JP2006228019A - Method and system for supporting business process extraction, and server - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a support method for automatically extracting a business process. <P>SOLUTION: Message monitoring parts 102-1 to 102-n included in respective nodes 101-1 to 101-n monitor messages transmitted/received among the nodes 101-1 to 101-n and their transmitting/receiving states, a message information collection part 201 registers the monitored results in a message information registering database 202 as message information and process estimation parts 204-1 to 204-m estimate business processes on the basis of the registered message information and a process extraction part 203 extracts the estimated business processes and registers the extracted business processes in a process information database 206. The registered business processes are displayed on a process information display part 302, an operator 301 can edit the displayed business processes through an input part 303 and the edited business processes are registered in the process information database 206 by a process information correction part 303. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to a business process extraction support method, a business process extraction support system, and a server that extract a business process from an existing business system in which the business process is not clearly defined.

  Conventionally, in order to clearly define a business process, the results of work and related information with other work have been input and input by an operator to an input device connected to the device that builds the business process. Business processes are built based on information.

  FIG. 14 is a diagram illustrating a configuration example of a conventional business process extraction system.

  As shown in FIG. 14, this configuration example includes an input device such as a terminal, and an input unit 403 for an operator 405 to input process information which is information related to business performance and other business, A process information display unit 402 having a function of displaying the process information input by the operator 405 via the input unit 403 in a graph, and a process information output for outputting the process information input by the worker 405 to the process information database 401 Part 404.

  In the business process extraction system configured as described above, when process information is input via the input unit 403 by the worker 405, the process information input via the input unit 403 is displayed by the process information display unit 402. Displayed in a graph. When the input of process information by the worker 405 is completed, the graph displayed on the process information display unit 402 is converted into process definition information by the process information output unit 404 and registered in the process information database 401.

In this way, by inputting process information by the worker 405, a business process is extracted and constructed (see, for example, Patent Document 1).
JP-A-9-259188

  However, in the business process extraction support system as described above, an operator must input process information such as work results directly to an input device such as a terminal, and business processes cannot be automatically extracted. There is a problem.

  Also, for messages that are sent and received between nodes that make up the business process, the worker must investigate the behavior and relevance to estimate the business process. There is a problem that it takes time and effort.

  The present invention has been made in view of the problems of the conventional technology as described above, and includes a business process extraction support method, a business process extraction support system, and a server that can automatically extract a business process. The purpose is to provide.

In order to achieve the above object, the present invention provides:
A business process extraction support method for extracting a flow of the work process in performing a series of work having a plurality of work processes,
A process for monitoring messages transmitted and received between the work steps and the transmission / reception status thereof; and
Processing to collect the monitoring results as message information;
Processing to estimate a business process based on the collected message information.

  Further, it is characterized in that the similarity of messages included in the message information is analyzed, categorization is performed for each similar message, and a business process is estimated for each divided category.

Further, in performing a series of work having a plurality of work steps, a business process extraction support system for extracting the flow of the work steps,
Message monitoring means for monitoring messages transmitted and received between the work steps and their transmission / reception status;
And a process analysis unit that analyzes a business process based on the message monitored by the message monitoring unit and a transmission / reception state thereof.

The process analysis means includes
Message information collecting means for collecting the message monitored by the message monitoring means and its transmission / reception status as message information, and registering the message information in a first database;
Process estimation means for estimating a business process based on message information registered in the first database;
It has a process extracting means for extracting the business process estimated by the process estimating means and registering it in a second database.

  The business process registered in the second database further includes a process input / output means for enabling editing from the outside.

The process input / output means includes
Process information display means for displaying a business process registered in the second database by the process extraction means;
Process information correction means for reflecting the contents of correction made from the outside and registering the corrected business process in the second database is provided.

  The process estimation means analyzes the similarity of messages included in the message information, performs category classification for each similar message, and estimates the business process for each divided category.

In addition, a server that analyzes a process based on messages transmitted and received between work processes and their transmission / reception status,
Message information collecting means for collecting the message and its transmission / reception status as message information, and registering the message information in a first database;
Process estimation means for estimating a business process based on message information registered in the first database;
A process extracting unit that extracts the business process estimated by the process estimating unit and registers the extracted business process in a second database.

