CN111860936A - Method for predicting defects of office business process - Google Patents
Method for predicting defects of office business process Download PDFInfo
- Publication number
- CN111860936A CN111860936A CN202010446235.9A CN202010446235A CN111860936A CN 111860936 A CN111860936 A CN 111860936A CN 202010446235 A CN202010446235 A CN 202010446235A CN 111860936 A CN111860936 A CN 111860936A
- Authority
- CN
- China
- Prior art keywords
- business process
- nodes
- office
- node
- service
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 193
- 230000008569 process Effects 0.000 title claims abstract description 157
- 230000007547 defect Effects 0.000 title claims abstract description 46
- 238000007689 inspection Methods 0.000 claims abstract description 33
- 238000012549 training Methods 0.000 claims abstract description 26
- 230000008520 organization Effects 0.000 claims abstract description 19
- 230000002159 abnormal effect Effects 0.000 claims abstract description 17
- 238000013461 design Methods 0.000 claims abstract description 8
- 238000012545 processing Methods 0.000 claims description 25
- 238000003066 decision tree Methods 0.000 claims description 17
- 238000004458 analytical method Methods 0.000 claims description 12
- 230000006870 function Effects 0.000 claims description 8
- 230000000694 effects Effects 0.000 claims description 7
- 238000012886 linear function Methods 0.000 claims description 6
- 238000007781 pre-processing Methods 0.000 claims description 6
- 238000005259 measurement Methods 0.000 claims description 5
- 238000012417 linear regression Methods 0.000 claims description 4
- 238000007619 statistical method Methods 0.000 claims description 4
- 230000007423 decrease Effects 0.000 claims description 3
- 239000011159 matrix material Substances 0.000 claims description 3
- 238000011002 quantification Methods 0.000 claims description 3
- 238000013468 resource allocation Methods 0.000 claims description 3
- 238000012546 transfer Methods 0.000 claims description 3
- 230000007704 transition Effects 0.000 claims description 3
- 238000004422 calculation algorithm Methods 0.000 claims description 2
- 238000005111 flow chemistry technique Methods 0.000 claims description 2
- 230000003993 interaction Effects 0.000 claims description 2
- 239000013598 vector Substances 0.000 claims description 2
- 238000012896 Statistical algorithm Methods 0.000 claims 1
- 238000004364 calculation method Methods 0.000 claims 1
- 238000010801 machine learning Methods 0.000 abstract description 4
- 238000007726 management method Methods 0.000 description 5
- 238000004891 communication Methods 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 4
- 238000012360 testing method Methods 0.000 description 4
- 238000010276 construction Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 230000005856 abnormality Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000000611 regression analysis Methods 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
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
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/24323—Tree-organised classifiers
-
- 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
- G06Q10/10—Office automation; Time management
- G06Q10/103—Workflow collaboration or project management
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Strategic Management (AREA)
- Human Resources & Organizations (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Economics (AREA)
- Entrepreneurship & Innovation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Computation (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Marketing (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Development Economics (AREA)
- Game Theory and Decision Science (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to a method for predicting defects of office business processes, which comprises the following steps: (1) analyzing the structural characteristics of the office business process inspection; (2) analyzing the organization collaboration characteristics of the office business process inspection; (3) analyzing the resource distribution characteristics of the office business process inspection; (4) analyzing the abnormal mode characteristics of the office business process inspection; (5) making a judging interval of the business process defects under different business lines by a judging group; (6) and taking the structural feature, the organization and cooperation feature, the resource distribution feature and the mode abnormity feature of the office business process inspection as feature dimensions. And taking the service process scores given by the panel as the marking data. And training an office business process defect prediction model through machine learning. The invention comprehensively considers the structural characteristics, the organization and cooperation characteristics, the resource distribution characteristics and the mode abnormity characteristics of the office business process, analyzes the irrationality and the error in the office business process from the design and execution two levels, and helps enterprises to avoid economic loss caused by the defects of the business process.
Description
Technical Field
The invention relates to the field of business process management in an office cooperative system, in particular to construction of a hierarchical decision tree for business process defect inspection and automatic process defect identification based on the hierarchical decision tree.
