JP4796876B2 - Risk prediction apparatus and method - Google Patents

Description

  The present invention relates to a technique for predicting a risk that occurs in the course of a project. The present invention is suitable for use in software development.

  Examples of conventional software quality management systems are proposed in Patent Documents 1-5. In this conventional software quality management system, a target (planned value) is set for each evaluation item, and the risk of software development is predicted by evaluating the difference between the target and the actual result. That is, a target is set based on abundant experiences in the past, and the difference between the target and the actual result is evaluated. For example, if the difference is large, the risk is large, and if the difference is small, the risk is small.

JP 2000-250749 A JP 2003-345918 A JP 2000-11048 A JP 2003-186696 A JP 2003-345918 A

  In the conventional technology described above, it is essential to set a target (planned value) based on abundant past experience. If this goal is set incorrectly, the risk prediction accuracy naturally falls. Therefore, a wealth of experience is required to set goals.

  Therefore, even for companies with little experience in entering the market, even if the above-mentioned conventional software quality management system is introduced, it is not possible to properly set targets that should become risk criteria, and the system functions sufficiently. I can't let you.

  The present invention has been made under such a background, and does not depend on the difference between the target (planned value) set based on abundant experiences in the past and the actual results, and even if the experience is small. An object of the present invention is to provide a risk prediction apparatus and method capable of predicting the risk occurrence of a project in progress.

  Since the present invention holds the occurrence status of the risk generated in the software development process in time series together with the information on the contents thereof, and uses this as the risk prediction criterion, when setting the risk prediction criterion, Because it is only necessary to register the actual risk occurrence status in the database as it is, there is no need for work such as setting the target (planned value) in advance based on abundant experience in the past. The most important feature is not to do.

  Naturally, in the present invention, as the past experience is richer, the risk predictable range becomes wider, and it can be said that the present invention is the same as the prior art in terms of easier risk prediction. Thus, even in the case where there is little past experience (for example, even once), the fact that risk prediction can be performed with high accuracy with respect to matters that have experience is a significant point of the present invention. .

  That is, the present invention is a risk prediction device that predicts a risk that occurs in the course of a project based on accumulated information of risk that has occurred in the past, and the occurrence status of the risk that has occurred in the past together with information on its contents. Risk information holding means for holding in series, predicted information holding means for holding information on results for each of a plurality of categories in the ongoing project, and risks held in the risk information holding means A time series comparison between the occurrence status of the occurrence and the information on the results for each of the plurality of categories held in the predicted information holding unit, and a change pattern approximated by the extracting unit When a pattern is extracted, there is a means for displaying a warning of risk occurrence along with information on the content of the risk. Characterized in that was.

  According to this, it is possible to warn of the occurrence of the risk when the change pattern of the results in the ongoing project is close to the situation where the risk has occurred in the past. Accordingly, it is not necessary to set a target (planned value) based on abundant experiences in the past, and it is only necessary to hold a past situation in which a risk has actually occurred in a database or the like as it is in time series.

  The present invention can also be viewed from the viewpoint of a risk prediction method. That is, the present invention is a risk prediction method that is executed in a risk prediction apparatus that predicts a risk that occurs in the course of a project based on accumulated information of a risk that has occurred in the past. Is stored in time series together with information on its contents, information on the performance of each category in the ongoing project is held in time series, and the occurrence status of held risks and Compares information related to the results for each category in time series, extracts approximate change patterns, and displays approximate risk warnings together with information about the contents of the risks when approximate change patterns are extracted. And

  The present invention can also be viewed from the viewpoint of a program. That is, the present invention is a program that, when installed in a general-purpose information processing apparatus, causes the information processing apparatus to realize a function corresponding to the risk prediction apparatus of the present invention. By recording the program of the present invention on a recording medium, the information processing apparatus can install the program of the present invention using the recording medium. Alternatively, the program of the present invention can be directly installed in the information processing apparatus via a network from a server holding the program of the present invention.

  Thereby, the risk prediction apparatus of this invention is realizable using a general purpose information processing apparatus.

  According to the present invention, it is possible to predict the risk occurrence of a project in progress even if the experience is small, without depending on the difference between the target (scheduled value) set based on the abundant experience in the past and the actual result. it can.

  A risk prediction apparatus and method according to an embodiment of the present invention will be described in detail with reference to FIGS. FIG. 1 is an overall configuration diagram of a risk prediction apparatus according to the present embodiment. FIG. 2 is a block diagram of the main part of the database management server of this embodiment.

  As shown in FIG. 1, the present embodiment is a risk prediction apparatus for predicting a risk that occurs in the course of a project based on accumulated risk information that has occurred in the past. The server 20 includes a development property database 21, an evaluation information database 22, and a risk information database 23.

