CN106980935B - Process instance set evolution management method based on configurable process model - Google Patents
Process instance set evolution management method based on configurable process model Download PDFInfo
- Publication number
- CN106980935B CN106980935B CN201710188272.2A CN201710188272A CN106980935B CN 106980935 B CN106980935 B CN 106980935B CN 201710188272 A CN201710188272 A CN 201710188272A CN 106980935 B CN106980935 B CN 106980935B
- Authority
- CN
- China
- Prior art keywords
- configurable
- node
- model
- process instance
- nodes
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0633—Workflow analysis
Abstract
The invention provides a process instance set evolution management method based on a configurable process model, and belongs to the technical field of business processes and information. The invention takes the similar process instance set describing the same service as input, automatically generates a configurable process model through process instance combination, and utilizes the model to uniformly describe and manage the process instance set. When the process instance set changes, namely a new process instance, a deleted process instance or a modified process instance occurs, the corresponding configurable process model automatically evolves to adapt to the change of the process instance set. The method is intuitive, simple and effective, the generation of the configurable process model has the characteristics of one-time model combination and configuration rule construction support, and the problems of high redundancy, inconsistent evolution and the like caused by similar process examples in a process model library are solved to a certain extent by establishing the uniform configurable process model.
Description
Technical Field
The invention relates to the technical field of business processes and information, in particular to a process instance set evolution management method based on a configurable process model.
Background
With the gradual improvement of the enterprise informatization level, the importance of the process in enterprise management is increasingly prominent, and the production process, the operation process, the management process and the like in the enterprise belong to the enterprise process, which is also called as a business process. The Process model is an abstraction of the Process in the real world, and is mainly used for describing, analyzing and improving the Process, and currently, mainstream Process Modeling languages include UML Activity Diagram (UML Activity Diagram), bpmn (business Process Modeling notification), EPC (Event-driven Process channels), yawl (yet other Workflow language), and the like. With the increasing popularity of process models, it is possible for an enterprise to own and manage a large number of business process models.
After long-term accumulation, a large number of Process Model instances (Process Model variants, Process instances or Process variants for short) describing similar processes in the field may exist in the Process Model library of the enterprise. For example, the Suncorp group is the largest insurance company in australia, and provides various types of insurance products, such as car insurance, house insurance, business insurance, and so on, and these products may also be provided separately by different subsidiaries under the group, so there are at least 30 more process examples describing processing of insurance claim application in the Suncorp group; as another example, a local airport visitors service company provides different but similar services for different types of passengers, such as government visitors, business visitors and other personal members, and has several member subsidiaries around the country, and the visitors service processes for different types of passengers and different subsidiaries are very similar but have certain differences. A set of similar process instances that describe the same business that exist in such an enterprise process model library is referred to as a process instance collection.
Since business of an enterprise may change at any time, the process model also evolves as business changes. Evolution inconsistencies between these process instances may occur due to the presence of similar process instances in the process model library. For example, according to business requirements, the activity a of all process models needs to be changed and modified into the activity B, but business personnel may forget to modify a certain process instance when changing the process model, thereby causing inconsistency between the process instance and other process instances in the set. In addition, the existence of the similar process examples also increases the redundancy of the process model library, and increases the difficulty for the effective management of the process model.
The prior art often solves the above problems by building a unified model that describes a collection of process instances, or implementing automatic propagation of changes between process instances. Configurable process models are reference model modeling techniques that have emerged in recent years, and users may obtain a particular process model by tailoring and configuring the reference model. The configurable process model describes the variability of the process through the configurable nodes, each configurable node can have various configuration options, association relations exist among different configuration options, and modeling personnel can obtain a series of similar process models through different configurations of the configurable nodes, so that the configurable process model can be used as a unified model to describe and manage a process instance set, so that the redundancy of a process model library is reduced, and the consistent evolution of the process instance set is realized.
The use of configurable process models as a unified model for describing a set of process instances essentially requires two things to be done: first, automatically generating a corresponding configurable process model based on a set of process instances; second, when a process instance set evolves, i.e., a new process instance is added to the set or an existing process instance is deleted from the set, the configurable process model corresponding to the set should also evolve automatically to adapt to the new process instance set. The existing process model merging technology can merge a pair of process instances into one configurable process model, but the technology cannot support the construction of the association relationship between the configurable nodes, only supports the merging of a pair of process models, and cannot simultaneously merge a group of process instances into the configurable process model. In addition, there is currently a lack of corresponding technologies to support the automatic incremental evolution of the structural and configuration associations of configurable process models.