  The process estimation means analyzes the similarity of messages included in the message information, performs category classification for each similar message, and estimates the business process for each divided category.

  In the present invention configured as described above, in performing a series of operations having a plurality of operation processes, messages transmitted and received and the transmission / reception status between the operation processes are monitored, and the monitored messages and transmission / reception thereof are monitored. A business process is estimated based on the situation.

  Thereby, the worker can automatically extract a business process without taking time and effort to input process information of work results directly to a terminal or the like. Further, by editing the automatically extracted business process from the outside, the automatically extracted business process and the worker's knowledge are merged to create an accurate business process definition.

  As described above, in the present invention, in performing a series of work having a plurality of work processes, messages transmitted / received between the work processes and their transmission / reception status are monitored, and the monitored messages and their transmission / reception statuses are monitored. Since the business process is estimated based on the configuration, the business process can be automatically extracted.

  Embodiments of the present invention will be described below with reference to the drawings.

  FIG. 1 is a diagram showing an embodiment of a business process extraction support system of the present invention.

  As shown in FIG. 1, the present embodiment includes a business system 100 from which a business process is extracted, a process analysis server 200 that analyzes the business process of the business system 100, a terminal, and the like. The process input / output unit 300 that displays the business process candidates analyzed in FIG. 4 to the worker 301 and enables editing from the outside of the worker 301 or the like.

  The business system 100 includes nodes 101-1 to 101-n, which are a plurality of work processes. The nodes 101-1 to 101-n are transmitted and received between the nodes 101-1 to 101-n. Message monitoring units 102-1 to 102-n that monitor messages and their transmission / reception status and transmit the monitoring results to the process analysis server 200 as message information, respectively.

  Further, the process analysis server 200 collects message information from the message monitoring units 102-1 to 102-n, and registers the collected message information in the message information database 202 which is a first database; Refers to message information registered in the message information database 202, extracts business process candidates, and is activated by the process extraction unit 203, which is registered in the process information database 206, which is the second database, and the process extraction unit 203. The received message information is statistically analyzed by a predetermined logic, and the process estimation units 204-1 to 204-m and the process estimation units 204-1 to 204-m for estimating business process candidates are newly registered, and , Already registered processes Tough 204-1 through 204-m and a Process estimator registering selector 205. notifying to the process extraction unit 203 selects the one to be used when the process extracted from the.

  The process input / output unit 300 includes a process information display unit 302 that displays the process information extracted by the process extraction unit 203 and registered in the process information database 206, and the process information is corrected by the worker 301. The process information correction unit 304 for re-registering the processed process information in the process information database 206, the external input from the worker 301 or the like is received, the process extraction unit 203, the process estimation unit registration selection unit 205, and the process information display unit 302 are received. And an input unit 303 that activates the process information correction unit 304.

  Below, the business process extraction support method in the business process extraction support system comprised as mentioned above is demonstrated with reference to a flowchart.

  First, the monitoring operation of messages transmitted and received between the nodes 101-1 to 101-n will be described.

  FIG. 2 is a flowchart for explaining a message monitoring operation in the business process extraction support system shown in FIG.

  First, message transmission / reception is performed between the nodes 101-1 to 101-n (step S1). Then, message transmission / reception is detected in the message monitoring units 102-1 to 102-n provided in the nodes 101-1 to 101-n, and the transmitted / received message information is acquired (step S2).

  The message information acquired in the message monitoring units 102-1 to 102-n is transmitted from the message monitoring units 102-1 to 102-n (step S3) and received by the message information collecting unit 201 (step S4). .

  The message information received by the message information collection unit 201 is stored in the message information database 202 (step S5).

  Next, an operation for extracting business process candidates from the message information described above will be described.

  FIG. 3 is a flowchart for explaining a business process candidate extraction operation in the business process extraction support system shown in FIG.

  When the process extraction unit 203 is activated by the operator 301 operating the input unit 303 (step S11), first, message information stored in the message information database 202 is acquired by the process extraction unit 203 (step S12). .

