JP6209931B2 - Software asset utilization apparatus, software asset utilization method and program - Google Patents

Description

  The present invention relates to a software asset utilization apparatus, a software asset utilization method, and a program.

  In order for companies to design, construct, and operate information systems, various information, for example, information (business processes) that defines the content (business processes) of business in which information systems (hereinafter simply referred to as systems) are used. Definition information), system design information, source code, test specification information, operation specification information, and the like. These pieces of information are generally called software assets.

  Various tools for managing and utilizing such software assets have been proposed.

JP 2012-256197 A

  In the course of using a system over a long period of time, companies and the like often accumulate system expansions and modifications in view of the purpose of the company's activities. As a result, the system becomes intricately intertwined with various new and old software, hardware, and development technologies, making it difficult to grasp the entire software assets. This makes it difficult to expand and modify the system (renewal), and companies cannot quickly make the best and best system changes according to the changes in the purpose and requirements of activities and miss business opportunities. Faced with such problems.

  Conventional software asset management tools do not adequately address this problem. The factors are as follows, for example.

  The relationship between business processes and system implementation (for example, design, source code, operation specifications, etc.) is not always written as a document, but is often formed as tacit knowledge. These tacit knowledge cannot be managed by conventional software management tools. For this reason, when the system is renewed, for example, it takes a lot of time to investigate the range of influence on the business by remodeling. Furthermore, if the affected range cannot be determined in advance, a problem is detected in the evaluation phase at the end of the system renewal, resulting in wasted time and costs.

  In some cases, the system design specifications do not match the system implementation (eg, source code, operation specifications, etc.), or the design document itself does not exist. In such a case, when renewing the system, it is necessary to directly decode the source code and specify the location to be modified and the influence range due to the modification, which requires a lot of time and cost.

  Even if the system renewal succeeds, the conventional software management tool does not provide an effective means for storing the information obtained in the process in a form that can be reused or maintained. Therefore, there is a high possibility that the same problem will occur again when a system renewal is required later. Therefore, the renewal cost cannot be sufficiently suppressed in the long term.

  When an existing software asset tool tries to manage software assets related to a large-scale system, a keyword may be associated with information to be registered for easy access. However, there are cases where keywords cannot be set appropriately due to differences in registrants when registering keywords and information in association, inadequate keyword naming rules, setting errors, and the like. In such a case, an information group to be extracted by a certain keyword may not be obtained accurately.

  For example, when information extraction using “generator” as a keyword is performed, information associated with “generator”, “generator”, “generation unit”, and “generator” as keywords may not be acquired. Here, it is conceivable to perform notation fluctuation detection processing using a dictionary or the like indicating the similarity of keywords (for example, Patent Document 1).

  However, in a repository that handles large-scale software assets, a similar character string may be assigned to keywords that should be attached to different information groups. Even in the case where the same keyword should be attached, a keyword that is completely unrelated may be set due to an interpretation error of the registrant information. For such a repository, the above-described detection method of notation fluctuation, that is, the notation fluctuation detection method of handling a keyword as a simple character string, has a problem that the detection accuracy is not sufficient.

  The present invention has been made to solve such problems, and a software asset utilization apparatus, a software asset utilization method, and a program capable of accurately detecting a keyword notation fluctuation in a repository for managing software assets. The main purpose is to provide

One aspect of the software asset utilization device according to the present invention is:
For each block divided according to abstraction and detail, information nodes that are individual information units constituting software assets are arranged, and an asset management matrix in which the relationship between each information node is set by a relationship link is held. A repository,
The keywords set in the information node at the start point and the end point of each of the related links in the asset management matrix are counted, and the relevance level of each keyword pair is calculated according to the count, An information analysis unit that calculates the occurrence probability of the fluctuation of each keyword pair according to,
It is provided.

One aspect of the software asset utilization method according to the present invention is as follows.
For an asset management matrix in which information nodes, which are individual information units that make up a software asset, are placed on blocks divided according to abstraction and detail, and the relationship between each information node is set by relationship links.
The keywords set in the information node at the start point and the end point of each of the related links in the asset management matrix are counted, and the relevance level of each keyword pair is calculated according to the count, And an information analysis step of calculating the occurrence probability of the notation fluctuation of each keyword pair in accordance with.