Background
With the continuous expansion of the scale of modern enterprises, the organization architecture of the enterprises becomes large and complex, the division of labor is thinner and thinner, and the cross-organization cooperation is more and more frequent. Although the enterprises can form written responsibility explanation and system flow through strict business flow management, the problems of unsmooth internal cooperation of organizations, long work decision time of cross-department flow, insufficient refinement degree of system flow, difficult execution of flow treatment and the like still occur.
The root cause of these problems is the defects in the design and implementation of enterprise business process management. These flow defects are various errors that prevent the smooth execution of the workflow, destroy the normal operation of each flow node, and cause the overtime of the work task. In order to improve the cooperation efficiency inside an enterprise and improve the business service quality of the enterprise, how to establish an accurate and reliable flow defect prediction mechanism through data analysis, mine and find potential problems existing in the business flow execution process of the enterprise, position defect types and the severity of the problems, and are key technologies for improving the business management level of the enterprise and realizing the transformation of a digital strategy, and have strong application and practice values.
The existing flow defect identification technology and flow optimization method mainly aim at the problems existing in certain application scenes of certain specific industries and have no universality. Secondly, a large amount of prior knowledge is needed in the defect identification process, and the flow data is not effectively utilized for deep analysis. Finally, many service instances of the business process are difficult to determine the defects in the design and simulation test stages, and accurate defect prediction can be given only by comprehensively considering the dynamic change characteristics of the business process on the premise of collecting enough log data of the process execution process.
Disclosure of Invention
Aiming at the problems in the prior art, the invention provides a method for predicting the defects of the office business process. The method utilizes a machine learning technology to train the characteristics of the flow design and the execution, and obtains a display representation model of the office business flow defect analysis, namely a hierarchical decision tree of the business flow defect inspection. The prediction of flow defects can be realized by utilizing the hierarchical decision tree.
The technical scheme adopted by the invention is as follows:
aiming at the problems in the prior art, the invention provides a method for predicting the defects of the office business process. The method utilizes a machine learning technology to train the characteristics of the flow design and the execution, obtains a display representation model of the office business flow defect analysis, namely a hierarchical decision tree for business flow defect inspection, and can realize the prediction of the flow defects by utilizing the hierarchical decision tree.
The technical scheme adopted by the invention is as follows:
preprocessing service process log data, analyzing a bottleneck node problem, a decision node position problem and a cycle structure problem caused by process rollback in a process structure, and obtaining structural characteristics of office service process inspection by using a corresponding characteristic quantification method;
analyzing an information flow network formed among organizations or individuals participating in the process node tasks at the service process execution level, evaluating the reasonability of personnel allocation and cooperation of the node tasks, and obtaining the organization cooperation characteristics of office service process inspection;
thirdly, basic strategies and rules of resource performance in the business process execution process are formulated through regression analysis, the rationality of resource distribution is measured through the deviation of the real situation and the performance reference value, and the resource distribution characteristics of office business process inspection are obtained;
step four, a normal transaction processing mode in the business process is established by utilizing the probability model, and the deviation degree of the business process variation is quantized through similarity comparison with the normal mode, so that the mode abnormal characteristic of the office business process inspection is obtained;
collecting information such as the execution timeout times of the service flow used for more than one month on line, the node task execution timeout times, the reporting of flow problems and the like, giving a grade of the service flow by a panel according to the collected data, and making a judgment interval of the service flow defects under different service lines;
And step six, taking the structural features, the organization and cooperation features, the resource distribution features and the mode abnormal features of the office business process inspection as feature dimensions of process defect prediction. And taking the business process scores given by the panel as labeled data, taking the data combined with the characteristic value data as a training set, and training a business process defect inspection hierarchical decision tree to obtain a prediction model capable of evaluating the office business process. And finally, judging the severity of the process defect according to the judging intervals of the service process defects under different service lines.
The analysis method for the office business process inspection structure characteristics further comprises the following steps:
1) taking a time interval between task arrival time and task processing start time as service waiting time, counting execution frequency and average service waiting time of each node in a business process, confirming nodes with longer service waiting time frequently occurring in a plurality of time intervals by using a scanning counting method, regarding the nodes as bottleneck nodes in the business process, recording the number of the bottleneck nodes, the average waiting time of the bottleneck nodes and the service waiting time peak value of the bottleneck nodes in the business process, and incorporating the number, the average waiting time and the service waiting time peak value into training parameters.