Disclosure of Invention
The invention provides a process instance set evolution management method based on a configurable process model, aiming at the problems of high redundancy, inconsistent evolution and the like caused by the existence of similar process instances in a process model library. The method of the invention better solves the problems, describes and manages the process instance set by establishing a uniform configurable process model so as to reduce the redundancy of the process model library and realize the consistent evolution of the process instance set, and when the process instance set changes, the corresponding configurable process model can also adaptively evolve.
The invention discloses a process instance set evolution management method based on a configurable process model, which comprises the steps of firstly, taking a process instance set for describing a specific service as input, automatically constructing the configurable process model which supports constraint rule description, and then automatically evolving and managing the configurable process model to adapt to the change of the process instance set.
The invention provides an automatic construction method of a configurable process model, which comprises the following steps:
the method comprises the following steps: a configurable process model is defined for uniformly managing a set of process instances.
Step two: merging all process instances in the process instance set, generating a structure of a configurable process model and labeling configurable nodes, comprising:
establishing a node matching relationship between process instances by taking all the process instances as input, combining the matched nodes, combining edges with the same source node and target node, inserting the edges into an XOR gateway optimization model structure for structural processing, and enabling all input/output control flows to be unique;
configurable nodes are generated, a node being configurable if and only if the labels of all input-output edges contained by the node are not exactly the same.
Step three: generating configuration association relations among the configurable nodes so as to obtain a complete configurable process model, wherein the configuration association relations include:
firstly, identifying a configuration result of a configurable node in each process instance;
then, a set of configuration rules is generated based on the configuration result matrix using an association rule mining algorithm.
The invention discloses an automatic construction method based on a configurable process model, and relates to a process instance set evolution management method based on a configurable process model.
The case of a process instance set change is: (1) adding a new process instance to the set of process instances; (2) deleting a process instance from the set of process instances; (3) the user modifies a process instance in the set of process instances.
The invention has the advantages and positive effects that: (1) the method is visual, simple and effective, can automatically generate the configurable process model based on the process instance set, the generation mode is one-time process model combination, the unified model is adopted to describe and manage the process instance set, the redundancy of the process model base can be reduced, and the consistent evolution of the process instance set can be realized; (2) the unified model constructed by the method can adaptively evolve along with the change of the process instance set, the evolution covers the structure and the configuration incidence relation of the configurable process model, and is incremental, namely, the evolution method only needs to consider the process instance which is changed currently, and has higher efficiency.
Drawings
FIG. 1 is a schematic overall flow chart of a process instance set evolution management method based on a configurable process model according to the present invention;
FIG. 2 is a simplified schematic diagram of the introduction of the configurable process model concept of the present invention;
FIG. 3 is a schematic diagram of a set of example insurance claim settlement processes in accordance with an embodiment of the method of the present invention;
FIG. 4 is a diagram illustrating pre-processing results of an embodiment of the present invention after merging of implementation process instances in step two;
FIG. 5 is a diagram illustrating the result of simple combination of process instances in step two of the present invention;
FIG. 6 is a diagram illustrating the result of the repeated edge merging in step two according to the present invention;
FIG. 7 is a schematic diagram illustrating the way an XOR gateway is inserted in step two of the present invention;
FIG. 8 is a diagram illustrating the result of inserting the XOR gateway in step two according to the present invention;
FIG. 9 is a diagram illustrating the result of applying the reduction rule 1 in step two according to the present invention;
FIG. 10 is a diagram illustrating the result of applying the reduction rule 2 in step two according to the present invention;
FIG. 11 is a schematic diagram of a configurable process model generated in step two of the present invention;
FIG. 12 is a schematic diagram of a configuration rule definition included in the configurable process model of the present invention;
FIG. 13 is a schematic diagram illustrating the results of configuration rules for a configurable process model generated in step three of the present invention;
FIG. 14 is a diagram of a new process instance from the set of process instances in step five according to an embodiment of the present invention;
FIG. 15 is a diagram illustrating the result of inserting new nodes and edges into the configurable process model in step five of the present invention;
FIG. 16 is a diagram illustrating an evolution result of a configurable process model structure in step five of the present invention when a new process instance is added;
FIG. 17 is a diagram illustrating the evolution of configurable process model configuration rules during the addition of new process instances in step eight of the present invention.