  Next, the process extraction unit 203 determines whether there is a node that has not yet been selected as the initial node among the nodes 101-1 to 101-n included in the business system 100 (step S13). If it is determined that there is no node that has not been selected, the business process candidate extraction operation ends (step S14).

  On the other hand, if it is determined in step S13 that there is a node that has not been selected as the initial node, one node is selected as the initial node from among the nodes that have not been selected as the initial node (step S15).

  Next, the process extraction unit 203 selects a target node corresponding to the initial node selected in step S15 from among the nodes 101-1 to 101-n included in the business system 100. First, it is determined whether there is a node that is not selected as a target node corresponding to the initial node among all the nodes 101-1 to 101-n (step S16), and the target node corresponding to the initial node is determined. When it is determined that there is no node that is not selected, the process returns to step S13.

  On the other hand, when it is determined that there is a node that is not selected as the target node corresponding to the initial node, one node among the nodes that are not selected as the target node corresponding to the initial node corresponds to the initial node. Selected as a target node (step S17).

  Next, from the message information acquired in step S12, a message transmitted from the initial node to the target node is extracted, and the extracted message is set as a reference message (step S18).

  Then, process extraction is started. It is determined whether or not any process estimation unit 204-1 to 204-m registered in advance by the process estimation unit registration / selection unit 205 has not been selected as the process extraction logic (step S19). If it is determined that there is something that is not selected as the extraction logic, one process estimation unit is selected from the process estimation units 204-1 to 204-m by the process estimation unit registration selection unit 205 (step S20). . Then, the process estimation logic of the selected process estimation unit is executed (step S21), and the process returns to step S19. Below, the execution method of the process extraction logic performed by step S21 is demonstrated in detail.

  FIG. 4 is a flowchart for explaining a process extraction logic execution method in the business process extraction support system shown in FIG.

  First, the similarity of each message is analyzed for the reference message transmitted from the initial node to the target node, and the reference message is categorized for each similar message (step S31). Next, a message group transmitted from the target node to another node is extracted (step S32), and the extracted message group is defined as a message group A.

  Next, a message group transmitted to another node immediately after the reference message is received at the target node is extracted from the message group A (step S33), and the extracted message group is defined as a message group B.

  Further, it is determined whether there is any node that has not been selected as the initial node or the target node among the nodes 101-1 to 101-n included in the business system 100 (step S34). If it is determined that there is a node that is neither selected as the initial node nor the target node, one of the nodes is selected as the node C (step S35). When the initial node or target node is changed, even the node that was selected as the initial node or target node before the change is selected as the initial node after change or the target node after change. Nodes that do not exist are selected as node C.

  Next, the ratio of messages transmitted to the node C included in the message group A and the ratio of messages transmitted to the node C included in the message group B are compared (step S36) and included in the message group B. It is determined whether the ratio of messages transmitted to the node C is significantly larger than the ratio of messages transmitted to the node C included in the message group A (step S37). Here, whether or not it is significantly large is determined by a statistical test, and “the ratio of messages transmitted to the node C included in the message group B is the message transmitted to the node C included in the message group A”. Is the null hypothesis, and “the ratio of messages transmitted to the node C included in the message group B is larger than the ratio of messages transmitted to the node C included in the message group A”. If the test is significant, the ratio of messages transmitted to the node C included in the message group B is greater than the ratio of messages transmitted to the node C included in the message group A. Is also judged to be significantly larger. Note that the significance level is appropriately determined according to the size of the message group. When it is determined that the ratio of messages transmitted to the node C included in the message group B is not significantly larger than the ratio of messages transmitted to the node C included in the message group A, the process of step S34 Return to. When it is determined that the ratio of messages transmitted to the node C included in the message group B is significantly larger than the ratio of messages transmitted to the node C included in the message group A, in step S31. It is determined whether there is a category that has not yet been selected as a process extraction target category among the classified categories (step S38). If it is determined that there is no category not selected as the process extraction target category, the process returns to step S34. If it is determined that there is an unselected category in the process extraction target category, one category is selected as the process extraction target category from the unselected categories (step S39). A message included in the category is set as a reference message (step S40).

  Next, a message group transmitted to another node immediately after the reference message is received at the target node is extracted from the message group A (step S41), and the extracted message group is set as a message group C.