2) All paths which are existed in the business process design and are started by a node to carry out task flow and can return to the node are used as a cycle structure in the business process. And counting the number of the loop structures in the business process, the trigger frequency when the loop structures are executed and the average path length of the loop structures, and incorporating the average path length into the training parameters.
3) And taking the decision nodes related to the flow trend of the business flow as decision nodes. The decision node controls the path of flow execution in the flow processing. By placing decision points with high decision efficiency at the initial stage of the process and deleting redundant decision points, the process structure can be simplified. And recording the number of unreasonable decision points in the position in the service process, the average path length from the decision point to the earliest decision point and the number of redundant decision points by positioning the earliest position of the decision node, and incorporating the number into training parameters.
The analysis method for the office business process checking organization collaboration characteristics further comprises the following steps:
1) extracting an information flow network formed by employee information transfer inside an enterprise from OA, Email, IM and other cooperative work systems;
2) according to a social role classification method in a social network, staff nodes in an information flow network are divided into contributors, influencers and coordinators;
3) Extracting an organization collaboration network formed by task participants of each node from workflows collected by office business processes;
4) judging whether people organizing key nodes in the collaborative network meet proper social roles or not, and further judging whether staff with proper roles exist in the same organization or not;
5) and counting the number of key nodes which do not conform to the social roles and the number of key nodes with replaceable roles, and incorporating the number into the training parameters.
The analysis method for the office business process checking resource distribution characteristics further comprises the following steps:
1) performing data preprocessing on the process logs to obtain data related to resource load and data related to process processing efficiency;
2) obtaining a linear function of the resource load and the process processing efficiency by using a linear regression statistical method;
3) calculating the least square between the predicted value and the actual value of the regression function, and judging the training accuracy of the regression function;
4) according to a linear function curve of the resource load and the process processing efficiency, the deviation degree of the actual node resource load and an ideal peak value (the process processing efficiency increases along with the increase of the resource load at the beginning, but decreases along with the increase of the resource load after reaching a certain degree) is calculated and is included in the training parameters.
The method for analyzing the abnormal mode characteristics of the office business process inspection further comprises the following steps:
1) extracting an office affair network formed by activities of all nodes from workflows collected by office business processes;
2) in the network, the routing relation of flow execution is described by adopting the conditional probability of the successive occurrence of the activities of two nodes;
3) because the execution probability of the current node activity is only influenced by the task of the previous node, a standardized model for office transaction processing can be calculated and represented by a Markov model;
4) by comparing with the standardized model, the nodes with abnormal activity execution states in the process can be found, the number of abnormal nodes, the average deviation degree of the abnormal nodes and the maximum value of the deviation degree of the abnormal nodes are counted, and the abnormal nodes are included in the training parameters.
Drawings
FIG. 1 is a flow chart of the embodiment
Fig. 2 is a flowchart of step S1
Fig. 3 is a flowchart of step S2
Fig. 4 is a flowchart of step S3
Fig. 5 is a flowchart of step S3
Detailed Description
In order to make the technical solutions and features of the present invention better understood by those skilled in the art, the present invention is described in further detail below with reference to the accompanying drawings.
FIG. 1 is a flowchart of a method for predicting defects in an office business process according to the present invention. Referring to fig. 1, an office business process defect prediction method based on a hierarchical decision tree according to an embodiment of the present invention specifically includes the following steps:
S1: preprocessing flow log data from a business process management system (BPM), analyzing a bottleneck node problem, a decision node position problem and a cycle structure problem caused by flow rollback in a flow structure, using a corresponding characteristic quantification method to obtain structural characteristics of office business process inspection, and bringing the structural characteristics into a training parameter set of a hierarchical decision tree.