Detailed Description
To facilitate an understanding and practice of the invention by those of ordinary skill in the art, specific embodiments thereof will now be described with reference to the accompanying drawings. The BPMN modeling language is widely used for modeling business processes, and the implementation examples introduced by the invention all adopt modeling icons of the BPMN language. As shown in FIG. 1, to implement the steps of the method of the present invention, the specific steps P01-P08 are described in detail as follows:
step P01: a configurable process model is defined for managing evolution of a set of process instances. A configurable process model is a process model that contains configurable nodes, which are nodes that need to make configuration decisions at design time. The configurable node may be an Activity (Activity) or Gateway node (Gateway). A configurable activity can be configured ON or OFF, which respectively means that this activity is present or ignored in the process instance, as shown in fig. 2 (a). A configurable gateway may be configured to tailor the behavior that the gateway contains, for example, a configurable XOR node may be configured as any of the branch paths it contains or the XOR node and several branch paths may be reserved in the configuration, as shown in fig. 2 (b); whereas for a configurable OR node, since it contains the most extensive behavioral semantics, it can be configured as any gateway type OR select any branching path, as shown in fig. 2 (c). Table 1 shows the configuration rules of different types of configurable mesh nodes, and the configuration result of one configurable node n is denoted as conf (n).
TABLE 1 configuration rules for different types of configurable mesh nodes
| OR | AND | XOR | SEQ | |
| Configurable OR node | √ | √ | √ | √ |
| Configurable AND node | √ | |||
| Configurable XOR node | √ | √ |
In table 1, an OR node represents an OR node, which indicates that any several paths can be selected for execution in subsequent branch paths; the AND node represents an AND node AND represents that all subsequent branch paths need to be executed; the XOR node represents an XOR node, which represents that one and only one path in the subsequent branch paths are selected to be executed; SEQ indicates the selection of one of the paths (Sequence) to connect with the configurable node and the deletion of the configurable node. For example, the last row of table 1 indicates a configuration rule that: the configurable XOR node in the configurable process model may be configured as any branch path (SEQ) to which it is connected in the process instance, or may be left configured as a normal XOR node.
The invention defines a configurable process model as a connected directed graph with nodes and attributes, elements such as activities and gateways in the process model are abstracted as nodes, control flow connections are abstracted as directed edges, and the nodes can have a plurality of attributes such as name, type, configurability, and the like, specifically, one configurable process model is a six-tuple (N, E, L, T, B, S), wherein:
n represents the set of all nodes in the model;
·
representing a set of directed edges between nodes;
n → label represents the name of each node in the model;
n → type represents the type of node, including activity, event, gateway, etc.;
n → { true, false } is a Boolean function for specifying whether the model node is a configurable node;
s is a set of configuration rules CR, one of which describes the association between the configuration results of two nodes, the configuration rules being in the form of "IF rule header THEN rule body", the description type being "when node n1Is configured as conf (n)1) Then node n1Will be configured as conf (n)2) "is used in the following description.
Step P02: all process instances in the process instance set are merged, the structure of the configurable process model is generated, and the configurable nodes are labeled.
The invention is described in detail in connection with a set of example processes for insurance claim processing as shown in FIG. 3. FIG. 3(1) illustrates a conventional process, in which after receiving a Request (Receive Claim), a clerk checks (valid Claim or Confirm Request) a Claim application of a client, and after the check is passed, two managers are assigned for Damage assessment (Review Damage by Adjuster 1 and Review Damage by Adjuster 2), and after the assessment is made, the clerk is handed over to Determine a Claim amount (delete receipt), and finally pays (Make Payment) a money to the client after passing through a Review by Mgr; FIG. 3(2) depicts the process when the loss amount is low, and when the loss amount is below 10 ten thousand dollars, the assessment of Damage can be made by only one rational operator (Review Damage by Adjuster 1); when the loss amount is larger and exceeds 30 ten thousand yuan, a step of Approval (Approval by Senior manager Mgr.) needs to be added, as shown in (3) of FIG. 3; if the application requires urgent processing, the two approval steps can be performed simultaneously to save time, as shown in fig. 3 (4); fig. 3(5) illustrates a case where the loss amount is particularly large, and when the loss amount exceeds 50 ten thousand yuan, the Approval of the second step needs to be performed by an assistant President (Approval by view President).
The goal of this step is to automatically generate a unified configurable process model based on these 5 process model instances. The steps for automatically generating a unified configurable process model are as follows.
Firstly, preprocessing an input process instance, matching the similarity of the activity names among the process instances by adopting a WordNet-based word similarity method (Lin similarity), establishing a node matching relationship among the process instances, and marking the identification number id of the process instance on the edges of all the process instances, as shown in a process instance (2) in fig. 4, wherein each edge is marked by using the id 2 of the process instance. The model of fig. 4 is the same as fig. 3, and letters are used instead of specific activity names for the convenience of the following description.
Second, the nodes that match each other are merged into one node, and the connection of edges between the nodes is maintained. If the attributes of the matched nodes are different, the attribute value with the highest frequency of occurrence in the process instance is reserved, and other attribute values are recorded in the form of node marks. The 5 process instances of FIG. 4 merge nodes to arrive at the model shown in FIG. 5.
