JP2002279119A - Business analyzing system and business analyzing method for workflow management system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To obtain analysis information required for improving the business for a workflow management system by precisely and objectively analyzing the operation business for the workflow management system. SOLUTION: A characteristic vector extracting part 201 obtains operational history data on a job and task specified by an analysis request from a person in charge of analysis from an analysis data base 115, and one or more characteristic vector components and the respective values are calculated. In a retrieval part 202, the similarlity between the pattern of the extracted one or more characteristic vector component values and the pattern of one or plural characteristic vector component values related to each of business samples registered in a business pattern data base 117 is determined by a pattern matching method, thereby retrieving the business sample whose pattern of distribution of characteristic vector component values is similar to the analysis object business from the business pattern data base 117. The analysis result information on the business pattern is presented to a person in charge of analysis.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a business analysis system and a business analysis method for a workflow management system, and more particularly, to the characteristics of the business operated by the workflow management system using operation history data of the workflow management system. The present invention relates to a business analysis system and a business analysis method for analyzing.

[0002]

2. Description of the Related Art In recent years, in offices and business offices, the introduction of groupware for exchanging and sharing information between individuals and groups using a computer network has been promoted. By using this groupware, the productivity of groups and individuals can be significantly increased.

A typical groupware is a workflow management system. This workflow management system supports the execution of work by automating the flow of a series of work necessary for business processing and the transfer of information necessary for the work, and is realized as client-server type software. I have. In this case, the server manages the entire task (job) to be executed, the individual tasks (tasks) required for the task processing, and the information (document) to be transferred. Is performed in a distributed manner by clients (user terminals).

[0004] The server manages the assignment of jobs and tasks to each user. A task to be processed by the client is notified from the server to each user terminal, and a process specified by the notified task information is executed on the user terminal. In this case, on the server, data indicating the processing status of each job and each task is accumulated and managed as operation history data.

[0005]

A business feature analysis system using operation history data of a workflow management system is proposed in Japanese Patent Application No. 2000-333726, which is a patent application filed by the present applicant. As a result, the characteristics of the actual operation work can be checked with the operation history data.

However, in this work analysis system, it is practically difficult to accurately and objectively analyze the operation work of the workflow management system because the final analysis result often depends on the skills of the analyst. Met. Hereinafter, this point will be specifically described.

That is, generally, the procedure of business analysis is divided into the following two. (1) Acquisition of data: Acquisition of data necessary for analysis from operation history (2) Analysis: Analysis of acquired data to find out business features, problems, etc. The point of analysis is the analysis of (2) At the stage, it depends on "how to grasp the characteristics from the processing status of the work and find out the problem hidden in the work".

In the data acquisition of (1), the type of data to be acquired is almost fixed and can be acquired automatically, so that no special skill is required.

However, in the analysis of (2), the processing
From the operation history data on tasks, their processing time, etc., it is necessary to characterize the work actually operated in the workflow management system and find problems.
It depends heavily on the experience and skills of analysts.

In other words, in the workflow management system, the contents of tasks and the like to be processed by each processing person are updated every moment in accordance with changes in the progress of task processing by each processing person and addition / change of work. May be done. This is because there are tasks that can be processed only after completion of a certain task process, and tasks that are additionally defined as tasks are added or changed. For example, in the case of a workflow that consists of three tasks of “Create quote” → “Approve” → “Quote issuance”, for the person in charge of the “Approve” task, “ Since it is necessary to refer to the “estimate”, the “approval” task can be processed only after the “person in charge of the“ estimate creation ”task creates the“ estimate ”. Similarly, the “estimate issue” task can be processed after the “estimate” has been approved by the “approval” task. In consideration of such points, from the operation history data on the person in charge of processing, the task, its processing time, etc.,
It is necessary to find out the characterization and problems of the work actually operated by the workflow management system. Therefore, conventionally, the following problem occurs.

It is difficult for an inexperienced analyst to characterize the work and find problems. Even an experienced analyst can find features and problems from various data. It is difficult, and the analysis takes some man-hours.

SUMMARY OF THE INVENTION The present invention has been made in view of the above circumstances, and enables a workflow management system to accurately and objectively analyze the operation of a workflow management system, such as reviewing a business definition in the workflow management system. It is an object of the present invention to provide a business analysis system and a business analysis method for a workflow management system capable of obtaining analysis information necessary for business improvement defined and operated in the above.

