CN115904748A - Method for detecting compliance of layered business process model based on alignment - Google Patents

Abstract

The invention discloses a method for detecting the compliance of a layered service process model based on alignment, which comprises the steps of mining a layered transition nesting relation; the hierarchical process model hpn is used as input to carry out excavation of nesting relation among models, and the output result is a hierarchical transition nesting relation tree ang; constructing a hierarchical event log; taking an event log xlog with a life cycle and a hierarchical transition nesting relation tree ang as input, and constructing a hierarchical event log hlog by analyzing a model hierarchical nesting relation; constructing an alignment sequence relation nested tree; taking a layering event log hlog and a layering process model hpn as input, and carrying out compliance detection on a layering structure to obtain an alignment sequence nesting relation tree hat; merging the layered alignment sequences; and merging the nested relation trees of the alignment sequences to obtain a final result. Compared with the existing method for detecting the compliance of the layered model, the method reduces the alignment time of the log and the model, and is beneficial to reducing the time overhead required by the repair and optimization of the model.

Description

Alignment-based layered business process model compliance detection method

Technical Field

The invention relates to the field of process mining, in particular to a method for detecting the compliance of a layered service process model based on alignment.

Background

Process mining, a new research hotspot in the Business Process Management (Business Process Management) field, aims to build a bridge between traditional model-driven methods (such as Business Process modeling and model correctness verification) and new data-driven methods (such as data mining and machine learning). Compliance testing is an important component of process mining, and is a technique for testing the degree of matching between a process model and a log, i.e., how much the model can replay (replay) the essence of a process performed by a log-related flow. Usually, enterprises do not implement systematic modeling work in the initial stage, but a large amount of business activity execution flows are recorded in internal software systems of the enterprises. When the scale of an enterprise is gradually increased and the business is more and more complicated, the enterprise adopts some automatic mining technologies to mine a working process model on the basis of a large amount of actual logs so as to better manage the performance so as to improve the working efficiency, and the quality of the mining model is generally required to be evaluated through compliance inspection. In addition, when the log generated by the process model execution has a large deviation from the original model, that is, the compatibility between the log and the original model is lower than a given threshold value, it is necessary to perform a correction operation on the original model.

The existing process model compliance detection technology comprises the following steps: (1) The royal ad riansyah proposes that a product net (product net) of a process model and a process model corresponding to a track is constructed, a reachable graph of the product net is constructed, the calculation problem of optimal alignment is converted into a problem of searching for an optimal path from an initial state to a termination state in the reachable graph, the problem can be solved by an a-search algorithm, and the quality of the process model is evaluated by using a fitness evaluation index. (2) de Leoni et al teaches a method of aligning event logs and declarative models and provides a sophisticated diagnostic method that accurately indicates where the deviation occurs and the severity of the deviation. The above method alignment is defined on the basis of a flat model, and is not suitable for a hierarchical model with a hierarchical structure. The relationship between the four measurement indexes of integrating degree, accuracy, generality and complexity and the importance of model quality evaluation discussed in detail by doctor Joos Buijs are all defined on the basis of a flat process model and cannot be applied to a layered model containing sub-process behaviors.

Aiming at the problem, a professor and the like of Wen Lijie of Qinghua university provides a compliance detection algorithm Acorn based on a BPMN model, and the compliance detection algorithm Acorn is only based on the BPMN model. Professor Liu Cong, university of Shandong science & engineering, teaches a method for converting a hierarchical model with sub-processes into a flat model. However, in the real world, log files and corresponding models are usually large, and this method of converting a hierarchical model into a flat model for compliance detection needs to consume a lot of time and memory.

Disclosure of Invention

Aiming at the problems in the prior art, the invention provides a method for detecting the compliance of a layered business process model based on alignment.

The invention adopts the following technical scheme:

the method for detecting the compliance of the layered service process model based on alignment comprises the following steps:

step 1: excavating a hierarchical transition nesting relation; the hierarchical process model hpn is used as input to carry out excavation of nesting relations among models, and the output result is a hierarchical transition nesting relation tree ang which is used for describing all possible nesting relations of the hierarchical models;

and 2, step: constructing a hierarchical event log; taking an event log xlog with a life cycle and a hierarchical transition nesting relation tree ang as input, and constructing a hierarchical event log hlog by analyzing a model hierarchical nesting relation;

and step 3: constructing an alignment sequence relation nested tree; taking a layering event log hlog and a layering process model hpn as input, and carrying out compliance detection on a layering structure to obtain an alignment sequence nesting relation tree hat;

and 4, step 4: merging the layered alignment sequences; and (4) merging the alignment sequence nested relation trees obtained in the step (3) to obtain a final result.