Thirdly, combining the edges with the same start and stop ends, traversing all the edges of the model, finding the edges with the same source node and target node, combining the edges into one edge respectively, combining the process instances id marked on the edges, and processing the model shown in fig. 5 to obtain the model shown in fig. 6.
Fourthly, in order to make the model more structured, an XOR gateway optimization model structure needs to be inserted, so that all branching or merging occurs on the gateway node, that is, the active node is not allowed to make implicit branching or merging, and therefore, the following processing is performed on all activities which are not unique to the input/output control flow: 1) if the input edge of the active node n is not unique, e1,e2,…,ek，k>1, inserting an XOR-Join type gateway node g before n, setting target nodes of all input edges of the active node n as g, and inserting an edge e taking g as a source node and n as a target nodenewThe label of the edge is the union of all input edge labels, as shown in the left diagram in fig. 7; 2) if the output edge of the active node n is not unique, e1,e2,…,ek，k>1, inserting an XOR-Split type gateway node g after n, setting the source nodes of all output edges of n as g, and inserting an edge e taking n as the source node and g as the target nodenewThe label of the edge is the union of all output edge labels, as shown in the right diagram of fig. 7. The model shown in fig. 6 is subjected to the above-described processing to obtain the model shown in fig. 8.
Fifthly, the simplification model provides two simplification rules to reduce the number of gateway nodes, including: 1) merging continuous homogeneous gateway nodes; 2) and deleting the useless gateway nodes. Reduction rule 1: merging successive homogeneous gateway nodes, since two successive gateway nodes of the branch (or merge) type can be merged into one branch (or merge) gateway node. For example, the part marked by the dashed box in fig. 8 is a gateway node that can be merged, and the merged model is shown in fig. 9. Simplification rule 2: the useless gateway nodes are deleted, that is, the gateway nodes with the number of input edges and the number of output edges both being 1 are deleted, and the simplified model is shown in fig. 10.
Sixth, configurable nodes are generated, and if a node is configurable and only if the labels of all the input and output edges included in the node are not completely consistent, the configurable process model after labeling the configurable nodes is as shown in fig. 11.
Step P03: and generating configuration incidence relations, namely configuration rules, among the configurable nodes in the configurable process model. The present invention formally describes the configuration rules using the EBNF shown in fig. 12. For example, there is a configuration rule CR between configurable process model nodes g6 and g3 in FIG. 111"g 6 ═ SEQ-g5 → g3 ═ SEQ-E ═ S ═ 0.5, and C ═ 1.0", the meaning of this arrangement rule is: when node g6 is configured as a single path with source node g5, node g3 is likely to be configured as a single path with target node active E, and the support and confidence of the configuration rule are 0.5 and 1.0, respectively.
FIG. 12 illustrates a configuration rule definition included in the configurable process model of the present invention, where the configuration rule definition is expressed as:
CR＝Config,“→”,Config,“:”,Param；
the parameters involved therein are explained as follows:
CR: configuration rules, e.g. CR1；
Config: rule heads or bodies, e.g. CR1"g 6 ═ SEQ-g 5" and "g 3 ═ SEQ-E" in (a);
param: parameters of configuration rules, e.g. CR1"S ═ 0.5, C ═ 1.0" in (a);
NID: configurable node id, e.g. CR1"g 6 ═ g6 in SEQ-g 5", and "g 3 ═ g3 in SEQ-E";
and (4) Result: configuration results, e.g. CR1"SEQ-g 5" in "g 6 ═ SEQ-g 5" and "SEQ-E" in "g 3 ═ SEQ-E";
ActRNAs: a configuration result of the configurable activity;
GtwRes: configuring a configuration result of the configurable gateway;
type: type of configuration result of configurable gateway, e.g. CR1"g 6 ═ SEQ" in SEQ-g5 ";
branches: configurable gateway configuration result selection paths, e.g. CR1"g 6 ═ g5 in SEQ-g 5";
value 1: the support degree of the configuration rule;
value 2: the confidence of the configuration rule.
The Support value Support of 0.5 indicates that in 50% of the process examples the configuration results "g 6 — SEQ-g 5" and "g 3 — SEQ-E" hold, while the Confidence value Confidence of 1.0 indicates that for process examples satisfying "g 6 — SEQ-g 5", they all also satisfy "g 3 — SEQ-E". The support degree represents the proportion of the process examples in which the rule head and the rule body are both established in the process example set, and the confidence degree represents the proportion of the process examples in which the rule head and the rule body are both established in the process example set.