[0013]

SUMMARY OF THE INVENTION In order to solve the above-mentioned problems, the present invention provides an operation history data of a workflow management system which supports a flow of a series of work necessary for business processing and information transfer necessary for the work. Is a business analysis system that analyzes the characteristics of the business operated by the workflow management system, using, for each of a plurality of business samples having different business characteristics, information on the analysis result of the business characteristics and business characteristics. A business pattern database in which values relating to a plurality of feature vector components predefined for analysis are registered, and a feature vector extracting means for extracting values relating to each of the plurality of feature vector components from operation history data relating to a business to be analyzed And a feature vector component extracted by the feature vector extracting means. And a search means for searching the business pattern database for a business sample similar to the business to be analyzed based on the above, and acquiring information on an analysis result regarding the similar business sample as the business characteristic of the business to be analyzed. And

In this business analysis system, for each of a plurality of business samples having different business characteristics, information on the analysis result of the business characteristics and a value regarding each of a plurality of feature vector components defined in advance for analyzing the business characteristics. A business pattern database in which is registered is prepared, and business characteristics relating to the analysis target business can be obtained by searching the business pattern database for business samples having a plurality of patterns of feature vector component value similarities. . In this case, first, a plurality of feature vector component values are extracted from the operation history data relating to the analysis target business, and a business sample similar to the analysis target business is searched from the business pattern database based on the extracted feature vector component values. Is done. As a result, the information of the analysis result registered in advance for the similar work sample is acquired as the work characteristic of the work to be analyzed, so that the operation work of the workflow management system can be accurately and objectively analyzed. In particular, in the search processing, a plurality of predefined feature vector component value patterns are used in order to analyze the business feature, so that the processing status of the business process can be diversified from various angles by a pattern matching method. Evaluate and find similar patterns.

Further, as the plurality of feature vector components, at least when each task or a task constituting the task arrives at a user terminal of the workflow management system, the processing of the task or task starts at the user terminal. Processing start time required until
Alternatively, it is preferable to define a component relating to the processing time required from the start of the work or the processing of the work to the completion thereof. Accordingly, for example, when the ratio of the processing start time is large, 1) there is a problem in the system design, the task or the method of notifying the work (eg, not aware that the task has arrived), 2) the flow If there is a problem with the design method and the assignment load of the user (there is a heavy load on other work, the hand is not turned, etc.), it is possible to recognize the work characteristics, and if the proportion of processing time is large, 1) Since it is possible to recognize the business characteristics of a problem with the user assignment (the task level is too high for the user skill, etc.), important information for reviewing the business definition in the workflow management system, etc. Can be obtained.

Also, the work constituting the business, the person in charge of processing,
By designating at least one of the analysis target periods to further narrow down and specify the range of the business to be analyzed, for example, it is possible to break down the entire job to individual work inside the job and to provide a feature vector. Can be extracted, which is suitable for finding potential problems.

Further, new analysis result information obtained by causing the analyst to add / modify information of the analysis result corresponding to the similar task sample obtained by the search means,
Using the feature vector component value of the analysis target business extracted by the feature vector extracting means, the analysis target business is additionally registered in the business pattern database as a new business sample. It is possible to easily realize additional registration of a business sample in a business pattern database while efficiently using the above data and feature vector component values extracted for search.

Further, a graph or a table showing the business characteristics of each business sample is registered in the business pattern database, and a graph or table corresponding to a business sample similar to the business to be analyzed at the time of searching from the business pattern database. Is obtained from the business pattern database and displayed and output, whereby the characteristics of the business to be analyzed can be visually presented.

[0019]

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

FIG. 1 shows the configuration of a task analysis system according to an embodiment of the present invention. This task analysis system is for analyzing features relating to tasks such as jobs and tasks operated by the workflow management system. Since this business analysis system analyzes business using the operation history data of the workflow management system, it can basically be executed on a computer system independent of the operation system of the workflow management system. In FIG. 1, for the sake of simplicity, an example is shown in which a task analysis system is implemented on a computer network system in which a workflow management system is operated.

The workflow management system uses a computer network to automate the flow of a series of tasks necessary for the job processing for each job and the transfer of information such as documents necessary for the task, and to support the execution of the job. And is realized by a workflow management system 111 which is server software executed on the server computer 11 and client software respectively executed on a plurality of user terminals 12 connected to the server computer 11 via the LAN 10. Is done.

The workflow management system 111 provides task information indicating the contents of the work to each of the processing staff on the plurality of user terminals 12 according to the work progress procedure defined as the workflow, and is necessary for each task processing. It controls various work screens and program execution, provides documents, and manages the processing status of jobs, task processing, and the like.

The management of jobs, tasks, documents, persons in charge of processing, and the correspondence between them by the workflow management system 111 is performed using information stored in the workflow database 112. As a result, the plurality of documents and the persons in charge are managed in association with each other, and the connection between each task which is a component of the workflow and the document necessary for the task processing is managed. In this case, the jobs constituting the flow are realized as objects of the object-oriented model, and the relationship between the job and the task is defined by the dependency between the objects.

The workflow management system 111
Has a function of sequentially collecting the processing status of a job and a task from the user terminal 12 and storing it in the workflow database 112 as operation history data and managing it. The operation history data includes, for example, as shown in FIG. 2, history data, object data relating to a user (a person in charge of processing), a task, and a job. The history data includes a completed object name (job name) and a user ID of a user who has processed the object, and a date / time when the object is completed.
This is data stored in chronological order together with time information and the like.
The object data relating to the user stores, for each user of the workflow management system 111, the object name and task name notified to the user in chronological order along with date / time information such as the notification date and time, the processing start date and time, and the completion date and time. Data. Similarly, the object data relating to a task stores, for each executed task, the task name and the user ID of the task processed in chronological order along with date / time information such as notification date / time, processing start date / time, and completion date / time. The object data relating to the job includes, for each executed job, the job name and the user ID of the executed job in time series with date / time information such as a notification date / time, a processing start date / time, and a completion date / time. This is the stored data. By using the operation history data, for example, which task is notified to which user, and when the processing is started and completed, it is possible to identify a processing status such as a job, a task, and a person in charge.