Preferably, the step 1 specifically includes:

inputting: a layered process model hpn;

and (3) outputting: a hierarchical transition nested relation tree ang;

step 1.1: calling an algorithm getActivityNestedSet (), getActivityPair (), and mining a nested transition set ActivityNestedSet [ ] and a nested transition association pair set ActivityPariSet [ ] in the hierarchical model;

step 1.2: constructing a hierarchical transition nesting relation tree ang according to the nesting transition association pair by calling an algorithm ActivityGraphConstruction ();

step 1.3: and returning to the hierarchical transition nested relation tree ang.

Preferably, the function of the invoked getactivtynestedset () function is to return all nested transitions in the hierarchical model, specifically described as:

inputting: a layered process model hpn;

and (3) outputting: nesting a transition set activityNestedSet [ ];

traversing the top Petri net, finding out the nested transitions contained in the top Petri net, storing the nested transitions into a set activitySet [ ], recursively traversing the nested hierarchical Petri net of the top Petri net, and returning all nested transition sets in the hierarchical model;

the function of the called getActivityPair () function is to return a nested transition association pair in the hierarchical model, and the specific description is as follows:

inputting: a layered process model hpn;

and (3) outputting: a set of nested transition association pairs activityPairSet [ ];

extracting nested transition t from the nested relation tree ang of the hierarchical transition, assigning the nested transition t to a variable source, and finding out a nested hierarchical model hpn of the nested transition t _i (ii) a Find the hierarchical model hpn _i Nested transition of lower t _i And storing the target; storing (source, target) nested transition association pairs into a set activityPairSet [ 2 ]]The preparation method comprises the following steps of (1) performing;

nested hierarchical Petri net hierarchical model hpn for recursively traversing nested transition t of top-level Petri net _i (ii) a Return set activityparilset [ 2 ]]；

The function of the invoked activity graph structure () function is to return a transition nested relation tree ang in the hierarchical model, and the specific description is as follows:

inputting: nesting migration association pair set activityPariSet [ ];

and (3) outputting: a hierarchical transition nested relation tree ang;

sequentially extracting nested transition association pairs from the activityPairSet [ ], and storing nodes and corresponding edges in the nested transition association pairs into a tree structure; and returning to the hierarchical transition nested relation tree ang.

Preferably, the step 2 specifically includes:

inputting: a hierarchical transition nested relation tree ang and an event log xlog with a life cycle;

and (3) outputting: a hierarchical event log hlog;

sequentially extracting nodes from the hierarchical transition nested relation tree ang and storing the nodes into all NestedActities [ ];

respectively establishing root log activities and top-layer nested log activities under the root log activities according to nodes in a hierarchical transition nested relation tree ang;

extracting log activities in an event log xlog with a life cycle corresponding to the root log activities and assigning the log activities to a mainLog;

log activities in an event log xlog with a life cycle corresponding to top-layer nested log activities under the root log activities are extracted and assigned to the submainLog;

recursively traversing other nodes in the hierarchical transition nested relation tree ang;

a hierarchical event log hlog is returned.

Preferably, the step 3 specifically includes:

inputting: a hierarchical process model hpn, a hierarchical event log hlog;

and (3) outputting: aligning a sequence nesting relation tree hat;

extracting a root log rootlog from a hierarchical event log hlog;

extracting a top-layer Petir net pn from the hierarchical model;

aligning pn and the roothlog to obtain a top-layer alignment sequence rootA;

nested hierarchical model for top Petri nets hpn _i Hierarchical event Log Roothlog nested with root Log _i And recursively obtaining an alignment sequence, and returning to the nesting relation tree hat of the alignment sequence.

Preferably, the step 4 specifically includes:

inputting: aligning a sequence nesting relation tree hat;

and (3) outputting: an alignment sequence r;

when the alignment sequence nesting relation tree hat is not empty, extracting a top-level alignment sequence in the alignment sequence nesting relation tree hat and pushing the top-level alignment sequence into a set rootA;

pushing the elements in the set rootA into a queue st;

and sequentially extracting the elements in the queue st and checking whether the elements are nested transitions, namely tags _i Whether or not it is 1;

if tag _i Not equal to 1, pushing the current node into a queue r;

if tag _i =1, push current node into queue r, and extract aligned sequence nested relation tree hat _i The top layer in (1) aligns the sequences and pushes them into the set rootA; recursive traversal hat _i

Up to the queue

Returning to the final alignment sequence r.