The generation of the configuration rule comprises two core steps: the configuration results of the configurable nodes in the various process instances need to be identified first, and then a configuration rule set is generated based on the configuration results. In order to identify the configuration result of the configurable node n in the configurable process model CG in the process instance G, a node matching relationship between CG and G is first established based on Lin similarity. If node n is a configurable activity, its configuration result conf (n) ═ ON if and only if there is an activity n' in G that has a node matching relationship with it, otherwise its configuration result conf (n) ═ OFF. If the node n is a configurable gateway and a gateway n 'exists in the G and has a node matching relationship with the node n', the configuration result conf (n) ═ Type (n ') -n' ·, where n is a branch Type and n '·denotesa subsequent node set of n'; or conf (n) ═ Type (n ') -. n', where n is the merge Type and n 'denotes the set of predecessor nodes for n'; if there is no matching relationship between the gateway and the node n in G, assuming that n is a branch (or merge) type gateway, it needs to find in its successor (or predecessor) nodeIf the node n meeting the condition is found out in the node matching relation1Then n is configured as result of conf (n) ═ SEQ-n1Otherwise node n is a configurable node that is not valid for process instance G.
The configuration result is identified to obtain an N × M matrix, where N is the number of process instances in the process instance set and M is the number of configurable nodes in the configurable process model, and the configuration result matrix is used as the basis for generating the configuration rule, table 2 shows the configuration result matrix for configuring the model shown in fig. 11 into 5 process instances included in fig. 4.
Table 2 configuration result matrix of the embodiment of the present invention
| ID | g1 | g2 | g3 | | g5 | g6 | |
| 1 | AND-B,C | AND-B,C | SEQ-E | N/A | SEQ-g6 | SEQ- |
|
| 2 | SEQ-B | SEQ-B | SEQ-E | N/A | SEQ-g6 | SEQ- |
|
| 3 | AND-B,C | AND-B,C | SEQ-E | SEQ-g5 | SEQ-g4 | SEQ- |
|
| 4 | AND-B,C | AND-B,C | AND-E,g4 | SEQ-g3 | SEQ-g6 | AND-g5, |
|
| 5 | AND-B,C | AND-B,C | SEQ-E | N/A | SEQ-H | SEQ-H |
Then, the invention adopts a method of association rule mining to generate a configuration rule based on the configuration result matrix. Association rule mining methods were initially applied to supermarket shopping basket analysis with the goal of finding associations between items purchased by customers, such as the well-known "beer and diaper" case. The embodiment of the invention adopts an association rule mining classic Apriori algorithm to process the configuration result matrix. The core idea of the Apriori algorithm is that all frequent item sets meeting the minimum support degree are generated from small to large according to the scale of the item sets, and then the rule meeting the minimum confidence degree among the items contained in the frequent item sets is generated on the basis of the frequent item sets. For example given support and confidence parameters: s-0.5 and C-0.75, with the configurable process model shown in fig. 11 and the set of process instances contained in fig. 4 as inputs, the result of the Apriori algorithm is shown in fig. 13. The support degree and the confidence coefficient parameters determine the result generated by the configuration rule, a user can set the parameters according to the self requirement, and when the association relation is strictly included, the confidence coefficient can be set at a higher level; when the configuration rule is required to have strong applicability, and only the commonly used configuration result is considered, the support degree parameter should be increased. To this point, configurable process models (including model structures and configuration rules) for uniformly managing a set of process instances have been completely built.
Step P04: and the modeler uses the constructed configurable process model to uniformly manage the process instance set, monitors the process model library and responds to the change of the process instance set at any time.
Step P05: when a new process instance is added to the set of process instances, the structure of the configurable process model evolves automatically in an incremental manner. As shown in fig. 14, a new process instance is added to the model base by replacing the original damage assessment tasks (activity B and activity C) with a new task (activity X) in the newly added processing scheme due to department business tuning. The configurable process model needs to evolve along with the change of the process instance set, so that the configurable process model can represent all 6 process instances including the original 5 process instances and the newly-added model, and the specific steps of the evolution are as follows:
first, a node matching relationship between the configurable process model and the newly added process instances is established based on word similarity (Lin similarity). Nodes that exist in the newly added process instance but do not have a matching relationship with any node in the configurable process model, such as activity X in fig. 14, are identified by node matching relationships and inserted into the configurable process model.
Second, after inserting a new node into the configurable process model, edges in the newly added process instance that do not conform to the configurable process model need to be identified. Edge e in the newly added process model<n1i,n1j>Conforming to a configurable process model G if and only if there is a node n in G2iAnd n2jSo that n is2iAnd n1iN is2jAnd n1jThere is a matching relationship between them, and node n2iAnd n2jThere is a path between and there are no non-configurable nodes on the path. Inserting edges that do not conform to the configurable process model into the configurable process model, such as edges that conform to the configurable process model in the newly added model of FIG. 14, includes:<start,A>、<D,E>、<E,F>and<F,end>the non-compliant edge includes<A,X>And<X,D>. The results obtained after inserting nodes and edges in the example configurable process model are shown in FIG. 15.