The entity of information such as documents is stored in the repository file 113 shown in FIG. 1, and documents to be transferred to each person in charge according to the work progress are read out from the repository file 113 and correspond to the documents. Provided to the person in charge.

The data acquisition tool 114 acquires operation history data relating to the job or task to be analyzed from the operation history data accumulated in the workflow database 112 as analysis data, and uses this as the analysis data.
5 is a program to be stored. The content of the data acquisition by the data acquisition tool 114 can be defined by a setting file defining the data. In this setting file, analysis item information indicating data items to be acquired from the operation history data, acquisition period information indicating from what time to when data should be acquired, and the like are defined. Since the above-mentioned operation history data is often deleted in order from the oldest one in the operation course, it is necessary to sequentially acquire necessary data by the data acquisition tool 114.

The automatic analysis tool 116 is a program for automatically analyzing the business characteristics of the business to be analyzed in response to the analysis request from the terminal 13 of the person in charge of analysis.
It has a function of automatically extracting a task feature related to a task to be analyzed from the operation history data stored in the analysis data database 115 and searching for a task pattern similar to the task pattern from the task pattern database 117. Here, the principle of the business analysis method using the automatic analysis tool 116 will be described.

As described above, the tasks defined and operated by the workflow management system 111 have various features,
There is a problem. The analyst will find problems and improvements by grasping these characteristics. At this time, how to find the characteristics of each task is the point of accurate task analysis, and this can lead to improvement of the task definition in the workflow management system 111. Therefore, in the present embodiment, the features of the business are classified into several types (patterns) by using the pattern recognition technology, and the analysis results (business characteristics, problems, etc.) of the business processing are preliminarily obtained for each pattern. A method of registering in the business pattern database 117 and automatically recognizing a pattern similar to the business to be analyzed from the patterns is used.

For pattern recognition, a feature vector is used. The feature vector refers to data that well represents the feature of each task, and is composed of a plurality of feature vector components defined in advance for evaluating the feature of the task processing. If this feature vector selection method is appropriate, it can be considered that feature vectors of tasks having similar features are similar to each other, and feature vectors of tasks having different features are different from each other.

First, in the business pattern database 117, for each of a plurality of business samples having different business characteristics, a plurality of business samples defined in advance for evaluating the information on the analysis result of the business characteristics and the progress of the business process are evaluated. A value relating to each feature vector component is registered. The automatic analysis tool 116 automatically extracts the values of the plurality of feature vector components from the operation history data of the business to be analyzed, and stores the values in the business pattern database 117 using a known pattern recognition technology (pattern matching). From similar business sample (business pattern)
Is searched. As a result, analysis results (business characteristics, problems, and the like) registered in advance for similar business samples can be presented to the person in charge of analysis as the business characteristic analysis results for the business to be analyzed.

Next, referring to FIG. 3, the automatic analysis tool 1
The sixteenth functional configuration will be described.

As shown in FIG. 3, the functions of the automatic analysis tool 116 are the feature vector extraction unit 201 and the search unit 2
02. The feature vector extraction unit 201 converts operation history data relating to a job, a task, and the like designated by an analysis request from an analysis staff into the analysis data database 11.
5 (analysis data acquisition), and a feature vector is extracted by calculating a value of each of the plurality of feature vector components from the acquired data. The value of each of the extracted plurality of feature vector components is sent to the search unit 202.

The retrieving unit 202 determines the similarity between the pattern of the plurality of extracted feature vector component values and the pattern of the plurality of feature vector component values for each of the business samples registered in the business pattern database 117 by using a pattern matching method. A business sample whose distribution pattern of the characteristic vector component value is similar to the business to be analyzed is determined by the business pattern database 11.
7 is searched. Then, analysis result information and the like on the business pattern are presented to the person in charge of analysis.

As described above, similar business patterns are stored in the business pattern database 1 using the automatic analysis tool 116.
By automatically recognizing from step 17, it is possible to reduce man-hours at the time of work analysis and to perform work analysis independent of the skill of the person in charge.

Here, an example of the feature vector component value will be described with reference to FIG.

FIG. 4 shows the relationship between the definition of each feature vector component used in task analysis and the pattern of feature vector component values for two specific business samples (business patterns 1 and 2). . Here, each feature vector component value assumes a value of 0 to 100% (normalization processing is performed if necessary).