  Thereafter, the ratio of messages transmitted to the node C included in the message group A and the ratio of messages transmitted to the node C included in the message group C are compared (step S42), and the nodes included in the message group C are compared. It is determined whether the ratio of messages transmitted to C is significantly greater than the ratio of messages transmitted to node C included in message group A (step S43). When it is determined that the ratio of messages transmitted to the node C included in the message group C is not significantly larger than the ratio of messages transmitted to the node C included in the message group A, the process of step S38 Return to. If it is determined that the ratio of messages transmitted to the node C included in the message group C is significantly larger than the ratio of messages transmitted to the node C included in the message group A, the selected category A business process of “initial node → target node → node C” is extracted.

  Next, the target node is set as the initial node, node C is set as the target node, the message group transmitted to the node C included in the message group C is set as the reference message, and the same process extraction is executed (step S44). ).

  Then, “initial node → target node → node C → process extracted in step S44” is added as an extraction result (step S45).

  This “initial node → target node → node C → process extracted in step S44” is obtained by connecting message paths presumed to constitute a process in order of transmission time.

  On the other hand, if it is determined in step S34 that neither the initial node nor the target node is selected, that is, all nodes have already been selected, the selected process extraction logic ends. Furthermore, if it is determined in step S19 that there is no item that has not been selected as the process extraction logic, that is, all the process estimation units 204-1 to 204-m registered in advance have already been selected. Then, the operation of the process extraction unit 203 ends, and the extraction result is registered in the process information database 206 (step S22).

  The process extraction logic execution process performed in step S21 will be described below with another example.

  FIG. 5 is a flowchart for explaining another execution method of the process extraction logic in the business process extraction support system shown in FIG.

  First, for the reference message transmitted from the initial node to the target node, the similarity of each message is analyzed, and the reference message is categorized for each similar message (step S51). Next, a message group transmitted from the target node to another node is extracted (step S52), and the extracted message group is set as a message group D.

  Next, after the reference message is received at the target node and before the message is returned from the target node to the initial node, the message group transmitted to the other node is extracted from the message group D (step S53). The extracted message group is defined as a message group E.

  Further, it is determined whether there is a node that is not selected as an initial node or a target node among the nodes 101-1 to 101-n included in the business system 100 (step S54). If it is determined that there is a node that is neither selected as the initial node nor the target node, one of the nodes is selected as the node C (step S55). When the initial node or target node is changed, even the node that was selected as the initial node or target node before the change is selected as the initial node after change or the target node after change. Nodes that do not exist are selected as node C.

  Next, the ratio of messages transmitted to the node C included in the message group D and the ratio of messages transmitted to the node C included in the message group E are compared (step S56), and the nodes included in the message group E are compared. It is determined whether or not the ratio of messages transmitted to C is significantly greater than the ratio of messages transmitted to node C included in message group D (step S57). When it is determined that the ratio of messages transmitted to the node C included in the message group E is not significantly larger than the ratio of messages transmitted to the node C included in the message group D, the process of step S54 Return to. If it is determined that the ratio of messages transmitted to the node C included in the message group E is significantly larger than the ratio of messages transmitted to the node C included in the message group D, in step S51. It is determined whether there is a category that has not yet been selected as a process extraction target category among the classified categories (step S58). If it is determined that there is no category not selected as the process extraction target category, the process returns to step S54. If it is determined that there is an unselected category in the process extraction target category, one category is selected as the process extraction target category from the unselected categories (step S59). A message included in the category is set as a reference message (step S60).

  Next, a message transmitted to another node is extracted after the reference message is received at the target node until the message is returned from the target node to the initial node (step S61). Let the group be a message group F.

  Thereafter, the ratio of messages transmitted to the node C included in the message group D and the ratio of messages transmitted to the node C included in the message group F are compared (step S62), and the node C included in the message group F is compared. It is determined whether or not the ratio of messages transmitted to is significantly larger than the ratio of messages transmitted to node C included in the message group D (step S63). If it is determined that the ratio of messages transmitted to the node C included in the message group F is not significantly larger than the ratio of messages transmitted to the node C included in the message group D, the process of step S58 Return to. When it is determined that the ratio of messages transmitted to the node C included in the message group F is significantly larger than the ratio of messages transmitted to the node C included in the message group D, the selected category A business process of “initial node → target node → node C” is extracted.

