Business processes based on a predictive model
First Claim
1. A method, comprising operating a computer processor to perform operations comprising:
- accessing process execution data derived from audit logs generated during execution of instances of an automated business process comprising business process entities and defined by a directed graph of interconnected nodes representing respective activities, wherein the business process entities comprise services and resources, and each of the activities is defined by a respective service and is performed by a respective set of one or more resources;
classifying the instances of the business process in accordance with a quality taxonomy defined by a set of quality categories each of which is associated with a respective condition on the process execution data and a respective quality score value, wherein the classifying comprises assigning the business process instances to respective ones of the quality categories based on application of the respective conditions to the accessed process execution data;
building a predictive model comprising a respective set of predictive rules for each of multiple logical partitions of the directed graph defining the business process based upon the assignment of the business process instances to respective ones of the categories, wherein each of the predictive rules assigns a respective probability to each of the quality categories of the quality taxonomy for each of the business process entities that are invocable at the respective logical partitions of the directed graph, and each of the probabilities represents a likelihood that a given business process instance that invokes the respective business process entity at the respective logical partition of the directed graph will have an outcome corresponding to the respective quality category; and
executing an active instance of the business process based on the predictive model.
5 Assignments
0 Petitions
Accused Products
Abstract
Systems and methods of improving business processes are described. These systems and methods enable service providers to improve the quality of services delivered to customers and employees by improving service execution through the optimal selection of resources (e.g., internal resources or external resources, or both) that are invoked to execute the delivered services. In one aspect, process execution data is accessed. Business process instances are classified in accordance with a quality taxonomy. A predictive model including a set of rules for scoring business process outcomes at different stages of the business process is built based upon the classified business process instances. In another aspect, process entities to be invoked at stages of an active business process instance are selected based upon the predictive model. In another aspect, a user is prompted to define a quality taxonomy for classifying outcomes of instances of a business process.
-
Citations
26 Claims
-
1. A method, comprising operating a computer processor to perform operations comprising:
-
accessing process execution data derived from audit logs generated during execution of instances of an automated business process comprising business process entities and defined by a directed graph of interconnected nodes representing respective activities, wherein the business process entities comprise services and resources, and each of the activities is defined by a respective service and is performed by a respective set of one or more resources; classifying the instances of the business process in accordance with a quality taxonomy defined by a set of quality categories each of which is associated with a respective condition on the process execution data and a respective quality score value, wherein the classifying comprises assigning the business process instances to respective ones of the quality categories based on application of the respective conditions to the accessed process execution data; building a predictive model comprising a respective set of predictive rules for each of multiple logical partitions of the directed graph defining the business process based upon the assignment of the business process instances to respective ones of the categories, wherein each of the predictive rules assigns a respective probability to each of the quality categories of the quality taxonomy for each of the business process entities that are invocable at the respective logical partitions of the directed graph, and each of the probabilities represents a likelihood that a given business process instance that invokes the respective business process entity at the respective logical partition of the directed graph will have an outcome corresponding to the respective quality category; and executing an active instance of the business process based on the predictive model. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16)
-
-
17. A computer-readable medium having computer-readable program code embodied therein, the computer-readable program code adapted to be executed by a computer to implement a method comprising:
-
accessing process execution data derived from audit logs generated during execution of instances of an automated business process comprising business process entities and defined by a directed graph of interconnected nodes representing respective activities, wherein the business process entities comprise services and resources, and each of the activities is defined by a respective service and is performed by a respective set of one or more resources; classifying the instances of the business process in accordance with a quality taxonomy defined by a set of quality categories each of which is associated with a respective condition on the process execution data and a respective quality score value, wherein the classifying comprises assigning the business process instances to respective ones of the quality categories based on application of the respective conditions to the accessed process execution data; building a predictive model comprising a respective set of predictive rules for each of multiple logical partitions of the directed graph defining the business process based upon the assignment of the business process instances to respective ones of the categories, wherein each of the predictive rules assigns a respective probability to each of the quality categories of the quality taxonomy for each of the business process entities that are invocable at the respective logical partitions of the directed graph, and each of the probabilities represents a likelihood that a given business process instance that invokes the respective business process entity at the respective logical partition of the directed graph will have an outcome corresponding to the respective quality category; and executing an active instance of the business process based on the predictive model. - View Dependent Claims (18)
-
-
19. Apparatus, comprising:
-
a memory storing program code; and a computer processor coupled to the memory, operable to execute the program code, and based at least in part on the execution of the program code operable to perform operations comprising accessing process execution data derived from audit logs generated during execution of instances of an automated business process comprising business process entities and defined by a directed graph of interconnected nodes representing respective activities, wherein the business process entities comprise services and resources, and each of the activities is defined by a respective service and is performed by a respective set of one or more resources; classifying the instances of the business process in accordance with a quality taxonomy defined by a set of quality categories each of which is associated with a respective condition on the process execution data and a respective quality score value, wherein the classifying comprises assigning the business process instances to respective ones of the quality categories based on application of the respective conditions to the accessed process execution data; building a predictive model comprising a respective set of predictive rules for each of multiple logical partitions of the directed graph defining the business process based upon the assignment of the business process instances to respective ones of the categories, wherein each of the predictive rules assigns a respective probability to each of the quality categories of the quality taxonomy for each of the business process entities that are invocable at the respective logical partitions of the directed graph, and each of the probabilities represents a likelihood that a given business process instance that invokes the respective business process entity at the respective logical partition of the directed graph will have an outcome corresponding to the respective quality category; and executing an active instance of the business process based on the predictive model.