As the feature vector component, a processing start time and a processing time are defined as main components of the feature vector. Here, the “processing start time” is defined as a period from the time when each task or the task constituting the task arrives at the user terminal 12 of the workflow management system 111 to the time when the task or task is started at the user terminal 12. For example, as shown in FIG. 5, the time from the arrival of a task (work) to the actual start of the task is the “processing start time”. In other words, the elapsed time from when the task becomes processable to when the task processing is started is the “processing start time”. The “processing time” indicates the time required from the start of the processing of the job or the work to the completion thereof, and as shown in FIG. 5, the time from the start of the task to the completion of the task. The actual working time is the “processing time”. In the operation history data, the task attachment date and time, task start date and time, and task completion date and time are managed for each task.
The “processing start time” and “processing time” can be checked for each task. Hereinafter, the processing start time and the processing time are used in this sense.

<Feature vector component # 1 (processing start time,
(Processing time ratio)> Feature vector component # 1 indicates the ratio between the processing start time and the processing time, and is expressed, for example, as a ratio of the processing start time to the entire time from arrival of the task to completion. "98%" of the business pattern 1 has a processing start time ratio of 9 to the total time from arrival of the task to completion.
8% means that the actual processing time is 2%. The feature vector component # 1 can be obtained as an average value of all tasks related to a certain task, as well as a task unit.

By using the feature vector component # 1, it is possible to study the following business problems.

A. When the ratio of the processing start time is large ・ There is a problem in the system design and task notification method (not aware that a task has arrived, etc.) ・ There is a problem in the flow design method and user assignment load (other The hand is not turned because the load on work is high) b. When the ratio of the processing time is large-There is a problem in user assignment (the task level is too advanced for the user skill, etc.) <Feature vector component # 2 (user ratio of the number of processes (top 1))>< Feature Vector Component # 3 (User Ratio of Number of Processes (Top 3))> These feature vector components # 2 and # 3 are used as indices for confirming whether or not users who have actually processed the tasks have no bias. That is, the feature vector component # 2 indicates the ratio of the number of tasks occupied by the user with the largest number of tasks processed out of the total number of tasks processed by all users in a certain period, and The vector component # 3 indicates the ratio of the total number of tasks processed by the top three users with the largest number of tasks processed to the total number of tasks processed by all users in a certain period. Feature vector components # 2, #
By using No. 3, it is possible to confirm whether there is a bias in processing for a specific user, and whether the bias in the number of processes does not adversely affect the processing start time and processing time (feature vector components # 4 to #). 7).

In particular, in the feature vector component # 2, it is possible to confirm a large deviation in the number of processes, and in the feature vector component # 3, it is possible to confirm whether or not the processes are equally allocated to the user. In addition, by determining the average of the number of processes per user, it is possible to confirm whether the processes are evenly applied to the user.

<Feature vector component # 4 (user ratio of total processing start time (top one))><Feature vector component # 5 (user ratio of total processing start time (top three))> These feature vector components # 4 and # 5 are used as indices for confirming whether or not there is a bias among specific users when viewing the entire time up to the processing start time. That is, feature vector component # 4 indicates the ratio of the total processing start time of the user having the largest total processing start time to the total of the processing start times for all tasks, and feature vector component # 5 indicates , The ratio of the total value of the total processing start times for the top three users having the highest total processing start times to the total of the processing start times for all tasks. These feature vector components # 4, # 5
Is particularly effective when it is recognized that the ratio of the processing start time is high in the feature vector component # 1 and there is a problem with the processing start time. That is, when the user ratio of the total processing start time is high (there is a bias), it can be analyzed that there is no problem in the task, but there is a problem in the processing of a specific user. Therefore, the problem can be considered focusing on the distribution of tasks to specific users. In addition, the processing load on other tasks can be considered together.

Particularly, in the feature vector component # 4, a large deviation in the processing start time can be confirmed, and the feature vector component # 5
Then, the uniformity of each user can be confirmed. Also,
In addition, if the processing start time is evenly taken by calculating the average of the processing start time per user, the problem of the task notification method by the workflow management system 111 and the problem of the fundamental work distribution. And so on.

By using the feature vector components # 4 and # 5 as described above, it becomes possible to find out the cause of the problem with respect to the task that is recognized as having a problem in the processing progress in the feature vector component # 1.

<Feature vector component # 6 (user ratio of total processing time (top one))><Feature vector component # 7 (user ratio of total processing time (top three))> These feature vector components # 6 and # 7 is used as an index for checking whether or not a specific user is biased when the entire time up to the processing time is viewed. That is, the feature vector component # 6 indicates the ratio of the total processing time of the user having the largest total processing time to the total processing time of all tasks, and the feature vector component # 7 indicates It shows the ratio of the total value of the total processing time for the top three users with the highest total processing time in the total processing time for tasks. The values of these feature vector components # 6 and # 7 are, for example, feature vector component # 1, and are particularly effective when it is recognized that the processing time is high and there is a problem in the processing time. That is, when the user ratio of the total processing time is high (there is a bias),
The problem is not with the entire task, but with a particular user. Therefore, the problem can be considered focusing on the distribution of tasks to specific users. At the same time, the skill of the user can be considered.