One aspect of the program according to the present invention is as follows:
A program for causing a computer to execute the above-described software asset utilization method.

  The present invention can provide a software asset utilization apparatus, a software asset utilization method, and a program capable of accurately detecting a keyword notation fluctuation in a repository that manages software assets.

It is a figure which shows the structure of the software asset utilization apparatus 100 concerning Embodiment 1. FIG. FIG. 3 is a diagram showing a configuration of an information analysis unit 7 according to the first exemplary embodiment. FIG. 3 is a diagram illustrating a software asset management method in the software asset utilization apparatus 100 according to the first embodiment. FIG. 3 is a diagram illustrating a software asset management method in the software asset utilization apparatus 100 according to the first embodiment. FIG. 3 is a diagram illustrating a software asset management method in the software asset utilization apparatus 100 according to the first embodiment. FIG. 3 is a diagram illustrating a software asset management method in the software asset utilization apparatus 100 according to the first embodiment. FIG. 3 is a diagram illustrating a software asset management method in the software asset utilization apparatus 100 according to the first embodiment. FIG. 3 is a diagram illustrating a software asset management method in the software asset utilization apparatus 100 according to the first embodiment. FIG. 3 is a diagram illustrating a software asset management method in the software asset utilization apparatus 100 according to the first embodiment. FIG. 3 is a diagram illustrating a software asset management method in the software asset utilization apparatus 100 according to the first embodiment. FIG. 3 is a diagram illustrating a software asset management method in the software asset utilization apparatus 100 according to the first embodiment. FIG. 3 is a diagram illustrating a software asset management method in the software asset utilization apparatus 100 according to the first embodiment. FIG. 3 is a diagram illustrating a software asset management method in the software asset utilization apparatus 100 according to the first embodiment. FIG. 3 is a diagram illustrating a software asset management method in the software asset utilization apparatus 100 according to the first embodiment. FIG. 3 is a diagram illustrating a software asset management method in the software asset utilization apparatus 100 according to the first embodiment. FIG. 3 is a diagram illustrating a software asset management method in the software asset utilization apparatus 100 according to the first embodiment. FIG. 3 is a diagram illustrating a software asset management method in the software asset utilization apparatus 100 according to the first embodiment. FIG. 3 is a diagram illustrating a software asset management method in the software asset utilization apparatus 100 according to the first embodiment. FIG. 3 is a diagram illustrating a software asset management method in the software asset utilization apparatus 100 according to the first embodiment. 3 is a diagram conceptually illustrating an example of data held by an information holding unit 8 and a keyword holding unit 9 according to the first embodiment. FIG. It is a figure which shows the table structure of the repository 4 concerning Embodiment 1. FIG. It is a figure which shows the table structure of the repository 4 concerning Embodiment 1. FIG. It is a figure which shows the count result of the related link count part 71 concerning Embodiment 1. FIG. FIG. 10 is a diagram illustrating a processing result of a relevance calculation unit 72 according to the first embodiment. FIG. 10 is a diagram illustrating a processing result of a relevance calculation unit 72 according to the first embodiment. It is a figure which shows the processing result of the relevance test | inspection part 73 concerning Embodiment 1. FIG. 6 is a flowchart showing a flow of processing for checking a keyword notation fluctuation by the information analysis unit 7 according to the first exemplary embodiment; 3 is a diagram conceptually illustrating an example of data held by an information holding unit 8 and a keyword holding unit 9 according to the first embodiment. FIG. It is a figure which shows the count result of the related link count part 71 concerning other embodiment. It is a figure which shows the processing result of the relevance calculation part 72 concerning other embodiment. It is a figure which shows the processing result of the relevance calculation part 72 concerning other embodiment. It is a figure which shows the processing result of the relevance test | inspection part 73 concerning other embodiment. It is a figure which shows the count result of the related link count part 71 concerning other embodiment. It is a figure which shows the processing result of the relevance calculation part 72 concerning other embodiment. It is a figure which shows the processing result of the relevance calculation part 72 concerning other embodiment. It is a figure which shows the processing result of the relevance test | inspection part 73 concerning other embodiment. It is a figure which shows the structure of the software asset utilization apparatus 100 concerning this invention.