-
-
20. A method, comprising operating a computer processor to perform operations comprising:
-
receiving context data associated with an active instance of an automated business process defined by a directed graph of interconnected nodes representing respective activities each defined by a respective service and performed by a respective set of one or more resources; ascertaining a current logical stage of the active instance of the business process based on the received context data; determining a set of ratings for ones of the resources that are invocable at the current logical stage based on a predictive model comprising a respective set of predictive rules for each of multiple logical partitions of the directed graph defining the business process, each of the predictive rules assigns a respective probability to each of multiple quality categories of a quality taxonomy for each of the resources that are invocable at the respective logical partitions of the directed graph, each of the quality categories is associated with a respective condition on process execution data derived from audit logs generated during execution of instances of the business process, each of the quality categories is associated with a respective quality score value, and each of the probabilities represents a likelihood that a given business process instance that invokes the respective resource at the respective logical partition of the directed graph will have an outcome corresponding to the respective quality category; and selecting at least one of the resources to invoke in the active business process instance based upon the determined set of ratings. - View Dependent Claims (21, 22, 23, 24)
-
-
25. A computer-readable medium having computer-readable program code embodied therein, the computer-readable program code adapted to be executed by a computer to implement a method comprising:
-
receiving context data associated with an active instance of an automated business process defined by a directed graph of interconnected nodes representing respective activities each defined by a respective service and performed by a respective set of one or more resources; ascertaining a current logical stage of the active instance of the business process based on the received context data; determining a set of ratings for ones of the resources that are invocable at the current logical stage based on a predictive model comprising a respective set of predictive rules for each of multiple logical partitions of the directed graph defining the business process, each of the predictive rules assigns a respective probability to each of multiple quality categories of a quality taxonomy for each of the resources that are invocable at the respective logical partitions of the directed graph, each of the quality categories is associated with a respective condition on process execution data derived from audit logs generated during execution of instances of the business process, each of the quality categories is associated with a respective quality score value, and each of the probabilities represents a likelihood that a given business process instance that invokes the respective resource at the respective logical partition of the directed graph will have an outcome corresponding to the respective quality category; and selecting at least one of the resources to invoke in the active business process instance based upon the determined set of ratings.
-
-
26. Apparatus, comprising:
-
a memory storing program code; and a computer processor coupled to the memory, operable to execute the program code, and based at least in part on the execution of the program code operable to perform operations comprising receiving context data associated with an active instance of an automated business process defined by a directed graph of interconnected nodes representing respective activities each defined by a respective service and performed by a respective set of one or more resources; ascertaining a current logical stage of the active instance of the business process based on the received context data; determining a set of ratings for ones of the resources that are invocable at the current logical stage based on a predictive model comprising a respective set of predictive rules for each of multiple logical partitions of the directed graph defining the business process, each of the predictive rules assigns a respective probability to each of multiple quality categories of a quality taxonomy for each of the resources that are invocable at the respective logical partitions of the directed graph, each of the quality categories is associated with a respective condition on process execution data derived from audit logs generated during execution of instances of the business process, each of the quality categories is associated with a respective quality score value, and each of the probabilities represents a likelihood that a given business process instance that invokes the respective resource at the respective logical partition of the directed graph will have an outcome corresponding to the respective quality category; and selecting at least one of the resources to invoke in the active business process instance based upon the determined set of ratings.
-
Specification