In an embodiment of the present invention, as shown in fig. 2, the implementing step S1 further includes the following steps:
s11: defining service latency TwaitTask arrival time TarriveTask processing start time TstartWherein, Twait=Tstart-Tarrive. Setting a unit time interval/time window TwindowThe length is days, weeks and months, and the tolerance threshold value of the service waiting time is Tlimit. It is assumed that the effective recording time T of the flow log can be divided into w time windows/time intervals TwindowAnd for the jth time period (where 1 ≦ j ≦ w), the threshold T is exceededlimitTime-return value YjIs 1, otherwise 0, then YjDistribution P (Y)j1-p satisfies bernoulli's theorem, then the probability of k out-of-threshold events occurring or exceeding for consecutive w time segments satisfies the following equation:
utilizing scanning statistics to locate bottleneck nodes with task service waiting time frequently exceeding a preset threshold value in a plurality of time intervals, wherein the bottleneck nodes meet the following formula:
N←2∑i<kb(i, w, p) -1- (2k-1-2wp) b (k, w, p) formula 2
S12: and defining a business flow graph G (V, E), wherein V represents a task node, and E represents a task flow or an information flow between two task nodes. If present
p:{vi∈V|pathc=<vi,eij,ejk,...,emi,vi>},
Then p is a closed path in the business process graph representing a loop structure in the business process. This patent uses, but is not limited to, a dynamic programming algorithm to find all p in the business flowchart G.
S13: defining decision node related to flow trend of business flow as decision node C, and other nodes as common task node T, prefixes (T)i) Indicates a certain oneTask node TiAll previous paths. Mapping the most preferred location problem of a positioning decision node to a rule-based classification problem, wherein the selection at decision node c is classification, the set of classification attributes AT (c) is output by a task node preceding node c by a variable voutIs composed of, i.e.
Therefore, the rule-based decision tree construction method comprises the following steps:
finally, according to the rule-based decision tree (class, L)C) And finding out the most prior position node which is in accordance with the rule before the decision node c.
S2: on the aspect of business process execution, a centrality analysis method in social network structure feature analysis is utilized to perform node centrality measurement on organizations participating in process node tasks or an information flow network formed among individuals, further, the reasonability of personnel allocation and cooperation of the node tasks is evaluated, and organization cooperation features of office business process inspection are obtained;
In an embodiment of the present invention, as shown in fig. 3, the implementing step S2 further includes the following steps:
s21: extracting an information flow network formed by transmitting employee information inside an enterprise from OA, Email, IM and other cooperative work systems, wherein nodes in the network are employees, information exchange occurs between the two employees, if communication data are generated, an edge is constructed, and the weight of the edge is set according to a communication mode, communication times and communication frequency;
s22: each employee has a corresponding social role in the community, wherein there are three special social roles of key contributors, influencers and coordinators:
1) the key contributors are responsible for processing major flow tasks, and adopt a centrality measure;
2) the influencers have higher leadership capability in the group, and the centrality measurement of the feature vectors is adopted;
3) the coordinator is responsible for information interaction among groups and adopts intermediate centrality measurement;
s23: extracting an organization collaboration network formed by task participants of each node from workflows collected by office business processes;
s24: judging whether the person of the node meets a proper social role or not according to the connection characteristics of the participants of the organization cooperative network and the social roles of the participants in the information flow network, and further judging whether employees with proper roles exist in the same organization (unit, department and work group);
S25: and counting the number of key nodes which do not conform to the social roles and the number of key nodes with replaceable roles, and incorporating the number into the training parameters.
S3: and analyzing the functional relation between the resource load and the process processing efficiency by utilizing a mathematical statistical method of linear regression. Making basic strategies and rules of resource performance in the business process execution process by referring to the function curve, measuring the rationality of resource allocation through the deviation of the real situation and the performance reference value, and obtaining the resource allocation characteristics of the office business process inspection;
in an embodiment of the present invention, as shown in fig. 4, the implementing step S3 further includes the following steps:
s31: performing data preprocessing on the process logs to obtain data related to resource load and data related to process processing efficiency;
s32: obtaining a linear function of the resource load and the process processing efficiency by using a linear regression statistical method;
s33: calculating the least square between the predicted value and the actual value of the regression function, judging the training accuracy of the regression function, bringing the regression model into a process inspection model base after the accuracy reaches the standard (the standard threshold of the accuracy is manually set), and otherwise, turning the model into manual calibration (parameter adjustment and data re-acquisition);
S34: according to a linear function curve of the resource load and the process processing efficiency, the deviation degree of the actual node resource load and an ideal peak value (the process processing efficiency increases along with the increase of the resource load at the beginning, but decreases along with the increase of the resource load after reaching a certain degree) is calculated, the deviation degree is limited in a (0,1) interval by using a Sigmoid function, and the deviation degree is included in a training parameter.