<Embodiment 1>
Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 shows a basic configuration of a software asset utilization apparatus 100 according to the present embodiment. The software asset utilization device 100 includes a repository 4 for holding information constituting software assets, and a processing device 3 that performs input, extraction, analysis, and the like of information to the repository 4. Moreover, it is preferable to provide the output device 2 which presents a user with the input device 1 for performing input operation with respect to a process part, the processing result of the processing device 3, etc. to a user.

  The input device 1 is an input interface from the user to the processing device 3, and functions as a unit for inputting information stored in the repository 4, for example. For example, it may be a keyboard, a mouse, a communication interface, or the like.

  The output device 2 is an output interface from the processing device 3 to the user, and is used for outputting various processing results. For example, it functions as a means for extracting information stored in the repository 4 to the outside. For example, it may be a display, a printer, a communication interface, or the like.

  The processing device 3 includes an information input unit 5 that stores information in the repository 4, an information extraction unit 6 that extracts information in the repository 4 according to various conditions and formats, and information that analyzes the extracted information using a predetermined method. An analysis unit 7 is included.

  The repository 4 includes, for example, an information holding unit 8 that stores information constituting software assets, a keyword holding unit 9 that holds keywords assigned to each information, and information on the relationship between pieces of information constituting software assets. It has a link table 10 for storing. The repository 4 is a logical storage unit provided in a storage device (not shown) included in the software asset utilization device 100.

(Repository)
Here, with reference to FIG. 3 to FIG. 19, what concept the repository 4 manages software assets will be described.

  As described above, a software asset is a collection of various pieces of information. Individual information units constituting these software assets are called information nodes. For example, a group of meaningful information (information unit) constituting business definition information, system design information, source code, test specification information, operation specification information, and the like is an information node. When documented software assets exist, all or part of the document data is an information node entity. On the other hand, an information node having no entity such as document data can be created for undocumented information. For example, even if there is no document describing the A function, if the information system requires the A function, the user creates a so-called empty information node with the name of the A function and stores it in the repository 4. Can be stored.

(Matrix and block)
The repository 4 stores each of these information nodes after classifying them according to two evaluation criteria of “abstract level” and “detail level”. The “abstract level” and “detail level” are each provided with a plurality of levels. Information nodes are assigned to any of these multiple levels. That is, the repository 4 allocates and stores information nodes at any position on the matrix defined by the two axes of “abstract level” and “detail level”.

  FIG. 3 illustrates a matrix assumed in the present embodiment. In this example, the vertical axis of the matrix indicates “abstraction” and the horizontal axis indicates “detail”. The “abstract level” and “detail level” increase as the distance from the origin increases. That is, an information node including more abstract or detailed contents is stored at a position further away from the origin. In the present embodiment, the level of “abstract level” and “detail level” is identified by a level number. That is, the level with the smallest “abstraction” and “detail” is called level number “1”. Hereinafter, as the “abstract level” and “detail level” increase, the level numbers “2”, “3”,.

  Note that the vertical and horizontal sizes (number of levels) of the matrix are not fixed, and can be set freely by the user according to the size and complexity of the software assets to be managed.

  The position on the matrix to which the information node is assigned is called a block. In the present embodiment, each block is specified by a set of two numbers in the format of “abstraction level number-detail level level number”. For example, in FIG. 3, the block to which the information node X is assigned is “1-2”.

(Abstract and detail)
Here, the significance of “abstraction” and “detail” will be described. The “abstraction level” is a scale for classifying information nodes according to the strength of dependency between information contents and system implementation. For example, when the five types of information nodes of business definition information, system design information, source code, test specification information, and operation specification information are classified into five levels of abstraction levels 1 to 5, the dependency on the implementation is the lowest. The business definition information is assigned to the highest abstraction level (abstraction level = 5). Thereafter, in order of increasing dependency on the implementation, abstraction level = 4 for system design information, abstraction level = 3 for source code, abstraction level = 2 for test specification information, and abstraction for operation specification information. Assign a degree level = 1.

  The “detail level” is a scale for classifying information nodes according to the level of design detail. For example, as shown in FIG. 5, as system design information, a function description in which an outline of the AA function is described in a natural language, a data relation diagram in which data transition in the AA function is described in, for example, a DFD (Data Flow Diagram), When assigning the two types of information nodes to the two levels of detail levels 1 and 2, the data relation diagram having the highest design detail is assigned to the highest detail level (detail level = 2). On the other hand, a detail level = 1 is assigned to a function description having a relatively low design detail.