The invention has the beneficial effects that:

the method for detecting the compliance of the layered service process model based on the alignment is a new method for detecting the compliance of the layered model, and compared with the existing method for detecting the compliance of the layered model, the method reduces the alignment time of the log and the model, and is beneficial to reducing the time overhead required by the repair and optimization of the model. The alignment-based hierarchical business process model compliance detection method requires less time overhead than existing hierarchical model compliance detection techniques.

Drawings

FIG. 1 is an example of a layered process model hpn in example 1.

FIG. 2 is a hierarchical transition nested relationship tree obtained from the hierarchical process model of FIG. 1.

FIG. 3 is the layered process model hpn in example 1 ₁ 。

FIG. 4 is a hierarchical event log hl ₁ 。

FIG. 5 is a layered process model hpn ₂ 。

FIG. 6 is a hierarchical event log hl ₂ 。

Fig. 7 is an alignment sequence nesting relationship tree hat obtained in example 1.

FIG. 8 is a schematic diagram of a merged alignment sequence.

FIG. 9 shows an integration r ₁ Schematic representation of alignment sequences.

FIG. 10 shows an integration r ₂ Schematic representation of alignment sequences.

FIG. 11 shows an integration r ₃ Schematic representation of alignment sequences.

Fig. 12 shows the time overhead for a noise threshold of 2.0.

Fig. 13 shows the time overhead for a noise threshold of 3.0.

Fig. 14 shows the time overhead for a noise threshold of 5.0.

FIG. 15 illustrates the overhead in time of different noise threshold generation times when log sizes are fixed.

Detailed Description

The following description of the embodiments of the present invention will be made with reference to the accompanying drawings:

with reference to fig. 1 to 11, the method for detecting compliance of a hierarchical business process model based on alignment includes the following steps:

step 1: excavating a hierarchical transition nesting relation; the hierarchical process model hpn is used as input to perform mining of nesting relations among models, and the output result is a hierarchical transition nesting relation tree ang which is used for describing all possible nesting relations of the hierarchical models.

The method specifically comprises the following steps:

inputting: a layered process model hpn;

and (3) outputting: a hierarchical transition nested relation tree ang;

step 1.1: calling an algorithm getActivityNestedSet (), getActivityPair (), and mining a nested transition set ActivityNestedSet [ ] and a nested transition association pair set ActivityPariSet [ ] in the hierarchical model;

step 1.2: constructing a hierarchical transition nesting relation tree ang according to a nesting transition association pair by calling an algorithm ActivityGraphConstruction ();

step 1.3: and returning to the hierarchical transition nested relation tree ang.

The called getActivityNestedSet () function has the main function of returning all nested transitions in the hierarchical model, and is specifically described as follows:

inputting: a layered process model hpn;

and (3) outputting: nesting an activity NestedSet [ ];

traversing the top Petri net, finding out the nested transitions contained in the top Petri net, storing the nested transitions into a set activitySet [ ], recursively traversing the nested hierarchical Petri net of the top Petri net, and returning all nested transition sets in the hierarchical model.

The main function of the invoked getActivityPair () function is to return a nested transition association pair in the hierarchical model, and the specific description is as follows:

inputting: a layered process model hpn;

and (3) outputting: a set of nested transition association pairs activityPairSet [ ];

extracting nested transition t from the nested relation tree ang of the hierarchical transition, assigning the nested transition t to a variable source, and finding out a nested hierarchical model hpn of the nested transition t _i (ii) a Find the hierarchical model hpn _i Nested transition of lower t _i And storing the target; storing a (source, target) nested transition association pair in the set activitypPairSet [ 2 ]]In (1).

Recursively traversing the top levelNested hierarchical Petri net hierarchical model hpn of Petri net nested transition t _i (ii) a Return set activityparilset [ 2 ]]。

The main function of the invoked activityGraphConstruction () function is to return to a transition nesting relationship tree ang in the hierarchical model, which is specifically described as follows:

inputting: nesting migration association pair set activityPariSet [ ];

and (3) outputting: a hierarchical transition nesting relation tree ang;

sequentially extracting nested transition associated pairs from activitypair set [ ], and storing nodes and corresponding edges in the nested transition associated pairs into a tree structure; and returning to the hierarchical transition nested relation tree ang.

