Business processes based on a predictive model

• US 7,565,304 B2
• Filed: 06/21/2002
• Issued: 07/21/2009
• Est. Priority Date: 06/21/2002
• Status: Active Grant

• Current Assignee: Hewlett Packard Enterprise Development LP (Hewlett-Packard Enterprise Company)
• Original Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Inventors: Dayal, Umeshwar, Casati, Fabio, Shan, Ming-Chien
• Primary Examiner(s): Boyce; Andre
• Assistant Examiner(s): PATS, JUSTIN
• Application Number:
  US10/176,505
• Publication Number:
  US 20030236691A1
• Time in Patent Office: 2,587 Days
• Field of Search: 705/7
• US Class Current: 705/7.27
• CPC Class Codes: G06Q 10/06312 Adjustment or analysis of established resource schedule, e.g. resource or task levelling, or dynamic rescheduling
• CPC Class Codes: G06Q 10/0633 Workflow analysis
• CPC Class Codes: G06Q 10/06375 Prediction of business process outcome or impact based on a proposed change
• CPC Class Codes: G06Q 10/06395 Quality analysis or management
• CPC Class Codes: G06Q 10/067 Enterprise or organisation modelling
• CPC Class Codes: G06Q 30/06 Buying, selling or leasing transactions
• CPC Class Codes: Y10S 707/99931 Database or file accessing

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.

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.

2. The method of claim 1, wherein the building comprises partitioning the directed graph into the logical partitions.

3. The method of claim 2, wherein the building comprises, at each of the logical partitions, generating the predictive rules based upon context data and a set of one or more of the business process entities that are invocable at the logical partition.

4. The method of claim 3, wherein the invocable business process entities are selected from services or resources, or both.

5. The method of claim 1, wherein the executing comprises selecting ones of the business process entities to be invoked at respective ones of the partitions of the active instance of the business process based upon the predictive model.

6. The method of claim 5, wherein the business process entities are selected for invocation in accordance with one or more specified quality goals.

7. The method of claim 6, wherein at least one of the one or more quality goals corresponds to one of:

maximizing average quality per business process instance; and

maintaining business process instance quality above a threshold value.

8. The method of claim 5, wherein the selecting comprises applying the predictive model to context data at a current one of the logical partitions of the active instance of the business process to obtain predictions of business process outcomes for each business process entity invocable at the current business process partition.

9. The method of claim 5, wherein the invocable business process entities are selected from services or resources, or both.

10. The method of claim 1, further comprising:

on a display, presenting a user interface prompting a user to define the quality taxonomy classifying outcomes of instances of the automated business process,

wherein the presenting comprises presenting the user interface to prompt the user to define the quality taxonomy in terms of a set of quality categories each of which is associated with (i) a respective condition on process execution data derived from audit logs generated during execution of instances of the business process, and (ii) a respective quantitative quality score value.

11. The method of claim 10, further comprising operating the computer processor to perform operations comprising prompting a user to define one or more quality goals in accordance with which business process entities are to be selected for invocation at stages of an active instance of the business process.

12. The method of claim 11, wherein at least one of the quality goals corresponds to one of:

maximizing average quality per business process instance; and

maintaining business process instance quality above a threshold value.

13. The method of claim 1, wherein the accessing comprising accessing the process execution data from a business process data warehouse.

14. The method of claim 1, wherein the executing comprises ascertaining the logical stage of an active instance of the business process from context data associated with the active instance, and determining a set of ratings for ones of the resources that are invocable at the ascertained logical stage based on the predictive model.

15. The method of claim 1, wherein the executing comprises, applying the predictive model to the active instance of the business process, ranking ones of the resources invocable at a given one of the stages of the active instance based on the respective probabilities assigned to the quality categories and the respective quality score values associated with the quality categories, and outputting the ranking.

16. The method of claim 1, wherein the executing comprises, selecting the resource to invoke at the given stage of the active instance of the business process based on application of a specified quality goal to the rankings.

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.

18. The computer-readable medium of claim 17, wherein the computer readable program code is adapted to be executed by the computer

to perform operations comprising:

on a display, presenting a user interface prompting a user to define the quality taxonomy classifying outcomes of instances of the automated business process;

wherein in the presenting the computer process performs operations comprising presenting the user interface to prompt the user to define the quality taxonomy in terms of a set of quality categories each of which is associated with (i) a respective condition on process execution data derived from audit logs generated during execution of instances of the business process, and (ii) a respective quantitative quality score value.

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.

21. The method of claim 20, wherein process entities are selected for invocation in accordance with one or more specified quality goals.

22. The method of claim 21, wherein a quality goal corresponds to one of:

maximizing average quality per business process instance; and

maintaining business process instance quality above a threshold value.

23. The method of claim 20, wherein the determining comprises applying the predictive model to context data to obtain the ratings representing predictions of business process outcomes for each business process entity invocable at the current logical stage of the active instance of the business process.

24. The method of claim 20, wherein the invocable business process entities are selected from services or resources, or both.

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.