In particular, in the feature vector component # 6, a large deviation in the processing time can be confirmed, and in the feature vector component # 7, the uniformity of each user can be confirmed. In addition, by averaging the processing time per user,
If the processing time is equally long, it is possible to focus on the fundamental user assignment problem.

<Feature vector component # 8 (process start time user ratio exceeding a certain period)><Feature vector component # 9 (process start time number ratio exceeding a certain period)> These feature vector components # 8 and # 9 Is an index for setting a certain period of time, and confirming the ratio of users who took a long time to start processing and the ratio of the number of processes after that period. For example, if it takes more than 5 days to start the process, and there is a serious problem, the fixed period may be set to 5 days. In the feature vector component # 8, it is possible to confirm whether or not a specific user is taking a long time to start processing, and in the feature vector component # 9, when a specific user is seen, It can be checked whether it takes time to start processing.

For example, feature vector components # 1 and # 4
If it is found that it takes a long time to start the processing and the ratio of the processing start time of a specific user is high, it may be necessary to check the following details in more detail.

It takes a long time to start the process constantly for a certain period of time or more (it takes 10 days every time). (The rest is all within one day.) Although the meaning of the above is significantly different, it is difficult to detect only one abnormal value by summing up the total processing start time with data. However, by confirming the ratio of the number of processing start times that exceeded the fixed period of the feature vector component # 9, whether the period exceeded the fixed period constantly or the ratio of exceeding the fixed period was low, but the number of days exceeded was large. Can be distinguished from each other.

<Feature vector component # 10 (user ratio of processing time exceeding a certain period)><Feature vector component # 11 (proportion of number of processing times exceeding a certain period)> These feature vector components # 10 and # 11 are present. This is an index for setting a certain period, and confirming the ratio of the users who took the processing time and the ratio of the number of processes for the period or more.
The processing start time of the feature vector components # 8 and # 9 is replaced with the processing time. These feature vector components # 1
Due to 0 and # 11, ・ Processing is constantly taking longer than a certain period of time (it takes 10 days each time) ・ Only once, after a certain period of time, the processing is started only once. And then went on a business trip)
Analysis becomes possible.

The above-described feature vector components are shown as examples, and the feature vector components to be used can be added or changed as appropriate. It goes without saying that not only a task but also a similar feature vector component may be defined for a job.

FIG. 6 shows the business pattern database 11
7 shows an example of the data structure.

As shown in FIG. 6, in the business pattern database 117, several typical businesses classified by analyzing the operation business of the workflow management system 111 in advance using the above-described feature vector are described. For the sample (business pattern), the values of the characteristic vector components # 1 to # 11 and the business pattern ID are registered. Further, for each business pattern, one or more graphs or tables that most clearly show the business characteristics of the business pattern, comment texts on feature points, comment texts on problems, comment texts on improvements, and the like are registered. . Each comment sentence is analysis result information obtained by an analyst with sufficient skills analyzing the values of the feature vector components # 1 to # 11. The comment sentence regarding the feature point describes, for example, the business feature described in the above-described feature vector component value, and the comment sentence regarding the problem indicates at what point in the target job what kind of problem is. The comment text on the improvement proposal describes a business definition improvement proposal or the like for improving business efficiency. For business patterns for which further analysis is deemed necessary, information such as guidance information indicating what is more effective if re-analyzed is registered as a problem or comment text for improvement proposals. By doing so, more accurate analysis results can be obtained.

Next, referring to the flowchart of FIG.
The procedure of the business analysis / recognition processing executed by the automatic analysis tool 116 will be described.

First, a list display screen or the like for designating an analysis item to the analysis staff is provided from the automatic analysis tool 116 to the analysis staff terminal 13, and the analysis staff selection operation performed on the list display screen is performed. The analysis items (jobs, tasks, etc.) are determined accordingly (step S10).
1). In this case, by specifying the narrowing-down conditions on the list display screen, it is possible to break down the entire job to internal tasks and specify the analysis items. In addition, a user name, a period to be analyzed, and the like, as well as a job, a task, and the like can be specified by the narrowing-down condition. As a result, the operation task of the workflow management system 111 can be analyzed from various viewpoints using any combination of the entire task, individual tasks, processing charge types, analysis periods, and the like.

Next, the data corresponding to the analysis item specified in step S101 is stored in the analysis data database 115.
From the operation history data (step S10).
2). Then, feature vector component values required for the analysis are calculated from the obtained operation history data of the analysis target operation (step S103). Basically, component values are calculated for all the feature vector components defined by the automatic analysis seal 116. However, a feature vector component for which a component value should be calculated can be selected as necessary. . Thereafter, based on the calculated feature vector component values, a search for a similar business pattern from the business pattern database 117 is performed (step S104). In this business pattern search, a business pattern having a distribution pattern of feature vector component values most similar to a distribution pattern of feature vector component values of a business to be analyzed is searched for by a pattern recognition technique. The pattern recognition technique uses Bayes's discrimination theory, which is a general technique, or a statistical pattern recognition technique such as a neural network. In pattern recognition, it is preferable to perform learning as needed to improve the recognition rate. Specifically, the learning uses a technique such as "perform normalization processing on vector component values", "perform weighting (learning) on each vector component".