For the example of a hierarchical process model hpn in fig. 1, a hierarchical transition nested relation tree is obtained as shown in fig. 2, where (1) a and c are root nodes; and (2) a and b satisfy the nesting relation, and b and e satisfy the nesting relation.

Step 2: constructing a hierarchical event log; and (3) taking an event log xlog with a life cycle and a hierarchical transition nesting relation tree ang as input, and constructing a hierarchical event log hlog by analyzing the hierarchical nesting relation of the model.

The method specifically comprises the following steps:

inputting: a hierarchical transition nested relation tree ang and an event log xlog with a life cycle;

and (3) outputting: a hierarchical event log hlog;

sequentially extracting nodes from the hierarchical transition nested relation tree ang and storing the nodes into all NestedActities [ ];

respectively establishing root log activities and top-level nested log activities under the root log activities according to nodes in a hierarchical transition nested relation tree ang;

extracting log activities in an event log xlog with a life cycle corresponding to the root log activities and assigning the log activities to a mainLog;

log activities in an event log xlog with a life cycle corresponding to top-layer nested log activities under the root log activities are extracted and assigned to a submainLog;

recursively traversing other nodes in the hierarchical transition nested relation tree ang;

a hierarchical event log hlog is returned.

The layered process model hpn in FIG. 3 ₁ And its event log L with life cycle ₁ ＝{<a _s ,b _s ,b _c ,a _c > ⁹⁰ ,<a _s ,b _s ,a _c ,b _c > ¹ Example, layered model hpn ₁ And its event log L with life cycle ₁ ＝{<a _s ,b _s ,b _c ,a _c > ⁹⁰ ,<a _s ,b _s ,a _c ,b _c > ¹ As input, a hierarchical event log hl can be obtained ₁ As shown in fig. 4, the root log thereof is rootLog = ∑ final record<a _s ,a _c > ⁹¹ Its nested task set NA (rootLog) = { a }. The sub-log corresponding to the nested task a is NLoga = &<b _s ,b _c > ⁹⁰ }。

And step 3: constructing an alignment sequence relation nested tree; taking a layering event log hlog and a layering process model hpn as input, and carrying out compliance detection on a layering structure to obtain an alignment sequence nesting relation tree hat;

inputting: a hierarchical process model hpn, a hierarchical event log hlog;

and (3) outputting: aligning a sequence nesting relation tree hat;

extracting a root log rootlog from a hierarchical event log hlog;

extracting a top-layer Petir net pn from the hierarchical model;

aligning pn and the roothlog to obtain a top-layer alignment sequence rootA;

nested hierarchical model for top Petri nets hpn _i Nested with root Log hierarchical event Log Roothlog _i And recursively calculating an alignment sequence, and returning to the nested relation tree hat of the alignment sequence.

Layered Process model hpn in FIG. 5 ₂ And its event log L with life cycle ₂ ＝{<a _s ,b _s ,d _s ,d _c ,b _c ,a _c ,c _s ,c _c > ⁹⁹ ,<c _s ,a _s ,b _s ,c _c ,d _s ,d _c ,b _c ,a _c > ⁹⁶ ,<a _s ,c _s ,b _s ,d _s ,d _c ,b _c ,a _c ,c _c > ⁸⁶ ,<a _s ,b _s ,d _s ,c _s ,d _c ,b _c ,a _c ,c _c > ⁷⁸ ,<a _s ,b _s ,d _s ,d _c ,c _s ,b _c ,a _c ,c _c > ⁷⁹ ,<a _s ,b _s ,d _s ,d _c ,b _c ,c _s ,a _c ,c _c > ⁸² ,<a _s ,c _s ,b _s ,d _s ,c _c ,d _c ,b _c ,a _c > ⁹⁸ ,<a _s ,b _s ,c _c ,d _s ,d _c ,b _c ,c _c ,a _c > ⁸⁵ ,<a _s ,b _s ,c _s ,d _s ,c _c ,d _c ,b _c ,a _c ,c _c > ¹⁰⁰ Take the example.