  Next, the target node is set as the initial node, node C is set as the target node, the message group transmitted to the node C included in the message group F is set as the reference message, and the same process extraction is executed (step S64). ).

  Then, “initial node → target node → node C → process extracted in step S64” is added as an extraction result (step S65).

  This “initial node → target node → node C → process extracted in step S64” is obtained by connecting message paths presumed to constitute a process in order of transmission time.

  On the other hand, if it is determined in step S54 that neither the initial node nor the target node is selected, that is, all nodes have already been selected, the selected process extraction logic ends. Furthermore, if it is determined in step S19 that there is no item that has not been selected as the process extraction logic, that is, all the process estimation units 204-1 to 204-m registered in advance have already been selected. Then, the operation of the process extraction unit 203 ends, and the extraction result is registered in the process information database 206 (step S22).

  Hereinafter, a process for displaying and correcting business process candidates registered in the process information database 206 will be described.

  FIG. 6 is a flowchart for explaining the process of displaying and correcting business process candidates registered in the process information database 206 shown in FIG.

  First, processing for displaying business process candidates will be described.

  When a business process candidate display instruction is input by the worker 301 via the input unit 303 (step S71), the process information registered in the process information database 206 is acquired by the process information display unit 302 (step S72). ). Then, the process information display unit 302 displays the acquired process information (step S73).

  Next, processing for correcting a business process candidate will be described.

  When a business process candidate correction command is input by the operator 301 via the input unit 303 (step S81), the process information displayed on the process information display unit 302 is corrected by the process information correction unit 304 (step S81). S82). Then, the corrected process information is stored in the process information database 206 by the process information correction unit 304 (step S83).

  While the example of the process extraction method in the configuration shown in FIG. 1 has been described above, the process extraction method in the configuration shown in FIG. 1 will be described below with a more specific example. In the following description, the node 101-1 shown in FIG. 1 is the node a, the node 101-2 is the node b, the node 101-3 is the node c, the node 101-4 is the node d, and the node 101-5 is the node 101-5. Assume that the node e and message information transmitted and received between the nodes are already stored in the message information database 202.

  FIG. 7 is a diagram showing an example of message information stored in the message information database 202 shown in FIG.

  As illustrated in FIG. 7, the transmission source, transmission destination, transmission time, message content, and the like are stored for messages transmitted and received between the nodes 101-1 to 101-5.

  FIG. 8 is a diagram illustrating a state in which the message information illustrated in FIG. 7 is classified and tabulated by the transmission source and the transmission destination.

  As shown in FIG. 8, the number and ratio of messages transmitted and received for each transmission source and transmission destination are classified and tabulated.

  FIG. 9 is a flowchart for explaining a specific example of the business process candidate extraction operation in the business process extraction support system shown in FIG.

  When the process extraction unit 203 is activated by the operator 301 operating the input unit 303 (step S101), first, message information stored in the message information database 202 is acquired by the process extraction unit 203 (step S102). .

  Next, the process extraction unit 203 selects the node a as an initial node (step S103), and the node b is selected as a target node corresponding to the node a that is the initial node (step S104).

  Next, the message transmitted from node a to node b is extracted from the message information acquired in step S102, and the extracted message is set as a reference message (step S105).

  Then, process extraction is started. Here, it is assumed that only one process estimation unit 204-1 is registered by the process estimation unit registration selection unit 205.

  The process estimation unit 204-1 registered by the process estimation unit registration selection unit 205 is selected by the process estimation unit registration selection unit 205 as a process extraction logic (step S106). Then, the process estimation logic of the selected process estimation unit 204-1 is executed (step S107), and the result of process extraction by the executed process estimation logic is registered in the process information database 206 (step S108).

  Hereinafter, a method for executing the process extraction logic performed in step S107 will be described in detail.

  FIG. 10 is a flowchart for explaining an example of a specific execution method of the process extraction logic in the business process extraction support system shown in FIG.