When a similar business pattern is not registered in the business pattern database 117 and the analyst determines that sufficient similarity is not found in the similar business pattern presented to the analyst as a search result. Can add / modify the business pattern in the business pattern database 117 at that time according to the judgment of the analyst.

That is, when the analysis person requests the business pattern to be corrected (the business pattern is corrected in step S104), a process of additionally registering the analysis target business as a new business pattern in the business pattern database 117 is performed. In this case, the business analysis input / edit screen is provided from the automatic analysis tool 116 to the analyst terminal 13, and the value of the feature vector relating to the business to be analyzed and the business analysis obtained as a search result are displayed on the screen. Information (business features, problems, improvement proposals, etc.) is displayed. Then, the analyst analyzes the business to be analyzed, and the business analysis information (business characteristics, business characteristics,
Problems, improvement proposals, etc.) are added or corrected (step S105). At the time of analysis, the analysis work is supported by displaying the feature vector component values related to the analysis target business in graphs or tables and displaying the feature vector component values of the business patterns obtained as search results. can do. Then, along with the new business pattern ID and the added / modified business analysis information, the feature vector component value relating to the analysis target business is additionally registered in the business pattern database 117 (step S106). In addition, when registering, from the graph or table displayed on the screen at the time of analysis,
1 that most clearly shows the business features of the business pattern to be registered
By allowing the analyst to select the above graph or table, it can be used for business pattern ID, business analysis information,
It can be registered in the business pattern database 117 together with the feature vector component value (step S107).

On the other hand, when a business pattern similar to the business to be analyzed is registered in the business pattern database 117 and the search result is used as an analysis result for the business to be analyzed (the business pattern is not corrected in step S104), The registration information (business pattern ID, business analysis information, feature vector component value) on the searched business pattern is acquired from the business pattern database 117, and is displayed on the screen of the analysis staff terminal 13 (step S108). This makes it possible to provide task analysis information such as a problem registered in advance for a similar task pattern as an analysis result of the task to be analyzed. Also,
When the analysis staff requests the display of the characteristic graph table (there is a display of the characteristic graph table), the characteristic graph table accurately representing the business pattern is acquired from the business pattern database 117, and is analyzed. Is displayed on the user terminal 13. Thereby, the feature of the business pattern can be visually recognized. At this time, the feature graph
A graph table not registered as a table can also be generated and displayed from a feature vector extracted from a task to be analyzed or a feature vector of a similar task pattern.

Next, the business pattern and the characteristic graph table will be described with reference to FIGS.

The feature graph table is a graph or table for visualizing and displaying business patterns, and is used for analyzing business characteristics in more detail.

For example, for the business pattern 1 described in relation to the characteristic vector component value in FIG. 4, the graph A (the ratio of the total processing time for each user) in FIG. 8 is registered in the business pattern database 117 as a characteristic graph table. By doing so, it is possible to simply present the feature of the business pattern 1 in which the ratio of the processing time is high, there is a problem in the processing time, and the user ratio of the total processing time is high (biased).

That is, in the business pattern 1, the characteristic vector in FIG. 4 shows that the “user ratio of the total processing time (top three users)” is high, and the processing is biased to a certain number of users. In graph A, the situation of the remaining users and
It is possible to visually grasp which user has a high processing load. Furthermore, by displaying the relationship between the number of cases in which the processing time exceeds a certain period and the user in a three-dimensional graph, it is possible to confirm whether there is no problem in the processing time with a high load.

For example, for the business pattern 2 described in relation to the characteristic vector component value in FIG. 4, the business pattern database 1 uses the graph B (the number of processing start time excesses for each user) in FIG. 9 as a characteristic graph table.
17, the number of cases where the processing start time has been exceeded can be confirmed for each user together with the number of days exceeded.

That is, in the business pattern 2, the feature vector of FIG. 4 shows that “the processing start time user ratio exceeding a certain period” is high, and that the processing start time exceeds a certain period for most users. I understand. In the graph B, the relationship between the user and the number of days of the processing start time can be visually grasped by the three-dimensional display, and it is possible to confirm which user has taken a long time to start the processing. Further, by displaying a graph of the ratio of the number of processes for each user, it is possible to confirm whether there is any problem in the relationship with the load.

As described above, the feature graph and the feature table assist in analyzing the features and problems of the tasks that cannot be completely expressed by the task analysis information in more detail and deeper. Therefore, it is preferable to register a graph and a table that represent the most characteristic and assist in detailed and deep analysis in association with the task analysis information.

As described above, in the analysis system of the present embodiment, for each of a plurality of business samples having different business characteristics, information relating to the analysis result of the business characteristics and a plurality of predefined business process characteristics are analyzed. A business pattern database 117 in which values relating to each of the feature vector components are registered is prepared, and a business sample having a similar pattern of a plurality of feature vector component values is searched from the business pattern database 117 using a pattern recognition processing technique. Accordingly, it is possible to easily acquire accurate business features related to the business to be analyzed.