S4: the method comprises the steps that a normal transaction processing mode in a business process is built by utilizing a Markov model, the deviation degree of business process variation is measured by comparing the distance between an executing process mode and the Markov chain probability distribution of the normal transaction processing mode, and the mode abnormal characteristic of office business process inspection is obtained;
in an embodiment of the present invention, as shown in fig. 5, the implementing step S4 further includes the following steps:
s41: extracting an office affair execution sequence formed by activities of all nodes from a workflow collected by an office business process;
s42: using markov chains to cluster the transaction execution sequences, each sequence should be assigned to a cluster with a higher transition probability. For sequence x ═ x1,x2,...,xL}, adding cluster ckThe probability of an associated Markov chain can be expressed as:
S43: to measure the distance between clusters on a Markov chain probability distribution, especially considering the state space distribution, S is setiIs ckOne state in the state space of a cluster, length | xrL sequence xrIs assigned to ckThe sequence of the cluster. Then ckMiddle SiThe marginal probability of (c) can be expressed as:
wherein the summation over r is based on all the symbols belonging to ckSequence of clusters, # Si(xr) For each sequence xrAppearance state SiThe number of times of (c) is counted.
Further, cluster ckAnd cluster clThe distance between can be measured by the following formula:
whereinIs a cluster ckThe transfer matrix of (2). By computing the cluster ckAnd cluster clThe distances between all the cluster pairs can obtain a K multiplied by K distance matrix D (c)k||cl),0≤D(ck||cl) Less than or equal to 1. The above is a method of measuring the degree of pattern anomaly by distance on a markov chain probability distribution.
S5: collecting information such as the number of times of execution of an online service process which uses more than one month, the number of times of node task execution overtime, process problem reporting and the like, giving a grade of the service process by a panel according to the collected data, and making a judgment interval of service process defects under different service lines;
s6: and taking the structural feature, the organization and cooperation feature, the resource distribution feature and the mode abnormity feature of the office business process inspection as the feature dimension of the process defect prediction. And taking the business process scores given by the panel as labeled data, selecting a part of the labeled data combined with the characteristic value data as a training set, and training a business process defect inspection hierarchical decision tree to obtain a prediction model capable of evaluating the office business process. And testing the trained business process defect inspection hierarchical decision tree by using the residual merged data as a test set, wherein the test indexes are recall rate, accuracy and F value commonly used for machine learning. And finally, judging the severity of the process defect according to the judgment intervals of the service process defects under different service lines by the output prediction result.
In summary, the invention provides an office business process defect prediction method based on a hierarchical decision tree, which comprehensively considers structural features, organization and cooperation features, resource distribution features and mode abnormality features of a business process during process inspection, and the process defect inspection hierarchical decision tree obtained by a computer through training can analyze unreasonable and errors in the business process from two levels of design and execution, thereby helping an enterprise to avoid economic loss caused by business process defects and improving the business operation efficiency of the enterprise. According to the hierarchy of the decision tree where the problem is located and the hierarchy of the leaf nodes, the bottleneck of enterprise business development can be positioned to a certain extent, and the enterprise can be helped to make a more scientific workflow service combination scheme and business development strategy.