  The database access terminal 10 is a computer that operates under program control, communicates with the database management server 20, and can view, add, or change information in the development property database 21, the evaluation information database 22, and the risk information database 23.

  The database management server 20 is a computer that operates by program control. Upon receiving a request from the database access terminal 10, the database management server 20 stores information on the development property database 21, the evaluation information database 22, and the risk information database 23. Respond to. Further, when an event similar to the risk in the risk information database 23 occurs, information on the risk occurrence is displayed on the database access terminal 10.

  The development property database 21 holds development property data created during software development. The held data is read and written by the database management server 20. FIG. 4 is a diagram illustrating an example of data held in the development property database 21. In the example of FIG. 4, the project name, development program, program content, and progress are recorded.

  The evaluation information database 22 holds quality information data evaluated during software development. The held data is read and written by the database management server 20. FIG. 5 is a diagram showing an example of data held in the evaluation information database 22. In the example of FIG. 5, the results are recorded in units of one day from the project start date for each evaluation item. That is, information on the results for each of a plurality of categories in the ongoing project is held in time series.

  The risk information database 23 holds risk information generated during software development. The stored data is read and written by the database management server 20. FIG. 6 is a diagram illustrating an example of data held in the risk information database 23. In the example of FIG. 6, a risk occurrence situation and a situation explanation are recorded for each registered name. That is, the occurrence status of the risk that occurred in the past is held in time series together with information on the contents.

  As shown in FIG. 2, the database management server 20 chronologically displays the risk occurrence status held in the risk information database 23 and the information on the results for each of the categories held in the evaluation information database 22. And an approximate pattern extraction unit 1 that extracts a change pattern that approximates, and when an approximate change pattern is extracted by the approximate pattern extraction unit 1, a warning that displays a risk warning together with information about the content of the risk And a launching unit 2.

  Moreover, the present embodiment can be implemented as a program that, when installed in a general-purpose information processing apparatus, causes the information processing apparatus to realize a function corresponding to the risk prediction apparatus of the present embodiment. This program is recorded on a recording medium and installed in the information processing apparatus, or installed in the information processing apparatus via a communication line, whereby the database access terminal 10, the database management server 20, Functions corresponding to the development property database 21, the evaluation information database 22, and the risk information database 23 can be realized.

  Next, the risk prediction method of a present Example is demonstrated in detail with reference to FIGS. FIG. 3 is a sequence diagram showing an overall processing procedure of the risk prediction method of the present embodiment. As shown in FIG. 3, first, risk information to be displayed for risk occurrence is registered in the risk information database 23 via the database management server 20 by the database access terminal 10. This risk information is information on the number of problems, progress and risk status before and after the reference date that became a risk of software development. In the example of FIG. 6, the number of defects of the program a in the project β, the development cost of the project γ, and the like are registered together with information before and after the reference dates (the 10th and 18th days) that became the risk of software development. In the example of FIG. 6, “before and after the reference date” refers to the 7th to 14th days and the 15th to 22nd days from the start of the project.

  Next, information on the developed property created during software development is registered in the developed property database 21 via the database management server 20 by the database access terminal 10. For example, the information registered in the development property database 21 can be viewed, modified, or deleted between project managers as shared data between project managers. It may also be viewable in a department that manages each project in an integrated manner.

  Next, the quality information evaluated during software development is registered in the evaluation information database 22 via the database management server 20 by the database access terminal 10. In the example of FIG. 5, the progress status of project α (overall progress status), development costs, progress status of development of programs A to C (partial progress status), and the number of defects in programs A to C at the end of each day be registered.

  Here, the risk occurrence display procedure in the database management server 20 will be described. FIG. 7 is a flowchart showing a risk occurrence display procedure in the database management server 20. The database management server 20 inputs the evaluation information registered in the evaluation information database 22 to the approximate pattern extraction unit 1 (S1).

  Subsequently, the approximate pattern extraction unit 1 searches the risk information database 23 and searches for items similar to the input evaluation information (S2). For example, in the example of FIG. 6, “the number of defects of program a in project β” and “development costs in project γ” are registered as risk information items in the risk information database 23. On the other hand, in the example of FIG. 5, “number of defects of program A in project α” and “development cost in project α” are registered as items of evaluation information in the evaluation information database 22. As described above, when an item similar to the item registered in the risk information database 23 is registered in the evaluation information database 22, the evaluation information and the risk information are compared with each other as having a similar item (S3) ( S4). As a result, if there is no approximate change pattern (S5), it is determined that no risk has occurred, and the process is terminated. If there is an approximate change pattern (S5), this is notified to the alarm issuing unit 2. The alarm issuing unit 2 displays the risk occurrence on the database access terminal 10 (S6).

