Back to results
US 11,775,901 B2Granted
Real time testing and monitoring for business process management systems
AssigneeInternational Business Machines Corporation
InventorsMichal Paluch, Karol Cisowski, Tomasz Ploskon, Jaroslaw Musial
Life of this patent
Estimating the term…
Abstract
Providing a method for real-time testing and monitoring of a business process management system in order to identify commonly failing parameters. Machine learning methods are used to run multiple rounds of testing of the business process management system to more accurately identify commonly failing parameters and to modify the business process by removing those identified parameters from future rounds of testing.
Claims18
A computer-implemented method (CIM) comprising:
- running, by a first user, a first plurality of runs of a first set of business processes, with each run of the first plurality of runs including collecting a set of internal runtime parameters and external parameters;
- detecting, by a business process failure detection module, a first business process management (BPM) failure, with the detection of the first BPM failure including detecting internal runtime and external parameters involved in the run associated with the first BPM failure, and with the first BPM failure being labeled in a manner that can be used to train a BPM machine-learning module to build high risk test scenarios;
- storing, in a reference data warehouse, the internal runtime and external parameters involved in the run associated with the first BPM failure;
- building, by the BPM machine-learning module, a first high risk test scenario, with the building of the first high risk test scenario including:
- collecting, by the BPM machine-learning module, a set of BPM failure training data,
- receiving, from a reference data warehouse, a set of parameters that the BPM machine-learning module uses for running the first plurality of runs, and
- responsive to the receipt of the set of parameters, identifying a sub-set of parameters as commonly causing the first plurality of runs to fail;
- training the BPM machine-learning module with the first labeled BPM failure parameters to recognize which set of parameters to use;
- running, by the BPM machine-learning module, a second plurality of runs of the first high risk test scenario;
- updating the collected set of BPM training data to include the sub-set of parameters identified as commonly causing the first plurality of runs to fail;
- and
- responsive to the running of the second plurality of runs, modifying, by the BPM machine-learning module, the first high risk test scenario by:
- identifying a sub-set of problematic parameters, and
- removing the sub-set of problematic parameters.
11 more elements
A computer program product (CPP) comprising:
- a machine-readable storage device;
- and
- computer code stored on the machine-readable storage device, with the computer code including instructions and data to cause a processor(s) set to perform operations including the following:
- running, by a first user, a first plurality of runs of a first set of business processes, with each run of the first plurality of runs including collecting a set of internal runtime parameters and external parameters,
- detecting, by a business process failure detection module, a first business process management (BPM) failure, with the detection of the first BPM failure including detecting internal runtime and external parameters involved in the run associated with the first BPM failure, and with the first BPM failure being labeled in a manner that can be used to train a BPM machine-learning module to build high risk test scenarios,
- storing, in a reference data warehouse, the internal runtime and external parameters involved in the run associated with the first BPM failure,
- building, by the BPM machine-learning module, a first high risk test scenario, with the building of the first high risk test scenario including:
- collecting, by the BPM machine-learning module, a set of BPM failure training data,
- receiving, from a reference data warehouse, a set of parameters that the BPM machine-learning module uses for running the first plurality of runs, and
- responsive to the receipt of the set of parameters, identifying a sub-set of parameters as commonly causing the first plurality of runs to fail,
- training the BPM machine-learning module with the first labeled BPM failure parameters to recognize which set of parameters to use,
- running, by the BPM machine-learning module, a second plurality of runs of the first high risk test scenario,
- updating the collected set of BPM training data to include the sub-set of parameters identified as commonly causing the first plurality of runs to fail, and
- responsive to the running of the second plurality of runs, modifying, by the BPM machine-learning module, the first high risk test scenario by:
- identifying a sub-set of problematic parameters, and
- removing the sub-set of problematic parameters.
13 more elements
A computer system (CS) comprising:
- a processor(s) set;
- a machine-readable storage device;
- and
- computer code stored on the machine-readable storage device, with the computer code including instructions and data for causing the processor(s) set to perform operations including the following:
- running, by a first user, a first plurality of runs of a first set of business processes, with each run of the first plurality of runs including collecting a set of internal runtime parameters and external parameters,
- detecting, by a business process failure detection module, a first business process management (BPM) failure, with the detection of the first BPM failure including detecting internal runtime and external parameters involved in the run associated with the first BPM failure, and with the first BPM failure being labeled in a manner that can be used to train a BPM machine-learning module to build high risk test scenarios,
- storing, in a reference data warehouse, the internal runtime and external parameters involved in the run associated with the first BPM failure,
- building, by the BPM machine-learning module, a first high risk test scenario, with the building of the first high risk test scenario including:
- collecting, by the BPM machine-learning module, a set of BPM failure training data,
- receiving, from a reference data warehouse, a set of parameters that the BPM machine-learning module uses for running the first plurality of runs, and
- responsive to the receipt of the set of parameters, identifying a sub-set of parameters as commonly causing the first plurality of runs to fail,
- training the BPM machine-learning module with the first labeled BPM failure parameters to recognize which set of parameters to use,
- running, by the BPM machine-learning module, a second plurality of runs of the first high risk test scenario,
- updating the collected set of BPM training data to include the sub-set of parameters identified as commonly causing the first plurality of runs to fail, and
- responsive to the running of the second plurality of runs, modifying, by the BPM machine-learning module, the first high risk test scenario by:
- identifying a sub-set of problematic parameters, and
- removing the sub-set of problematic parameters.
14 more elements
Dependencies are inferred from the claim text.