Claims (10)
1. A method for predicting defects of office business processes comprises the following steps:
1) preprocessing the service process log data, analyzing the bottleneck node problem, the decision node position problem and the cycle structure problem caused by process rollback in the process structure, and obtaining the structural characteristics of office service process inspection by using a corresponding characteristic quantification method;
2) On the aspect of business process execution, analyzing an information flow network formed among organizations or individuals participating in the process node task, evaluating the reasonability of personnel allocation and cooperation of the node task, and obtaining the organization cooperation characteristics of office business process inspection;
3) basic strategies and rules of resource performance in the business process execution process are formulated, the rationality of resource distribution is measured through the deviation of the real situation and the performance reference value, and the resource distribution characteristics of office business process inspection are obtained;
4) a normal transaction processing mode in the business process is established by utilizing a probability model, and the deviation of business process variation is quantified through similarity comparison with a normal mode, so that the mode abnormal characteristic of office business process inspection is obtained;
5) collecting information such as the number of times of execution of an online service process which uses more than one month, the number of times of node task execution overtime, process problem reporting and the like, giving a grade of the service process by a panel according to the collected data, and making a judgment interval of service process defects under different service lines;
6) and taking the structural feature, the organization and cooperation feature, the resource distribution feature and the mode abnormity feature of the office business process inspection as the feature dimension of the process defect prediction. And taking the business process scores given by the panel as labeled data, taking the data combined with the characteristic value data as a training set, and training a business process defect inspection hierarchical decision tree to obtain a prediction model capable of evaluating the office business process. And finally, judging the severity of the process defect according to the judging intervals of the service process defects under different service lines.
2. The method according to claim 1, wherein the structural feature analysis method for office business process inspection further comprises:
taking a time interval between task arrival time and task processing start time as service waiting time, counting execution frequency and average service waiting time of each node in a business process, confirming nodes with longer service waiting time frequently occurring in a plurality of time intervals by using a scanning counting method, regarding the nodes as bottleneck nodes in the business process, recording the number of the bottleneck nodes, the average waiting time of the bottleneck nodes and the service waiting time peak value of the bottleneck nodes in the business process, and incorporating the number, the average waiting time and the service waiting time peak value into training parameters.
3. The method of claim 2, wherein the bottleneck node problem analysis method further comprises:
the service waiting time is defined as the waiting time consumed from the task arrival time to the task processing start time. After manually setting the length of the unit time interval/time window, the tolerance threshold of the service waiting time and the parameter values of the number of the observation time windows, the bottleneck nodes of which the task service waiting time frequently exceeds the preset threshold in a plurality of time intervals are positioned by using but not limited to a scanning statistical algorithm based on Bernoulli distribution.
4. The method of claim 1, wherein the structural feature analysis of the office business process survey further comprises:
all paths which are existed in the business process design and are started by a node to carry out task flow and can return to the node are used as a cycle structure in the business process. And searching for a loop structure existing in the flow by using but not limited to a dynamic programming algorithm, counting the number of the loop structures in the business flow, triggering frequency when the loop structures are executed and average path length of the loop structures, and incorporating the trigger frequency and the average path length into training parameters.
5. The method of claim 1, wherein the structural feature analysis of the office business process survey further comprises:
and taking the decision nodes related to the flow trend of the business flow as decision nodes. The decision node controls the path of flow execution in the flow processing. By placing decision points with high decision efficiency at the initial stage of the process and deleting redundant decision points, the process structure can be simplified. And recording the number of unreasonable decision points in the position in the service process, the average path length from the decision point to the earliest decision point and the number of redundant decision points by positioning the earliest position of the decision node, and incorporating the number into training parameters.
6. The method of claim 1, wherein the method for analyzing the organizational collaboration feature of the office business process examination further comprises:
1) extracting an information flow network formed by employee information transfer inside an enterprise from OA, Email, IM and other cooperative work systems;
2) according to a social role classification method in a social network, staff nodes in an information flow network are divided into contributors, influencers and coordinators;
3) extracting an organization collaboration network formed by task participants of each node from workflows collected by office business processes;
4) judging whether people organizing key nodes in the collaborative network meet proper social roles or not, and further judging whether staff with proper roles exist in the same organization or not;
5) and counting the number of key nodes which do not conform to the social roles and the number of key nodes with replaceable roles, and incorporating the number into the training parameters.
7. The method of claim 6, wherein the social role classification method further comprises:
1) the key contributors are responsible for processing major flow tasks, and adopt a centrality measure;
2) the influencers have higher leadership capability in the group, and the centrality measurement of the feature vectors is adopted;
3) The coordinator is responsible for information interaction among the groups and adopts intermediate centrality measurement.