  First, the similarity of each message is analyzed for the reference message transmitted from the initial node to the target node, and the reference message is divided into category 1 and category 2 for each similar message (step S121). Next, a message group transmitted from the target node b to another node is extracted (step S122), and the extracted message group is defined as a message group G. In FIG. 8, the column where the transmission source is b is the number of messages in the message group G and the ratio thereof.

  Next, a message group transmitted to another node immediately after the reference message is received at the node b is extracted from the message group G (step S123), and the extracted message group is defined as a message group H.

  Furthermore, among the nodes 101-1 to 101-n included in the business system 100, the node c that is not selected as the initial node or the target node is selected as the node C (step S124).

  Next, the ratio of messages transmitted to the node c included in the message group G and the ratio of messages transmitted to the node c included in the message group H are compared, and whether there is a significant difference between the two ratios. Is determined (step S125).

  FIG. 11 is a diagram showing a breakdown of messages transmitted to other nodes immediately after the message from node a is received at node b. The numerical values shown in FIG. 11 are the number of messages in the message group H and the ratio thereof.

  As shown in FIG. 8, the number of messages transmitted to the node c included in the message group G is 4000, and the ratio is 26%. On the other hand, as shown in FIG. 11, the number of messages transmitted to the node c included in the message group H is 3500, and the ratio is 35%.

  Therefore, it is determined that the ratio of messages transmitted to the node c included in the message group H is significantly larger than the ratio of messages transmitted to the node c included in the message group G.

  Thereafter, when category 1 is first selected as a process extraction target category from category 1 and category 2 divided in step S121 (step S126), a message included in category 1 is set as a reference message ( In step S127), a message group transmitted to another node immediately after the reference message is received by the node b is extracted from the message group G, and the extracted message group is defined as a message group I (step S128).

  After that, the ratio of messages transmitted to the node c included in the message group G and the ratio of messages transmitted to the node c included in the message group I are compared, and whether there is a significant difference between the two ratios. Judgment is made (step S129).

  FIG. 12 is a diagram showing a breakdown for each category of messages transmitted to other nodes immediately after the message from node a is received at node b. In FIG. 12, the category 1 column indicates the number of messages in the message group I and the ratio thereof.

  As shown in FIG. 8, the number of messages transmitted to the node c included in the message group G is 4000, and the ratio is 26%. On the other hand, as shown in FIG. 12, the number of messages transmitted to the node c included in the message group I is 3000, and the ratio is 60%.

  Accordingly, it is determined that the ratio of messages transmitted to the node c included in the message group I is significantly larger than the ratio of messages transmitted to the node c included in the message group G. A business process of “node b → node c” is extracted.

  Next, the node b is set as the initial node, the node c is set as the target node, the message group transmitted to the node c included in the message group I is set as the reference message, and the same process extraction is executed ( Step S130). However, as shown in FIG. 8, since no message is transmitted from the node c to any node, a business process from the node c to another node is not extracted.

  In this way, business process candidates “node a → node b → node c” for category 1 are determined and added to the business process extraction result (step S131).

  Next, category 2 is selected as a process extraction target category (step S132), and a business process is extracted using the same logic as in category 1. A message included in category 2 is set as a reference message (step S133). A message group transmitted to another node immediately after the reference message is received at node b is extracted from message group G. A message group J is set (step S134). In FIG. 12, the category 2 column indicates the number of messages in the message group J and the ratio thereof.

  Thereafter, the ratio of messages transmitted to the node c included in the message group G and the ratio of messages transmitted to the node c included in the message group J are compared, and whether there is a significant difference between the two ratios. Determination is made (step S135).

  As shown in FIG. 8, the number of messages transmitted to the node c included in the message group G is 4000, and the ratio is 26%. On the other hand, as shown in FIG. 12, the number of messages transmitted to the node c included in the message group J is 500, and the ratio is 10%.

  Accordingly, it is determined that the ratio of messages transmitted to the node c included in the message group J is not significantly larger than the ratio of messages transmitted to the node c included in the message group G. It is not recognized as a process candidate.