Thirdly, performing structural optimization on the configurable process model according to the fourth and fifth steps in step P02, that is, inserting the configurable XOR gateway node so that all branches or merges occur at the gateway node, and then simplifying the redundant gateway nodes, where the finally evolved configurable process model obtained after the structural optimization of the configurable process model in fig. 15 is shown in fig. 16.
Step P06: when a process is deleted from the process instance set, the structure of the configurable process model automatically evolves, and the evolution mechanism is similar to that of the newly added process instance, but the edge id mark is additionally needed as an auxiliary. The manner of evolution is incremental. The method comprises the steps of deleting id of a process instance needing to be deleted from marks of all edges of a configurable process model, directly deleting the edge when a certain edge does not contain any mark, then deleting an isolated node in the model, and simplifying useless gateway nodes (namely gateways with one edge being in and one edge being out) to obtain a new configurable process model structure.
Step P07: the structure of the configurable process model evolves automatically when a process instance from the set of process instances is modified. Since the modification of the process instance can be equivalent to two steps of deleting the original process instance and adding a new process instance, the invention directly performs the combination of the step P05 and the step P06 when processing the evolution of the configurable process model in the situation.
Step P08: the configuration rules contained in the configurable process model evolve automatically. And generating a new configuration rule according to a configuration result corresponding to the new process instance set by utilizing an association rule mining method, wherein the specific algorithm adopts an association rule incremental updating algorithm FUP to improve the processing efficiency. For example, the results of the evolution of the configuration rules for the configurable process model after the change of the newly added process instance as shown in FIG. 14 are shown in FIG. 17. And the configurable process model continues to enter a waiting state in order to adapt to the completion of the evolution of the process instance set, monitors the process model library and responds to the next change of the process instance set at any time.
The method of the invention provides a process instance set evolution management method based on a configurable process model in an intuitive, simple and effective mode, solves the problems of high redundancy, inconsistent evolution and the like caused by the existence of similar process instances in a process model library to a certain extent, constructs the configurable process model containing configuration rules by a one-time process instance combination method, adopts the model to carry out uniform description and management on the process instance set, can automatically carry out incremental evolution along with the change of the process instance set, and covers the structure and the configuration rules of the configurable process model.
Claims (9)
1. An automatic construction method of a configurable process model, taking a process instance set describing a specific service as an input, is characterized by comprising the following steps:
step one, defining a configurable process model for uniformly managing a process instance set;
step two, merging all process instances in the process instance set, generating a structure of the configurable process model and marking the configurable nodes, wherein the step comprises the following steps:
establishing a node matching relationship between process instances by taking all the process instances as input, combining the matched nodes, combining edges with the same source node and target node, inserting the edges into an XOR gateway optimization model structure for structural processing, and enabling all input/output control flows to be unique;
generating configurable nodes, a node being configurable if and only if the node contains all input-output edges whose labels are not identical;
step three, generating a configuration incidence relation between the configurable nodes to obtain a complete configurable process model, comprising the following steps:
firstly, identifying a configuration result of a configurable node in each process instance;
then, a set of configuration rules is generated based on the configuration result matrix using an association rule mining algorithm.
2. The method of claim 1, wherein the configurable process model defined in step one is a six-tuple (N, E, L, T, B, S), wherein:
n represents the set of all nodes in the model;
·
representing a set of directed edges between nodes;
n → label represents the name of each node in the model, label;
n → type represents the type of the node, including activity, event, gateway;
n → { true, false } is a Boolean function for specifying whether the model node is a configurable node;
s is a series of sets of configuration rules CR, one describing the association between the two node configuration results.
3. The method for automatically constructing a configurable process model according to claim 1, wherein the step two is implemented by:
(2.1) preprocessing the input process examples, matching the similarity of the activity names among the process examples by adopting a word similarity method, establishing a node matching relation among the process examples, and marking identification numbers id of the process examples on the edges of all the process examples;
(2.2) combining the nodes matched with each other into a node, and keeping the connection of edges between the nodes; if the attributes of the matched nodes are different, the attribute value with the highest frequency of occurrence in the process instance is reserved, and other attribute values are recorded in the form of node marks;
(2.3) merging edges with the same starting end and the same stopping end, traversing all edges of the model, merging the edges with the same source node and the same target node into one edge, and merging process instance ids marked on the edges;
(2.4) inserting an XOR gateway optimization model structure such that all branching or merging occurs at the gateway node, the XOR gateway optimization model structure comprising: 1) if the input edge of the active node n is not unique, inserting an XOR-Join type gateway node g before n, setting the target nodes of all the input edges of the active node n as g, and inserting an edge e taking g as a source node and n as a target nodenewEdge enewIs marked as the union of all the input edge marks; 2) if the output edge of the active node n is not unique, inserting an XOR-Split type gateway node g after n, setting the source nodes of all the output edges of n as g, and inserting an edge e taking n as the source node and g as the target nodenewEdge enewIs marked as the union of all output edge marks;
(2.5) simplifying the model according to the following two simplification rules: 1) merging continuous homogeneous gateway nodes; 2) deleting useless gateway nodes;
and (2.6) generating a configurable node, and obtaining the configurable process model marked with the configurable node.