8. The method of claim 1, wherein the method for analyzing the resource allocation characteristics of the office business process further comprises:
1) performing data preprocessing on the process logs to obtain data related to resource load and data related to process processing efficiency;
2) obtaining a linear function of the resource load and the process processing efficiency by using a linear regression statistical method;
3) calculating the least square between the predicted value and the actual value of the regression function, and judging the training accuracy of the regression function;
4) according to a linear function curve of the resource load and the process processing efficiency, the deviation degree of the actual node resource load and an ideal peak value (the process processing efficiency increases along with the increase of the resource load at the beginning, but decreases along with the increase of the resource load after reaching a certain degree) is calculated and is included in the training parameters.
9. The method of claim 1, wherein the method for analyzing the abnormal pattern feature of the office business process check further comprises:
1) extracting an office affair network formed by activities of all nodes from workflows collected by office business processes;
2) In the network, the routing relation of flow execution is described by adopting the conditional probability of the successive occurrence of the activities of two nodes;
3) because the execution probability of the current node activity is only influenced by the task of the previous node, a standardized model for office transaction processing can be calculated and represented by a Markov model;
4) by comparing with the standardized model, the nodes with abnormal activity execution states in the process can be found, the number of abnormal nodes, the average deviation degree of the abnormal nodes and the maximum value of the deviation degree of the abnormal nodes are counted, and the abnormal nodes are included in the training parameters.
10. The method of claim 9, wherein the method for calculating the mode state anomaly further comprises:
using markov chains to cluster the transaction execution sequences, each sequence should be assigned to a cluster with a higher transition probability. And calculating the deviation condition between the clusters obtained by clustering by using the distance on the probability distribution of the Markov chain, and simultaneously considering the condition of state space distribution, wherein a state transition matrix of the clusters needs to be added into the calculation formula.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010446235.9A CN111860936A (en) | 2020-05-25 | 2020-05-25 | Method for predicting defects of office business process |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010446235.9A CN111860936A (en) | 2020-05-25 | 2020-05-25 | Method for predicting defects of office business process |
Publications (1)
Publication Number | Publication Date |
---|---|
CN111860936A true CN111860936A (en) | 2020-10-30 |
Family
ID=72985218
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202010446235.9A Pending CN111860936A (en) | 2020-05-25 | 2020-05-25 | Method for predicting defects of office business process |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN111860936A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113780844A (en) * | 2021-09-14 | 2021-12-10 | 山东理工大学 | Cross-organization business process model mining and compliance checking method and system |
Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20120078678A1 (en) * | 2010-09-23 | 2012-03-29 | Infosys Technologies Limited | Method and system for estimation and analysis of operational parameters in workflow processes |
KR20130130400A (en) * | 2012-05-22 | 2013-12-02 | 김종수 | Method and system for the business standardization work |
CN103679384A (en) * | 2013-12-25 | 2014-03-26 | 武汉武船信息集成有限公司 | Method for workflow cooperative office work |
CN104714941A (en) * | 2013-12-12 | 2015-06-17 | 国际商业机器公司 | Method and system augmenting bussiness process execution using natural language processing |
KR20150094561A (en) * | 2015-07-24 | 2015-08-19 | 경기대학교 산학협력단 | Method and system of analysing workflow-supported social network based on closeness centrality |
US20150370467A1 (en) * | 2014-06-18 | 2015-12-24 | Alfresco Software, Inc. | Configurable and self-optimizing business process applications |
CN109753408A (en) * | 2018-12-11 | 2019-05-14 | 江阴逐日信息科技有限公司 | A kind of process predicting abnormality method based on machine learning |
CN110503211A (en) * | 2019-08-22 | 2019-11-26 | 贵州电网有限责任公司 | Failure prediction method based on machine learning |
CN111079997A (en) * | 2019-12-03 | 2020-04-28 | 北京仿真中心 | Modeling and collaborative optimization method |
-
2020
