaining a tabular data set with columns that respectively characterize items relating to behavior of components of a communication network, wherein records within the tabular data set represent events in the communication network;

constructing a frequent-pattern tree, wherein each respective node is associated with: (i) an item-name for representing an item in the tabular data set, (ii) a count of transactions from a root node of the frequent-pattern tree to the respective node, and (iii) node-links that refer to other nodes in the frequent-pattern tree that represent items having the same item-name;

traversing the frequent-pattern tree in a bottom-up fashion to identify, as part of frequent patterns, a set of nodes with counts greater than a predefined support threshold;

selecting a target item of the items;

generating, from the set of nodes, association-rules that are based on antecedent items associated with the target item; and

reducing the association-rules by combining two or more of the association-rules that have antecedent items that are at least partially disjoint and conditionally dependent.

13. The article of manufacture of claim 12 , wherein the operations performed by the computing system when executing the program instructions further comprise:

determining, from the association-rules, respective confidence values each representing a likelihood that the antecedent items of a given association-rule are is associated with the target item of the given association-rule.

14. The article of manufacture of claim 13 , wherein the operations performed by the computing system when executing the program instructions further comprise:

organizing numerical values within the tabular data into bins; and

reducing the association-rules by combining groups of the association-rules with adjacent bins and confidence values that are within a predetermined range.

15. The article of manufacture of claim 13 , wherein the operations performed by the computing system when executing the program instructions further comprise reducing the association-rules by removing the association-rules in which the antecedent items thereof are a superset or subset of the antecedent items of a further association-rule, and wherein removing the association-rules comprises determining that the association-rules to remove have confidence values that are within a predetermined range of that of the further association-rule.

16. The article of manufacture of claim 13 , wherein combining two or more of the association-rules comprises determining that the two or more of the association-rules have confidence values that are within a predetermined range.

17. The article of manufacture of claim 12 , wherein each of the antecedent items and the target item respectively appear in the columns of the tabular data set, wherein each of the antecedent items and the target item specify information about an event or state in the communication network.

18. The article of manufacture of claim 12 , wherein combining two or more of the association-rules that have antecedent items that are at least partially disjoint and conditionally dependent comprises determining whether two or more of the association-rules have antecedent items that are at least partially disjoint and conditionally dependent by:

determining a first value, wherein the first value corresponds to an absolute value of a difference between a first confidence value and a second confidence value, and wherein the first confidence value is associated with a first association-rule and the second confidence value is associated with a second association-rule;

determining a second value, wherein the second value corresponds to a quotient between: (i) an absolute value of a difference between a first cover value and a second cover value and (ii) the second cover value, wherein the first cover value is a cover value of a union between the antecedent item of the first association-rule and the antecedent item of the second association-rule, and wherein the second cover value is a cover value of the antecedent item of the first association-rule;

determining a third value, wherein the third value corresponds to a quotient between: (i) an absolute value of a difference between the first cover value and a third cover value and (ii) the second cover value, and wherein the third cover value is a cover value of the antecedent item of the second association-rule;

comparing the first value to a confidence threshold;

comparing the second value to a support threshold; and

comparing the third value to the support threshold.

19. The article of manufacture of claim 12 , wherein generating the association-rules comprises storing, in a memory, the association-rules.

20. A computing system comprising:

one or more processors; and

memory containing program instructions that, when executed by the one or more processors, cause the computing system to perform operations comprising:

obtaining a tabular data set with columns that respectively characterize items relating to behavior of components of a communication network, wherein records within the tabular data set represent events in the communication network;

constructing a frequent-pattern tree, wherein each respective node is associated with: (i) an item-name for representing an item in the tabular data set, (ii) a count of transactions from a root node of the frequent-pattern tree to the respective node, and (iii) node-links that refer to other nodes in the frequent-pattern tree that represent items having the same item-name;

traversing the frequent-pattern tree in a bottom-up fashion to identify, as part of frequent patterns, a set of nodes with counts greater than a predefined support threshold;

selecting a target item of the items;

generating, from the set of nodes, association-rules that are based on antecedent items associated with the target item; and

reducing the association-rules by combining two or more of the association-rules that have antecedent items that are at least partially disjoint and conditionally dependent.