  FIG. 8 and FIG. 9 are diagrams illustrating an example of the approximate change pattern. In the example of FIG. 8, “the number of defects of program A in project α” is extracted from the evaluation information input to the approximate pattern extraction unit 1, and “program a in project β is extracted from the risk information registered in the risk information database 23. "Number of bugs in" is searched. The number in parentheses is the cumulative number of defects since the start of the project, and the number in parentheses indicates the number of new defects that occurred on that day.

  The number of defects of the program A in the project α is extracted for the 11th to 18th days from the start of the project, and the number of 5 defects has already been confirmed by the 11th day. Furthermore, on the 12th day, one new defect was confirmed and the cumulative total is six. Furthermore, on the 13th day, one new defect was confirmed, and the cumulative total is seven. Furthermore, on the 14th day, one new defect was confirmed, and the cumulative total is 8. Furthermore, on the 16th day, one new defect was confirmed, and the cumulative total is nine. Furthermore, on the 18th day, one defect was confirmed and the cumulative total is 10. This is a change pattern A.

  The number of defects in program β in project β is registered as risk information for the 7th to 14th days (base date is 10 days) from the start of the project, and 3 defects have already been confirmed by the 7th day. . Furthermore, on the 8th day, one new defect was confirmed, and the cumulative total is 4. Furthermore, on the 9th day, one new defect was confirmed, and the cumulative total is 5. Furthermore, on the 10th day, one new defect was confirmed, and the cumulative total is six. Furthermore, on the 13th day, 2 new defects were confirmed, and the cumulative total is 8. Furthermore, on the 14th day, one defect was confirmed and the cumulative total is nine. This is a change pattern a.

  In this example, the approximate pattern extraction unit 1 extracts the change pattern A and a as a change pattern that approximates by focusing on the fact that the number of defects increases more than twice in 8 days. The alarm issuing unit 2 is notified. Upon receiving this notification, the alarm issuing unit 2 displays on the database access terminal 10 “The number of problems with Program A in Project α is abnormal and requires attention.” And further, “Project β "There is a similar problem during the development of program a. A bug was found and solved by the simulator".

  The registrant of the evaluation information can take measures to prevent the occurrence of risk by referring to this display, recognizing the risk occurrence prediction, and taking appropriate measures.

  In the example of FIG. 9, “development cost in project α” is extracted from the evaluation information input to the approximate pattern extraction unit 1, and “development cost in project γ” is extracted from the risk information registered in the risk information database 23. Has been searched. The amount in parentheses is the cumulative amount from the start of the project, and the amount in parentheses indicates the amount of new occurrence on that day.

  The development costs in the project α are extracted for the 10th to 17th days from the start of the project, and an amount of 700,000 yen has already been used by the 10th day. Furthermore, on the 11th day, a newly generated amount of 100,000 yen is added and the total amount is 800,000 yen. Furthermore, on the 12th day, a newly generated amount of 100,000 yen is added and the cumulative total is 900,000 yen. Furthermore, on the 13th day, a newly generated amount of 100,000 yen is added and the cumulative total is 1 million yen. Furthermore, on the 14th day, a newly generated amount of 150,000 yen was added, for a total of 1.15 million yen. Furthermore, on the 15th day, a newly generated amount of 100,000 yen is added, and the total amount is 1.25 million yen. Furthermore, on the 16th day, a newly generated amount of 100,000 yen was added and the total amount is 1.35 million yen. Furthermore, on the 17th day, a new amount of 150,000 yen was added and the total amount is 1.5 million yen. This is a change pattern α.

  The development cost in the project γ is registered as risk information on the 15th to 22nd days (the reference date is 18th day) from the start of the project, and the amount of 1 million yen has already been used by the 15th day. Furthermore, on the 16th day, a new amount of 150,000 yen was added and the total amount is 1.15 million yen. Furthermore, on the 17th day, a newly generated amount of 100,000 yen was added, and the cumulative total is 1.25 million yen. Furthermore, on the 18th day, a new amount of 150,000 yen was added and the cumulative total is 1.4 million yen. Furthermore, on the 19th day, a newly generated amount of 200,000 yen was added, and the total amount is 1.6 million yen. Furthermore, on the 20th day, a new amount of 150,000 yen was added and the total amount reached 1.75 million yen. Furthermore, on the 21st day, a new amount of 150,000 yen was added, and the cumulative total is 1.9 million yen. Furthermore, on the 22nd day, a newly generated amount of 100,000 yen was added and the total amount is 2 million yen. This is a change pattern γ.

  In this example, the approximate pattern extraction unit 1 extracts the change patterns α and γ as approximate change patterns by paying attention to the fact that the budget has increased more than twice in 8 days, and warns that a risk has occurred. The issuing unit 2 is notified. Upon receiving this notification, the warning issuing unit 2 displays on the database access terminal 10 “development cost in project α is abnormal and requires attention.” And further, “development cost for project γ” "There is a similar abnormality in the case. Cause due to wrong selection of subcontractor".

