Rating interestingness of profiling data subsets

1. A computer system for rating interestingness of profiling data subsets comprising:

a memory to store computer instructions; and

a processor coupled to said memory to execute said computer instructions to:

receive a plurality of nodes associated with at least one executed computer program;

receive a first set of scores for said plurality of nodes, wherein a score of said first set of scores indicates a resource consumption of a corresponding node;

calculating a score of a second set of scores for said corresponding node by summing a plurality of scores from said first set of scores, wherein said plurality of scores corresponds to said corresponding node or to at least one node related to said corresponding node, wherein the score of said second set of scores indicates a combined resource consumption of said corresponding node, and wherein the combined resource consumption comprises the resource consumption of the corresponding node and resource consumption of the at least one node related to said corresponding node;

removing a node from said plurality of nodes when a proportion between a score of said second set of scores corresponding to said node and a score of said second set of scores corresponding to a related node is lower than a second predefined value;

calculate an interestingness rate for a group of nodes from said plurality of nodes based on one or more of said first set of scores and said second set of scores, wherein said interestingness rate is associated with a cardinality of said group of nodes, the combined resource consumption of a node from said group of nodes, and a corrected total resource consumption of said node from said group of nodes; and

compare said calculated interestingness rate with at least one threshold rate to verify interestingness for said group of nodes.

2. The system of claim 1 , wherein receiving the plurality of nodes comprises: retrieving at least one relationship between at least two nodes wherein said at least one relationship corresponds to execution flow of said at least one computer program.

3. The system of claim 1 , wherein comparing the calculated interestingness rate comprises:

calculating a new interestingness rate for a new group of nodes of said plurality of nodes based on corresponding scores from one or more of said first set of scores and said second set of scores; and

comparing the interestingness rate calculated for the initial group of nodes with the new interestingness rate calculated for said new group of nodes.

4. An article of manufacture including a non-transitory computer readable storage medium storing instructions, which when executed by a computer, cause the computer to:

receive profiling data for at least one executed computer program, wherein said profiling data includes a plurality of nodes associated with said at least one computer program and a first set of scores, wherein a score of said first set of scores indicates a resource consumption of a corresponding node;

calculate a score of a second set of scores for said corresponding node by summing a plurality of scores from said first set of scores, wherein said plurality of scores corresponds to said corresponding node or at least one node related to said corresponding node, wherein the score of said second set of scores indicates a combined resource consumption of said corresponding node, wherein the combined resource consumption comprises the resource consumption of the corresponding node and resource consumption of the at least one node related to said corresponding node;

remove a node from said plurality of nodes when a proportion between a score of said second set of scores corresponding to said node and a score of said second set of scores corresponding to a related node is lower than a second predefined value;

calculate an interestingness rate for a group of nodes from said plurality of nodes based on one or more of said first set of scores and said second set of scores, wherein said interestingness rate is associated with a cardinality of said group of nodes, the combined resource consumption of a node from said group of nodes, and a corrected total resource consumption of said node from said group of nodes; andreport said group of nodes when said interestingness rate is higher than a predefined value.

5. The article of manufacture of claim 4 , wherein receiving the profiling data comprises:

retrieving at least one relationship between at least two nodes, wherein said at least one relationship corresponds to an execution thread of said at least one computer program.

6. The article of manufacture of claim 4 , wherein calculating the interestingness rate for the group of nodes comprises: removing a node from said plurality of nodes when a score of said first set of scores or said second set of scores corresponding to said node is lower than a predefined fraction of overall resource consumption of said plurality of nodes.

7. The article of manufacture of claim 4 , wherein calculating the interestingness rate for the group of nodes comprises: computing a mathematical expression combining a group of scores from said second set of scores corresponding to said group of nodes to evaluate resource consumption distribution within said group of nodes.

8. The article of manufacture of claim 4 , wherein calculating the interestingness rate for the group of nodes comprises:

decreasing a score from said second set of scores corresponding to a first node of by a score of said second set of scores corresponding to a second node, wherein said first node and said second node are related; and

based on the decreased score evaluating overlapping between the nodes within said group of nodes.

9. The article of manufacture of claim 4 , wherein reporting said group of nodes comprises:

calculate a new interestingness rate for a new group of nodes from said plurality of nodes based on a score from one or more of said first set of scores and said second set of scores; and

assigning said new interestingness rate to said predefined value.

10. A computerized method for rating interestingness of profiling data subsets, comprising:

receiving at a computer memory, a plurality of nodes associated with at least one computer program executed in a computer system;

receiving at said computer memory, a first set of scores, wherein a score of said first set of scores indicates a resource consumption of a corresponding node;

calculating at said computer memory a score of a second set of scores for said corresponding node by summing a plurality of scores from said first set of scores, wherein said plurality of scores corresponds to said corresponding node or at least one node related to said corresponding node, wherein the score of said second set of scores indicates a combined resource consumption of said corresponding node, wherein the combined resource consumption comprises the resource consumption of the corresponding node and resource consumption of the at least one node related to said corresponding node;

removing a node from said plurality of nodes when a proportion between a score of said second set of scores corresponding to said node and a score of said second set of scores corresponding to a related node is lower than a second predefined value;

calculating by a processor, an interestingness rate for a group of nodes from said plurality of nodes based on one or more of said first set of scores and said second set of scores, wherein said interestingness rate is associated with a cardinality of said group of nodes, the combined resource consumption of a node from said group of nodes, and a corrected total resource consumption of said node from said group of nodes; and

rendering on a user interface said group of nodes based on said interestingness rate.

11. The method of claim 10 , wherein receiving the plurality of nodes comprises: retrieving a hierarchical relationship between at least two nodes of said plurality of nodes, wherein an execution of said at least one computer program moves from a node of said at least two nodes at a higher hierarchical level to a node of said at least two nodes at a lower hierarchical level.

12. The method of claim 11 , wherein at least one node of said plurality of nodes corresponds to a first method of said at least one computer program, and wherein said hierarchical relationship corresponds to a second method.

13. The method of claim 10 , wherein calculating the interestingness rate for the group of nodes comprises: removing a node from said plurality of nodes when a score of said second set of scores corresponding to said node is lower than a value based on overall resource consumption of said plurality of nodes.

14. The method of claim 10 , wherein calculating the interestingness rate for the group of nodes comprises: removing a node from said plurality of nodes when said node is dominated by a related node.

15. The method of claim 10 , wherein calculating the interestingness rate for the group of nodes comprises: computing a plurality of scores from said second set of scores corresponding to said group of nodes to evaluate resource consumption distribution within said group of nodes.

16. The method of claim 10 , wherein calculating the interestingness rate for the group of nodes comprises: decreasing a score from said second set of scores corresponding to a first node by a score of said second set of score corresponding to a second node to evaluate overlapping between the nodes of the group, wherein said first node and said second node are related.

17. The method of claim 10 , wherein calculating the interestingness rate for the group of nodes comprises: decreasing a score from said second set of scores corresponding to a first node by a score of said second set of scores corresponding to a second node to evaluate said group of nodes for an overall resource consumption of said at least one computer program, wherein said first node and said second node are related.