  Next, the same logic as when node d is selected as node C (step S136) and node c is selected as node C is executed. The ratio of messages transmitted to node d included in message group G is compared with the ratio of messages transmitted to node d included in message group H to determine whether the two ratios are significantly different. (Step S137).

  As shown in FIG. 8, the number of messages transmitted to the node d included in the message group G is 6000, and the ratio is 40%. On the other hand, as shown in FIG. 11, the number of messages transmitted to the node d included in the message group H is 5000, and the ratio is 50%.

  Therefore, it is determined that the ratio of messages transmitted to the node d included in the message group H is significantly larger than the ratio of messages transmitted to the node d included in the message group G.

  Then, when category 1 is first selected as a process extraction target category from category 1 and category 2 divided in step S121 (step S138), a message included in category 1 is set as a reference message ( In step S139), a message group transmitted to another node immediately after the reference message is received by the node b is extracted from the message group G, and the extracted message group is defined as a message group I (step S140).

  Thereafter, the ratio of messages transmitted to the node d included in the message group G and the ratio of messages transmitted to the node d included in the message group I are compared, and whether there is a significant difference between the two ratios. Judgment is made (step S141).

  As shown in FIG. 8, the number of messages transmitted to the node d included in the message group G is 6000, and the ratio is 40%. On the other hand, as shown in FIG. 12, the number of messages transmitted to the node d included in the message group I is 1000, and the ratio is 20%.

  Therefore, it is determined that the ratio of messages transmitted to the node d included in the message group I is not significantly larger than the ratio of messages transmitted to the node d included in the message group G. It is not recognized as a process candidate.

  Next, category 2 is selected as a process extraction target category (step S142), and a business process is extracted using the same logic as in category 1. A message included in category 2 is set as a reference message (step S143). A message group transmitted to another node immediately after the reference message is received at node b is extracted from message group G. The message group is J (step S144).

  Thereafter, the ratio of messages transmitted to the node d included in the message group G and the ratio of messages transmitted to the node d included in the message group J are compared, and whether there is a significant difference between the two ratios. Determination is made (step S145).

  As shown in FIG. 8, the number of messages transmitted to the node d included in the message group G is 6000, and the ratio is 40%. On the other hand, as shown in FIG. 12, the number of messages transmitted to the node d included in the message group J is 4000, and the ratio is 80%.

  Accordingly, it is determined that the ratio of messages transmitted to the node d included in the message group J is significantly larger than the ratio of messages transmitted to the node d included in the message group G. A business process of “node b → node d” is extracted.

  Next, the node b is set as the initial node, the node d is set as the target node, the message group transmitted to the node d included in the message group J is set as the reference message, and the same process extraction is executed ( Step S146).

  FIG. 13 is a diagram showing a breakdown of messages transmitted to other nodes immediately after the message from node b is received at node d. The message group shown in FIG.

  The ratio of messages sent to other nodes included in the message group G and the ratio of messages sent to other nodes included in the message group K are compared to determine whether the two ratios are significantly different. (Step S147).

  As shown in FIGS. 8 and 13, it is determined that the ratio of messages transmitted to other nodes included in the message group K is significantly larger than the ratio of messages transmitted to other nodes included in the message group G. It can be seen that the transmission destination is only node e which is 4000 (20%) in FIG. 8 and 3500 (35%) in FIG.

  Accordingly, it is determined that the ratio of messages transmitted to the node e included in the message group K is significantly larger than the ratio of messages transmitted to the node e included in the message group G. A business process of “node b → node d → node e” is extracted.

  Next, the node d is set as an initial node, the node e is set as a target node, the message group transmitted to the node e included in the message group K is set as a reference message, and the same process extraction is executed ( Step S148). However, as shown in FIG. 8, since the message is hardly transmitted from the node e to the other nodes, the statistical significance is not recognized and the business process candidate is not extracted. Then, the business process estimation logic ends.

  In this manner, business process candidates “node a → node b → node d → node e” for category 2 are determined and added to the business process extraction result (step S149).

  As described above, business process candidates “node a → node b → node c” and “node a → node b → node d → node e” are registered in the process information database 206 as extraction results (step S150). .

  Thereafter, the process extraction process is performed while sequentially changing the initial node and the target node.