Using a hierarchical model hpn ₂ And a log L with a life cycle ₂ For input, get hierarchical event Log hl ₂ . Wherein its root log is rootLog = &<a _s ,a _c ,c _s ,c _c > ⁹⁹ ,<c _s ,a _s ,c _c ,a _c > ⁹⁶ ,<a _s ,c _s ,a _c ,c _c > ⁸⁶ ,<a _s ,c _s ,a _c ,c _c > ⁷⁸ ,<a _s ,c _s ,a _c ,c _c > ⁷⁹ ,<a _s ,c _s ,a _c ,c _c > ⁸² ,<a _s ,c _s ,c _c ,a _c > ⁹⁸ ,<a _s ,c _c ,c _c ,a _c > ⁸⁵ ,<a _s ,c _s ,c _c ,a _c ,c _c > ¹⁰⁰ Its nested task set NA (rootLog) = { a }. The sub-log corresponding to the nested task a is NLog _a ＝{<b _s ,b _c > ⁹⁹ ,<b _s ,b _c > ⁹⁶ ,<b _s ,b _c > ⁸⁶ ,<b _s ,b _c > ⁷⁸ ,<b _s ,b _c > ⁷⁹ ,<b _s ,b _c > ⁸² ,<b _s ,b _c > ⁹⁸ ,<b _s ,b _c > ⁸⁵ ,<b _s ,b _c > ¹⁰⁰ }, its nested task set NA (NLog) _a ) = b. Sub-log NLog corresponding to nested task b _b ＝{<d _s ,d _c > ⁹⁹ ,<d _s ,d _c > ⁹⁶ ,<d _s ,d _c > ⁸⁶ ,<d _s ,d _c > ⁷⁸ ,<d _s ,d _c > ⁷⁹ ,<d _s ,d _c > ⁸² ,<d _s ,d _c > ⁹⁸ ,<d _s ,d _c > ⁸⁵ ,<d _s ,d _c > ¹⁰⁰ Their existing nesting relationship is shown in fig. 6.

Using a layered process model hpn ₂ And a hierarchical event log hl ₂ An aligned sequence nesting relationship tree hat is obtained for the input, as shown in FIG. 7.

And 4, step 4: merging the layered alignment sequences; and (4) merging the alignment sequence nested relation trees obtained in the step (3) to obtain a final result.

The method specifically comprises the following steps:

inputting: aligning a sequence nesting relation tree hat;

and (3) outputting: an alignment sequence r;

when the alignment sequence nesting relation tree hat is not empty, extracting a top-level alignment sequence in the alignment sequence nesting relation tree hat and pushing the top-level alignment sequence into a set rootA;

pushing the elements in the set rootA into a queue st;

and sequentially extracting the elements in the queue st and checking whether the elements are nested transitions, namely tags _i Whether the value is 1;

if tag _i Not equal to 1, pushing the current node into a queue r;

if tag _i =1, push current node into queue r, and extract aligned sequence nested relation tree hat _i The top layer in (1) aligns the sequences and pushes into the set rootA; recursive traversal hat _i

Up to the queue

Returning to the final alignment sequence r.

Taking the nested relation tree hat of the alignment sequence in fig. 7 as an example, the hierarchical alignment sequence is integrated. rootA = { r = { (r) ₁ ,r ₂ ,r ₃ … } the elements are sequentially pushed into a queue st; head of line element r ₁ Pushing out the queue st, pointing the queue head pointer to the successor node r of the queue head element, and traversing r ₁ Tag of node a _i =1, that is, the node a is a nested transition, the element before the node a is pushed into the queue, the hierarchical alignment sequence nested by the node a is recursively traversed, and the other nodes in the rootA are sequentially traversed similarly, and the specific steps are as shown in fig. 8, fig. 9, fig. 10, and fig. 11.

Example 1

With reference to fig. 12 to fig. 15, the input data of the hierarchical business process mining method is an event log in which information of a life cycle and a sub-process is recorded, and 2 simulation hierarchical business process models and 2 real hierarchical business process models are selected as evaluation data in the embodiment of the present invention. The following is a detailed introduction of the hierarchical business process model:

two simulation models are the events hpn referred to above ₁ And hpn ₂ . And the two data sets corresponding to the real hierarchical business process model are a public TSEC Log Log and a CRMC Log Log.

The source of the data set:

(1) TSEC Log the data set is generated based on a cross-country e-commerce scenario, which involves two sub-processes.