  The registrant of the evaluation information can take measures to prevent the occurrence of risk by referring to this display, recognizing the risk occurrence prediction, and taking appropriate measures.

  In the examples of FIGS. 8 and 9, the change patterns A and a or α and γ are compared for the same number of days for easy understanding, but when performing actual risk prediction work, In most cases, an approximate change pattern is found when the number of days is less than the number of days of risk information registered in the risk information database 23, and risk occurrence can be avoided in advance. That is, a person in charge of evaluation information registers the evaluation information in the evaluation information database 22 at the end of each day. The approximate pattern extracting unit 1 displays the risk occurrence by the alarm issuing unit 2 when the approximation of the change pattern is recognized even if the number of days is less than the number of days registered in the risk information database 23. Therefore, normally, an approximate change pattern is often found when the number of days is less than the number of days of risk information registered in the risk information database 23, and risk occurrence can be avoided in advance.

  Also, in this example, in order to explain the basic technical idea of the present invention in an easy-to-understand manner, evaluation information and risk information include daily progress (%), development costs (yen), number of defects (cases) However, if the evaluation information and risk information are text information such as a work diary or a programming language, the keywords included in the text information are extracted. Thus, it is possible to extract a change pattern that approximates. Alternatively, if the evaluation information and the risk information are graphic information such as a graph or image information such as a photograph, a change pattern approximated by shape recognition of the graphic or image can be extracted.

  FIG. 10 is a sequence diagram showing a processing procedure when changing risk information already registered in the risk information database 23. As shown in FIG. 10, when changing risk information already registered in the risk information database 23, the registrant of risk information changes the risk information from the database access terminal 10 to the database management server 20. Request. Upon receiving the risk information change request from the database access terminal 10, the database management server 20 reads out the corresponding risk information from the risk information database 23 and displays it on the display unit of the database access terminal 10. The registrant of risk information can register the changed risk information in the risk information database 23 again after changing the risk information displayed on the display unit of the database access terminal 10.

  According to the present invention, it is possible to predict the risk occurrence of a project in progress even if the experience is small, without depending on the difference between the target (scheduled value) set based on the abundant experience in the past and the actual result. Therefore, the risk prediction apparatus of the present invention can be introduced for software development management even in companies with little experience, such as newly entered companies, and the efficiency of software development can be improved.

The whole block diagram of the risk prediction apparatus of a present Example. The principal part block block diagram of the database management server of a present Example. The sequence diagram which shows the whole process sequence of the risk prediction method of a present Example. The figure which shows an example of the information currently hold | maintained in a development property database. The figure which shows an example of the information hold | maintained at the evaluation information database. The figure which shows an example of the information hold | maintained at a risk information database. The flowchart which shows the risk occurrence display procedure in the database management server. The figure which shows an example (number of malfunctions) of the change pattern to approximate. The figure which shows an example (development cost) of the change pattern to approximate. The sequence diagram which shows the change procedure of risk information.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Approximate pattern extraction part 2 Alarm issuing part 10 Database access terminal 20 Database management server 21 Development property database 22 Evaluation information database 23 Risk information database

Claims (3)

A risk prediction device that predicts risks that occur in the course of a project based on accumulated information of risks that have occurred in the past,
Risk information holding means for holding the occurrence status of risks that occurred in the past along with information on their contents in time series,
Predicted information holding means for holding information on the results for each of a plurality of categories in the ongoing project in time series,
A risk occurrence state held in the risk information holding unit and information on the results for each of a plurality of categories held in the predicted information holding unit are compared in time series, and an approximate change pattern is extracted. Means,
A risk prediction apparatus comprising: means for displaying a warning of occurrence of a risk together with information relating to the content of the risk when an approximate change pattern is extracted by the extracting means. A risk prediction method executed in a risk prediction device that predicts a risk that occurs in the course of a project based on accumulated risk information that has occurred in the past,
The risk information holding means holds the occurrence status of the risk that occurred in the past in time series with the information about its contents,
Predicted information holding means holds information related to the performance of each category in the ongoing project in time series,
The approximating pattern extracting means compares the occurrence status of the risk held in the risk information holding means and the information on the results for each of the categories held in the predicted information holding means in time series and approximates them. Extract change patterns,
A risk prediction method, characterized in that, when a change pattern approximated by an approximate pattern extraction means is extracted by a display means , a risk occurrence warning is displayed together with information on the contents of the risk. A program for causing an information processing apparatus to function as the risk prediction apparatus according to claim 1 by being installed in a general-purpose information processing apparatus.

Images