It is a figure which shows one Embodiment of the business process extraction assistance system of this invention. It is a flowchart for demonstrating the message monitoring operation | movement in the business process extraction assistance system shown in FIG. It is a flowchart for demonstrating the business process candidate extraction operation | movement in the business process extraction assistance system shown in FIG. It is a flowchart for demonstrating the execution method of the process extraction logic in the business process extraction assistance system shown in FIG. It is a flowchart for demonstrating the other execution method of the process extraction logic in the business process extraction assistance system shown in FIG. It is a flowchart for demonstrating the process of a display and correction of the business process candidate registered into the process information database shown in FIG. It is a figure which shows an example of the message information stored in the message information database shown in FIG. FIG. 8 is a diagram illustrating a state in which the message information illustrated in FIG. 7 is classified and tabulated by a transmission source and a transmission destination. It is a flowchart for demonstrating the specific example of the business process candidate extraction operation | movement in the business process extraction assistance system shown in FIG. It is a flowchart for demonstrating an example of the concrete execution method of the process extraction logic in the business process extraction assistance system shown in FIG. It is a figure which shows the breakdown of the message transmitted to the other node immediately after the message from the node a was received in the node b. It is a figure which shows the breakdown for every category of the message transmitted to the other node immediately after the message from the node a was received in the node b. It is a figure which shows the breakdown of the message transmitted to the other node immediately after the message from the node b was received in the node d. It is a figure which shows the structural example of the conventional business process extraction system.

Explanation of symbols

DESCRIPTION OF SYMBOLS 100 Business system 101-1 to 101-n Node 102-1 to 102-n Message monitoring part 200 Process analysis server 201 Message information collection part 202 Message information database 203 Process extraction part 204-1 to 204-m Process estimation part 205 Process Estimating means registration / selection unit 206 Process information database 300 Process input / output unit 301 Worker 302 Process information display unit 303 Input unit 304 Process information correction unit

Claims (9)

A business process extraction support method for extracting a flow of the work process in performing a series of work having a plurality of work processes,
A process for monitoring messages transmitted and received between the work steps and the transmission / reception status thereof; and
Processing to collect the monitoring results as message information;
And a process for estimating a business process based on the collected message information. The business process extraction support method according to claim 1,
A business process extraction support method characterized by analyzing the similarity of messages included in the message information, categorizing each similar message, and estimating a business process for each divided category. A business process extraction support system for extracting a flow of the work process when performing a series of work having a plurality of work processes,
Message monitoring means for monitoring messages transmitted and received between the work steps and their transmission / reception status;
A business process extraction support system comprising process analysis means for analyzing a business process based on a message monitored by the message monitoring means and a transmission / reception state thereof. In the business process extraction support system according to claim 3,
The process analysis means includes
Message information collecting means for collecting the message monitored by the message monitoring means and its transmission / reception status as message information, and registering the message information in a first database;
Process estimation means for estimating a business process based on message information registered in the first database;
A business process extraction support system, comprising: a process extraction unit that extracts a business process estimated by the process estimation unit and registers it in a second database. In the business process extraction support system according to claim 4,
A business process extraction support system comprising a process input / output means for enabling editing of a business process registered in the second database from the outside. In the business process extraction support system according to claim 5,
The process input / output means includes
Process information display means for displaying a business process registered in the second database by the process extraction means;
A business process extraction support system, comprising: process information correction means for reflecting a correction content made from outside and registering the corrected business process in the second database. In the business process extraction support system according to any one of claims 4 to 6,
The process estimation means analyzes the similarity of messages included in the message information, performs category classification for each similar message, and estimates the business process for each divided category. Support system. A server that analyzes a process based on messages sent and received between work processes and their sending and receiving status,
Message information collecting means for collecting the message and its transmission / reception status as message information, and registering the message information in a first database;
Process estimation means for estimating a business process based on message information registered in the first database;
A server having process extraction means for extracting the business process estimated by the process estimation means and registering it in a second database; The server according to claim 8, wherein
The server, wherein the process estimation unit analyzes the similarity of messages included in the message information, performs category classification for each similar message, and estimates the business process for each divided category.

Images