The data acquisition tool 114 and the analysis tool 116 can be used by incorporating them into the workflow management system 111 as one of their functions. Basically, if there is only the operation history of the workflow management system 111, it can be analyzed. Therefore, it can be realized as a system independent of the workflow management system 111. Further, since the analysis tool 116, which is a main element of the analysis system, is a computer program for causing a computer to execute the processing, a computer-readable storage medium is stored in the analysis tool 116, and a normal computer is stored through the storage medium. , The same effect as in the present embodiment can be obtained. Further, a mechanism for automatically transmitting the operation history data acquired from the workflow management system 111 by the data acquisition tool 114 to an analysis system at a remote place through an electronic mail or the like can be used.

It should be noted that the present invention is not limited to the above-described embodiment, and can be variously modified in an implementation stage without departing from the gist of the invention. Further, the embodiments include inventions at various stages, and various inventions can be extracted by appropriately combining a plurality of disclosed constituent elements. For example, even if some components are deleted from all the components shown in the embodiment, the problem described in the column of the problem to be solved by the invention can be solved, and the effects described in the column of the effect of the invention can be solved. If you get
A configuration from which this configuration requirement is deleted can be extracted as an invention.

[0070]

As described above, according to the present invention,
It is possible to accurately and objectively analyze the workflow management system operation work, and to obtain analysis information necessary for improving the work defined and operated by the work flow management system, such as reviewing the work definition in the work flow management system. It becomes possible.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a configuration of a business analysis system according to an embodiment of the present invention and a workflow management system using the business analysis system.

FIG. 2 is a view showing an example of operation history data managed by the workflow management system of FIG. 1;

FIG. 3 is a block diagram showing a functional configuration of an automatic analysis tool included in the business analysis system of FIG. 1;

FIG. 4 is a view for explaining a feature vector used in the task analysis system of FIG. 1;

FIG. 5 is a view for explaining a relationship between a processing start time and a processing time defined by the feature vector in FIG. 4;

FIG. 6 is a view showing an example of a data structure of a business pattern database used in the business analysis system of FIG. 1;

FIG. 7 is an exemplary flowchart showing the procedure of a task analysis / recognition process executed by the task analysis system of FIG. 1;

FIG. 8 is a diagram showing an example of a feature graph used in the business analysis system of FIG. 1;

FIG. 9 is a view showing another example of a feature graph used in the task analysis system of FIG. 1;

FIG. 10 is a diagram showing an example of a feature table used in the task analysis system of FIG. 1;

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 11 ... Server computer 12 ... User terminal 13 ... Analytical person terminal 111 ... Workflow management system 112 ... Workflow database 113 ... Repository file 114 ... Data acquisition tool 115 ... Analysis data database 116 ... Automatic analysis tool 117 ... Business pattern database 201 ... Features Vector extraction unit 202: search unit

Claims (14)