(2) And CRMC Log, wherein the data set is generated based on an upgrading process of an open source cloud resource management tool Netflix Asgard on the Amazon web service, and the process relates to a sub-process.

The basic information of the data set is shown in table 1.

TABLE 1 basic information of data set

Name of log	Total number of traces	Total number of events	Number of movements
				L 1	101	214	2
L 2	803	3223	4
				TSEC Log	522	14616	10
CRMC Log	626	27544	17

Evaluation index, experiment result and analysis

Log scale evaluation

In order to detect the effectiveness and the availability of the compliance detection technology, logs with different scales and different noise thresholds are used as input and compared with the existing method (Convert a historical Petri Net to a Flat Petri Net), and the performance of the algorithm is displayed by comparing the time overhead.

As shown in fig. 12, it can be found through experiments that, in the case that the noise threshold is 2.0, by using the alignment-based hierarchical business process model compliance detection method, as the log scale is continuously enlarged, the time overhead is continuously increased, and the log scale is smaller, L ₁ The time overhead is obviously smaller than the CRMC Log with larger Log scale; by comparison, it is obvious that the average time consumption of the method provided by the invention is less than that of the proposed Convert a Hierarchical Petri Net to a Flat Petri Net algorithm.

As shown in fig. 13, it can be seen through experiments that in the case of the noise threshold value being 3.0, the time overhead of the alignment-based hierarchical business process model compliance detection method increases as the log scale increases; by contrast, the average time consumption of the method provided by the invention is smaller than that of the proposed Convert a Hierarchical Petri Net to a Flat Petri Net algorithm.

As shown in fig. 14, it can be known through experiments that, in the case of a noise threshold of 5.0, the time overhead increases as the log scale increases by using the alignment-based hierarchical business process model compliance detection method; the average time consumption of the method provided by the invention is less than that of the provided Convert a Hierarchical Petri Net to a Flat Petri Net algorithm.

From the experimental comparison results in fig. 12, 13, and 14, the following conclusions can be drawn: neglecting other situations, only considering alignment, under the condition that the noise threshold is the same, the larger the log scale is, the longer the consumed time is, and the average consumed time of the method provided by the invention is smaller than that of the provided Convert a Hierarchical Petri Net to a Flat Petri Net algorithm.

From the experimental results in fig. 15, the following conclusions can be drawn from comparative analysis: ignoring other things, only alignment is considered, the larger the noise threshold, the longer the elapsed time, and the average elapsed time of the algorithm proposed herein is less than that of the proposed Convert a Hierarchical Petri Net to a Flat Petri Net algorithm, for the same log size.

Through the above experiments, it can be found that the alignment-based hierarchical business process model compliance detection technology provided herein requires less time overhead than existing hierarchical model compliance detection technologies, and the effectiveness and availability of the alignment-based hierarchical business process model compliance detection technology are verified.

It is to be understood that the above description is not intended to limit the present invention, and the present invention is not limited to the above examples, and those skilled in the art may make modifications, alterations, additions or substitutions within the spirit and scope of the present invention.

Claims (6)

1. The method for detecting the compliance of the layered service process model based on alignment is characterized by comprising the following steps of:

step 1: excavating a hierarchical transition nesting relation; the hierarchical process model hpn is used as input to carry out excavation of nesting relations among models, and the output result is a hierarchical transition nesting relation tree ang which is used for describing all possible nesting relations of the hierarchical models;

step 2: constructing a hierarchical event log; taking an event log xlog with a life cycle and a hierarchical transition nesting relation tree ang as input, and constructing a hierarchical event log hlog by analyzing a model hierarchical nesting relation;

and 3, step 3: constructing an alignment sequence relation nested tree; taking a layering event log hlog and a layering process model hpn as input, and carrying out compliance detection on a layering structure to obtain an alignment sequence nesting relation tree hat;

and 4, step 4: merging the layered alignment sequences; and (4) merging the alignment sequence nested relation trees obtained in the step (3) to obtain a final result.

2. The method for detecting the compliance of the alignment-based hierarchical business process model according to claim 1, wherein the step 1 specifically comprises:

inputting: a layered process model hpn;

and (3) outputting: a hierarchical transition nested relation tree ang;

step 1.1: calling an algorithm getActivityNestedSet (), getActivityPair (), and mining a nested transition set ActivityNestedSet [ ] and a nested transition association pair set ActivityPariSet [ ] in the hierarchical model;

step 1.2: constructing a hierarchical transition nesting relation tree ang according to a nesting transition association pair by calling an algorithm ActivityGraphConstruction ();

step 1.3: and returning to the hierarchical transition nested relation tree ang.