- 2020-05-25 CN CN202010446235.9A patent/CN111860936A/en active Pending
Patent Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20120078678A1 (en) * | 2010-09-23 | 2012-03-29 | Infosys Technologies Limited | Method and system for estimation and analysis of operational parameters in workflow processes |
KR20130130400A (en) * | 2012-05-22 | 2013-12-02 | 김종수 | Method and system for the business standardization work |
CN104714941A (en) * | 2013-12-12 | 2015-06-17 | 国际商业机器公司 | Method and system augmenting bussiness process execution using natural language processing |
CN103679384A (en) * | 2013-12-25 | 2014-03-26 | 武汉武船信息集成有限公司 | Method for workflow cooperative office work |
US20150370467A1 (en) * | 2014-06-18 | 2015-12-24 | Alfresco Software, Inc. | Configurable and self-optimizing business process applications |
KR20150094561A (en) * | 2015-07-24 | 2015-08-19 | 경기대학교 산학협력단 | Method and system of analysing workflow-supported social network based on closeness centrality |
CN109753408A (en) * | 2018-12-11 | 2019-05-14 | 江阴逐日信息科技有限公司 | A kind of process predicting abnormality method based on machine learning |
CN110503211A (en) * | 2019-08-22 | 2019-11-26 | 贵州电网有限责任公司 | Failure prediction method based on machine learning |
CN111079997A (en) * | 2019-12-03 | 2020-04-28 | 北京仿真中心 | Modeling and collaborative optimization method |
Non-Patent Citations (4)
Title |
---|
MALAK AL HATTAB等: "Simulating the dynamics of social agents and information flows in BIM-based design", AUTOMATION IN CONSTRUCTION, vol. 92, pages 1 - 22, XP085402200, DOI: 10.1016/j.autcon.2018.03.024 * |
刘冰川: "软件缺陷分析与管理系统的设计与实现", 中国优秀硕士学位论文全文数据库 信息科技辑, pages 138 - 366 * |
孔建寿, 张友良, 汪惠芬, 陈石灵: "协同开发环境中项目管理与工作流管理的集成", 中国机械工程, no. 13, pages 6 * |
魏懿;曹健;: "基于机器学习的流程异常预测方法", 计算机集成制造系统, no. 04, pages 864 - 872 * |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113780844A (en) * | 2021-09-14 | 2021-12-10 | 山东理工大学 | Cross-organization business process model mining and compliance checking method and system |
CN113780844B (en) * | 2021-09-14 | 2024-03-01 | 北京杰成合力科技有限公司 | Cross-organization business process model mining and compliance checking method and system |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Samoilenko et al. | Using Data Envelopment Analysis (DEA) for monitoring efficiency-based performance of productivity-driven organizations: Design and implementation of a decision support system | |
US8751436B2 (en) | Analyzing data quality | |
Biller et al. | Introduction to simulation input modeling | |
Ebnother et al. | Operational risk: A practitioner's view | |
CN117557078B (en) | Intelligent production plan management method and system based on time nodes | |
CN106713267A (en) | Network security assessment method and system | |
Nafreen et al. | Connecting software reliability growth models to software defect tracking | |
US20240185159A1 (en) | Systems And Methods For Identifying An Officer At Risk Of An Adverse Event | |
CN111860936A (en) | Method for predicting defects of office business process | |
CN117217634B (en) | Enterprise cooperation community discovery method based on complex network | |
CN113570118A (en) | Workshop scheduling and analyzing method based on scheduling rule | |
CN116611733A (en) | Project evaluation index management method based on operation and maintenance project | |
Murad et al. | Application Engineer Selection using Simple Additive Weighting Method Approach | |
CN115204501A (en) | Enterprise evaluation method and device, computer equipment and storage medium | |
CN113743994A (en) | Provider's season-busy prediction method, system, equipment and storage medium | |
Madlenak et al. | Reliability enhancement using optimization analysis | |
Zhang | Research of boosting algorithm machine learning in logistics enterprise financial risk prediction | |
CN117114911B (en) | Automatic statistical settlement method and system for financial statement | |
CN116342300B (en) | Method, device and equipment for analyzing characteristics of insurance claim settlement personnel | |
KR102669893B1 (en) | Automatic calculation system for profit and loss by IT project using smart work log | |
CN114611627A (en) | Enterprise classification method and system | |
CN118247038A (en) | Financial evaluation method for constructing enterprise financial credit figures based on neural network | |
Xu | Research on Strategic Decision Model of Human Resource Management based on Biological Neural Network | |
CN117333001A (en) | Work order flow defect identification method based on flow mining technology | |
Gui | Enterprise Accounting Risk Early Warning Model Based on Artificial Intelligence System Economics |
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 | ||
WD01 | Invention patent application deemed withdrawn after publication | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20201030 |