[Claims] 1. A feature of a job operated by the workflow management system by using a flow of a series of operations required for the job processing and operation history data of the workflow management system for supporting transfer of information necessary for the job. A business analysis system for analyzing a business characteristic, for each of a plurality of business samples having different business characteristics, information on an analysis result of the business characteristic and a value for each of a plurality of feature vector components defined in advance for analyzing the business characteristic. A business pattern database in which are registered, a feature vector extraction unit that extracts values relating to each of the plurality of feature vector components from operation history data regarding the analysis target business, and a feature vector component value extracted by the feature vector extraction unit. Business sample similar to the business to be analyzed based on From the business pattern database,
A business analysis system comprising: a search unit configured to acquire information on an analysis result regarding the similar business sample as business characteristics of the business to be analyzed. 2. The method according to claim 1, wherein the search unit is configured to determine a distribution pattern of the plurality of feature vector component values extracted by the feature vector extracting unit and a plurality of feature vector component values for each of the business samples registered in the business pattern database. The method according to claim 1, wherein a similarity between the plurality of feature vector component values and a distribution pattern of the plurality of feature vector components is similar to the analysis target task. 1. The business analysis system according to 1. 3. The plurality of feature vector components include, at least, when a task or a task constituting the task arrives at a user terminal of the workflow management system, processing of the task or task starts at the user terminal. 2. The method according to claim 1, further comprising a component related to a processing start time required until the processing is completed and / or a processing time required from completion of the processing of the task or operation to completion of the processing. Business analysis system. 4. A feature vector extracting means, further comprising: a designating means for narrowing down and specifying a range of a task to be analyzed by designating at least one of a task constituting a task, a person in charge of processing, and a period to be analyzed. 2. The business analysis system according to claim 1, wherein the extracting unit extracts the feature vector component value from operation history data relating to a range of the business to be analyzed narrowed down and specified by the specification unit. 5. A new analysis result information obtained by adding / correcting an analysis result information corresponding to the similar work sample obtained by the search means to an analyst, and the feature vector extracting means 2. The system according to claim 1, further comprising: means for additionally registering the analysis target business as a new business sample in the business pattern database using the extracted feature vector component values of the analysis target business. Business analysis system. 6. The business pattern database has registered therein a graph or a table showing the business characteristics of each business sample, and the search means performs a business similar to the analysis target business when searching from the business pattern database. 2. The business analysis system according to claim 1, further comprising means for obtaining a graph or a table corresponding to the sample from the business pattern database and displaying and outputting the graph or table. 7. A flow of a series of work necessary for business processing performed by a plurality of processing staffs and the transfer of information necessary for the work using operation history data of a workflow management system that supports using a computer network. A business analysis system for analyzing characteristics of a business operated by the workflow management system, wherein each business operated by the workflow management system or each work constituting the business is handled by each processing of the workflow management system. Focus on the processing start time required from the time of arrival at the terminal of the user to the start of the corresponding processing on the terminal, and the processing time required from the start of the processing of the task or work to the completion. The values for each of the plurality of feature vector components defined in advance are Feature vector extracting means for extracting from a plurality of business samples having different business features, information relating to the analysis result of the business feature, and a business pattern in which values regarding each of the plurality of feature vector components are registered. A similarity between a distribution pattern of a plurality of feature vector component values extracted by the feature vector extraction unit and a distribution pattern of a plurality of feature vector component values for each of the business samples registered in the business pattern database; By retrieving from the business pattern database a business sample in which the distribution pattern of the plurality of feature vector component values is similar to the business to be analyzed, and analyzing information on an analysis result regarding the similar business sample as the analysis target. Acquired as the business feature of the business Business analysis system, characterized by comprising a search unit that. 8. A feature of a job operated by the workflow management system using operation history data of a workflow management system that supports a flow of a series of operations necessary for the job processing and information transfer necessary for the operation. A business analysis method for causing a computer to analyze, for each of a plurality of business samples having different business characteristics, information on the analysis result of the business characteristics and a plurality of feature vector components defined in advance for analyzing the business characteristics. A business pattern database in which values of the plurality of feature vector components are registered, and a feature vector extraction step of extracting the values of the plurality of feature vector components from operation history data related to the work to be analyzed; Based on the feature vector component values, A search step of retrieving a similar business sample from the business pattern database and acquiring information of an analysis result on the similar business sample as a business characteristic of the business to be analyzed. 9. The method according to claim 1, wherein the retrieving step comprises: determining a distribution pattern of the plurality of feature vector component values extracted in the feature vector extracting step and a plurality of feature vector component values relating to each of the business samples registered in the business pattern database. The method according to claim 1, wherein a similarity between the plurality of feature vector component values and a distribution pattern of the plurality of feature vector components is similar to the analysis target task. 8. The business analysis method according to 8. 10. The plurality of feature vector components include, at least, when a task or a task constituting the task arrives at a user terminal of the workflow management system, processing of the task or task starts at the user terminal. And / or a component related to the processing start time required until the processing is completed and / or the processing time required from the start of the processing of the task or the work to the completion thereof. The business analysis method according to claim 8. 11. A feature vector extracting step, further comprising: specifying at least one of a task constituting a task, a person in charge of processing, and a period to be analyzed to narrow down and designate a range of the task to be analyzed. 9. The business analysis method according to claim 8, wherein the extracting includes extracting the feature vector component value from operation history data relating to a range of the analysis target business narrowed down and specified in the specifying step. 12. A new analysis result information obtained by allowing an analyst to add / modify information of an analysis result corresponding to the similar work sample obtained in the search step, and the feature vector extraction step. 9. The method according to claim 8, further comprising a step of additionally registering the analysis target business as a new business sample in the business pattern database using the extracted feature vector component values of the analysis target business. Business analysis method. 13. The business pattern database has registered therein a graph or a table indicating business characteristics of each of the business samples, and the search step includes a business similar to the analysis target business when searching from the business pattern database. 9. The business analysis method according to claim 8, further comprising a step of acquiring a graph or a table corresponding to the sample from the business pattern database and displaying and outputting the graph or table. 14. A flow of a series of work necessary for business processing performed by a plurality of processing staffs and the transfer of information necessary for the work using operation history data of a workflow management system that supports using a computer network. A business analysis method for causing a computer to analyze characteristics of a business operated by the workflow management system, wherein each business operated by the workflow management system or each work constituting the same is performed by each of the workflow management systems. The processing start time required from the time of arrival at the terminal of the processing staff to the start of the corresponding processing at the terminal, and the processing time required from the start of the processing of the task or work to the completion thereof. By paying attention, the value for each of a plurality of feature vector components defined in advance is A feature vector extraction step for extracting from the operation history data on the target task, information on an analysis result of the task feature and a value on each of the plurality of feature vector components are registered for each of a plurality of task samples having different task features. A business sample having a pattern similar to the distribution pattern of the plurality of feature vector component values extracted in the feature vector extraction step is searched from the business pattern database, and information of an analysis result on the similar business sample is analyzed. A search step of acquiring as a task characteristic of the target task.