3. The alignment-based layered business process model compliance detection method of claim 2,

the function of the called getActivatyNestedSet () function is to return all nested transitions in the hierarchical model, and is specifically described as follows:

inputting: a layered process model hpn;

and (3) outputting: nesting a transition set activityNestedSet [ ];

traversing the top Petri net, finding out the nested transitions contained in the top Petri net, storing the nested transitions into a set activitySet [ ], recursively traversing the nested hierarchical Petri net of the top Petri net, and returning all nested transition sets in the hierarchical model;

the function of the called getActivityPair () function is to return a nested transition association pair in the hierarchical model, and the specific description is as follows: inputting: a layered process model hpn;

and (3) outputting: a set of nested transition association pairs activityPairSet [ ];

extracting nested transition t from the nested relation tree ang of the hierarchical transition, assigning the nested transition t to a variable source, and finding out a nested hierarchical model hpn of the nested transition t _i (ii) a Find the hierarchical model hpn _i Nested transition of lower t _i And storing the target; storing a (source, target) nested transition association pair in the set activitypPairSet [ 2 ]]Performing the following steps;

nested hierarchical Petri net hierarchical model for recursively traversing nested transition t of top-level Petri nethpn _i (ii) a Return set activitypariSet [ 2 ]]；

The function of the invoked activity graph structure () function is to return a transition nested relation tree ang in the hierarchical model, and the specific description is as follows:

inputting: nesting migration association pair set activityPariSet [ ];

and (3) outputting: a hierarchical transition nested relation tree ang;

sequentially extracting nested transition associated pairs from activitypair set [ ], and storing nodes and corresponding edges in the nested transition associated pairs into a tree structure; and returning to the hierarchical transition nested relation tree ang.

4. The alignment-based layered business process model compliance detection method of claim 1, wherein the step 2 specifically comprises:

inputting: a hierarchical transition nested relation tree ang and an event log xlog with a life cycle;

and (3) outputting: a hierarchical event log hlog;

sequentially extracting nodes from the hierarchical transition nested relation tree ang and storing the nodes into all NestedActities [ ];

respectively establishing root log activities and top-level nested log activities under the root log activities according to nodes in a hierarchical transition nested relation tree ang;

extracting log activities in an event log xlog with a life cycle corresponding to the root log activities and assigning the log activities to a mainLog;

log activities in an event log xlog with a life cycle corresponding to top-layer nested log activities under the root log activities are extracted and assigned to a submainLog;

recursively traversing other nodes in the hierarchical transition nested relation tree ang;

a hierarchical event log hlog is returned.

5. The alignment-based layered business process model compliance detection method of claim 1, wherein the step 3 specifically comprises:

inputting: a hierarchical process model hpn, a hierarchical event log hlog;

and (3) outputting: aligning a sequence nesting relation tree hat;

extracting a root log rootlog from a hierarchical event log hlog;

extracting a top-layer Petir net pn from the hierarchical model;

aligning pn and the roothlog to obtain a top-layer alignment sequence rootA;

nested hierarchical model for top Petri nets hpn _i Nested with root Log hierarchical event Log Roothlog _i And recursively obtaining an alignment sequence, and returning to the nesting relation tree hat of the alignment sequence.

6. The method for alignment-based hierarchical business process model compliance detection as claimed in claim 5, wherein the step 4 specifically includes:

inputting: aligning a sequence nesting relation tree hat;

and (3) outputting: an alignment sequence r;

when the alignment sequence nesting relation tree hat is not empty, extracting a top-level alignment sequence in the alignment sequence nesting relation tree hat and pushing the top-level alignment sequence into a set rootA;

pushing the elements in the set rootA into a queue st;

and sequentially extracting the elements in the queue st and checking whether the elements are nested transitions, namely tags _i Whether or not it is 1;

if tag _i Not equal to 1, pushing the current node into a queue r;

if tag _i =1, push current node into queue r, and extract aligned sequence nested relation tree hat _i The top layer in (1) aligns the sequences and pushes into the set rootA; recursive traversal hat _i

Up to the queue

Returning to the final alignment sequence r.

Images