4. The method for automatically constructing a configurable process model according to claim 1, wherein the step three of identifying the configuration result of the configurable node in each process instance specifically comprises: establishing a node matching relation between a configurable process model CG and a process instance G based on word similarity, and identifying a configuration result of a configurable node n as follows:
if node n is a configurable activity, its configuration result conf (n) ═ ON if and only if there is an activity n' in G that has a node matching relationship with it, otherwise its configuration result conf (n) ═ OFF;
if the node n is a configurable gateway and a gateway n 'exists in the G and has a node matching relationship with the node n', the configuration result conf (n) ═ Type (n ') -n' ·, where n is a branch Type and n '·denotesa subsequent node set of n'; or n configuration result conf (n) ═ Type (n ') - · n', where n is merge Type, and · n 'represents the set of predecessor nodes of n'; if there is no matching relationship between the gateway and the node n in G, if n is a branch or merge type gateway, it is necessary to search for the node existing in the matching relationship in the successor or predecessor node, if the node n meeting the condition is found1Then n is configured as result of conf (n) ═ SEQ-n1Else node n is a configurable node that is invalid for process instance G; SEQ indicates that one of the paths connected to the configurable node is selected and the configurable node is deleted.
5. The method of claim 1, wherein the configuration rule is expressed in the form of:
CR＝Config,“→”,Config,“:”,Param；
wherein, Config is a rule header or a rule body, and is denoted as Config ═ NID, "═ Result; NID is id of the configurable node, and Result is the configuration Result;
result is denoted as Result ═ ActRes | GtwRes; ActRes is a configuration result of configurable activity, GtwRes is a configuration result of configurable gateway, GtwRes is denoted as GtwRes ═ Type, "-", Branches; the Type is the Type of the configuration result of the configurable gateway, and the Branches is the path selected by the configuration result of the configurable gateway;
param is a parameter of the configuration rule, and is denoted as Param ═ S ═ Value1, ",", "C ═ Value 2; value1 indicates the support of the configuration rule, and Value2 indicates the confidence of the configuration rule.
6. A process instance set evolution management method is realized on the basis of the configurable process model generated by the automatic construction method of the configurable process model as claimed in any one of claims 1 to 5, and is characterized in that a process model library is monitored, the process instance set is waited for to change, when the process instance set changes, the structure of the configurable process model automatically evolves, and the configuration incidence relation contained in the configurable process model automatically evolves;
the case of a process instance set change is: (1) adding a new process instance to the set of process instances; (2) deleting a process instance from the set of process instances; (3) the user modifies a process instance in the set of process instances.
7. The process instance collection evolution management method of claim 6, wherein the automatic evolution of the configurable process model structure occurs when the new process instance is added to the process instance collection, and comprises:
firstly, establishing a node matching relationship between a configurable process model and a newly added process instance based on word similarity, identifying a newly added node, and inserting the newly added node into the configurable process model;
secondly, identifying edges which do not conform to the configurable process model in the newly added process instance, and inserting the edges into the configurable process model;
and finally, inserting an XOR gateway optimization model structure for structural processing to enable all input/output control flows to be unique, and combining and deleting redundant gateway nodes to obtain the evolved configurable process model.
8. The process instance collection evolution management method of claim 6, wherein the automatic evolution of the configurable process model structure when a process instance is deleted from the process instance collection comprises: deleting the id of the process instance needing to be deleted from the marks of all edges of the configurable process model; deleting an edge directly when the edge does not contain any mark; deleting isolated nodes in the configurable process model; redundant gateway nodes are simplified.
9. The evolution management method for process instance set according to claim 6, wherein the automatic evolution of the configuration association relationship included in the configurable process model is: and generating a new configuration rule according to the configuration result corresponding to the new process instance set by using an association rule mining method.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710188272.2A CN106980935B (en) | 2017-03-27 | 2017-03-27 | Process instance set evolution management method based on configurable process model |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710188272.2A CN106980935B (en) | 2017-03-27 | 2017-03-27 | Process instance set evolution management method based on configurable process model |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106980935A CN106980935A (en) | 2017-07-25 |
| CN106980935B true CN106980935B (en) | 2020-06-30 |
Family
ID=59339054
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710188272.2A Active CN106980935B (en) | 2017-03-27 | 2017-03-27 | Process instance set evolution management method based on configurable process model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106980935B (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107479893B (en) * | 2017-08-21 | 2020-12-22 | 淮阴师范学院 | Formalized Petri network model automatic evolution method adapting to demand change |
| CN107943569B (en) * | 2017-10-31 | 2021-07-06 | 北京航空航天大学 | Time analysis method for early decision in real-time system based on feature model |
| CN112579574B (en) * | 2020-12-25 | 2022-08-09 | 上海交通大学 | Configurable process mining method and system based on event log |
| CN112333292B (en) * | 2021-01-06 | 2021-05-04 | 苏州光格科技股份有限公司 | Electric power internet of things gateway edge calculation method |
| CN113139794A (en) * | 2021-01-29 | 2021-07-20 | 江阴逐日信息科技有限公司 | Automatic configurable process evolution method, system and equipment |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102779296A (en) * | 2012-06-21 | 2012-11-14 | 浙江工商大学 | Process execution cost determination method based on combination of activities and resources |
| CN102799960A (en) * | 2012-06-18 | 2012-11-28 | 北京大学 | Parallel operation flow anomaly detection method oriented to data model |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10176444B2 (en) * | 2010-05-28 | 2019-01-08 | International Business Machines Corporation | Computer-implemented method, computer program product and system for analyzing a control-flow in a business process model |
-
2017
- 2017-03-27 CN CN201710188272.2A patent/CN106980935B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102799960A (en) * | 2012-06-18 | 2012-11-28 | 北京大学 | Parallel operation flow anomaly detection method oriented to data model |
| CN102779296A (en) * | 2012-06-21 | 2012-11-14 | 浙江工商大学 | Process execution cost determination method based on combination of activities and resources |
Non-Patent Citations (2)
| Title |
|---|
| Generating Hierarchical Reference Process Model using Fragments Clustering;Jimin Ling et al;《2015 Asia-Pacific Software Engineering Conference》;20151204;第40-47页 * |
| 业务过程的自动配置与管理;吴亚洲等;《计算机科学》;20150731;第42卷(第7期);第129-141页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106980935A (en) | 2017-07-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106980935B (en) | Process instance set evolution management method based on configurable process model | |
| US11107036B2 (en) | Systems and methods for business processing modelling | |
| US7848942B2 (en) | Distribution of integrated business process models | |
| US20060143057A1 (en) | Integration of distributed business process models | |
| US8776009B2 (en) | Method and system for task modeling of mobile phone applications | |
| Zimmermann | Architectural refactoring for the cloud: a decision-centric view on cloud migration | |
| Ferreira et al. | Using logical decision trees to discover the cause of process delays from event logs | |
| Assy et al. | An automated approach for assisting the design of configurable process models | |
| CN108536718A (en) | A kind of method and system for the IT application in management realized based on input and output semantization | |
| CN106325887A (en) | Service process management method based on Activiti | |
| WO2022048164A1 (en) | Smart robot dialogue policy generation method based on granular computing | |
| Kotusev | Eight essential enterprise architecture artifacts | |
| CN112989059A (en) | Method and device for identifying potential customer, equipment and readable computer storage medium | |
| CN106779336A (en) | Engineering change (EC) method and device | |
| Park et al. | A goal-oriented big data analytics framework for aligning with business | |
| CN107479893B (en) | Formalized Petri network model automatic evolution method adapting to demand change | |
| CN107220044A (en) | Method and apparatus based on meta-model driven business object | |
| Weske et al. | Process Orchestrations | |
| US20220245153A1 (en) | Automated code generation for computer software | |
| Souza et al. | Automating the identification of opportunities for business process improvement patterns application | |
| CN114548916A (en) | Approval method for complex workflow and specified approver | |
| Wexler et al. | Time Is Precious, So Are Your Models. SAS provides solutions to streamline deployment | |
| US20190079967A1 (en) | Aggregation and deduplication engine | |
| US9141650B1 (en) | Methods and systems for filtering tree node data to determine part and assembly relationships | |
| Ahmadikatouli et al. | Enrichment of object oriented Petri net and object Z aiming at business process optimization |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20220120 Address after: 215488 No. 301, building 11, phase II, Taicang University Science Park, No. 27, Zigang Road, science and education new town, Taicang City, Suzhou City, Jiangsu Province Patentee after: Tianhang Changying (Jiangsu) Technology Co.,Ltd. Address before: 100191 No. 37, Haidian District, Beijing, Xueyuan Road Patentee before: BEIHANG University |
|
| TR01 | Transfer of patent right |