  In the repository 4, although the level of detail is different, information nodes having the same level of abstraction have the same level of abstraction, and information nodes having the same level of detail have the same level of detail. It is preferable to assign.

  FIG. 4 shows an example of a table structure for concretely realizing the repository 4. The software asset utilization apparatus 100 can include an information holding unit 8 having the above table structure in a storage area on a storage device (not shown), for example. The information holding unit 8 stores the information name for identifying the information node stored in the repository 4, the block indicating the position of the information node, and the actual storage location of the information node entity (information entity, for example, document data). It is possible to hold a plurality of records having at least the location of the information entity indicated. In the present embodiment, the entity of the information node is stored in an arbitrary storage location outside the information holding unit 8. The entity of the information node may be held in the record of the information holding unit 8.

(keyword)
In the repository 4, a keyword can be assigned to each information node. The relationship between information nodes and keywords will be described with reference to FIGS.

  FIG. 5 shows a state where information nodes AA, BB, CC, and DD are registered in the two blocks (1-1 and 1-2) of the information holding unit 8, respectively. The same logical configuration and terminology are used for the information entity (for example, function description) of the information node included in block 1-1 and the information entity (for example, data relation diagram) of the information node included in block 1-2. Such an event can occur because it is often the case. Here, the information node AA in the block 1-1 and the information node AA in the block 1-2 often have some semantic relationship. Therefore, in the repository 4, the same keyword can be assigned to these information nodes as a tag for identifying such a semantic relation.

  FIG. 6 shows an example of a table structure for specifically realizing the association between information nodes and keywords. The information holding unit 8 in FIG. 6 is given a unique number to each information node (record) as compared to the information holding unit 8 in FIG. 4, and identifies even if there are a plurality of information nodes with the same information name. It is different in that it is configured to be possible. In this example, there are two information names AA, two BBs, two CCs, and two DDs.

  The keyword holding unit 9 is a table including at least one record that associates at least a keyword name with an information number list. The information number list specifies a plurality of information nodes associated with a keyword by numbers in the information holding unit 8.

(Related link)
The repository 4 can hold information regarding the relationship between individual information nodes. This relationship is called a relationship link.

  The relational link can be set between a certain information node and another information node existing in the same block as the information node, a block having a different abstraction level, and a block having a different detail level only.

  In the present embodiment, at least the following four types of relational links can be set. The equality relationship is a relationship link that associates information nodes that represent the same object or event based on different purposes or viewpoints. The subordinate relationship is a relational link that associates an information node that is a certain superordinate concept with an information node that is a subordinate concept or component of the information node. The refinement relationship is a relationship link that associates a certain information node with an information node that is a more detailed information node. The materialization relationship is a relationship link that associates an information node that represents an abstract concept with an information node that further materializes the information node. In addition to the above, an arbitrary relation link can be set in order to express the relevance of data between information nodes, the processing order, and the like.

  Note that the refinement relationship may be set between information nodes having different detail levels, or may be set between information nodes having the same detail level. The materialization relationship may be set between information nodes having different abstraction levels, or may be set between information nodes having the same abstraction level. In addition, it is preferable that the repository 4 holds a flag indicating whether the detailing relationship or the materializing relationship crosses the level of detail or the level of abstraction.

  FIG. 7 shows an organization structure and a business structure of a certain virtual company using two relationship links of a subordinate relationship and “related”. Here, it is assumed that “related” is a relational link indicating input / output information related to a certain business. This figure shows the following facts: This company has software assets (eg, documented manuals) related to “business flow”. This “business flow” includes a business flow related to the “sales department”. The business flow of the “sales department” is composed of three tasks, “order received”, “inventory check”, and “shipment instruction”. The input / output information in each business is “Order Form”, “Manufacturing Instruction”, and “Shipping Instruction”.

  FIG. 8 is an example of a table structure for specifically realizing the relational link of FIG. The information holding unit 8 of FIG. 8 is different from the information holding unit 8 of FIG. 6 in that it has “outgoing link” and “incoming link” elements. Here, “outgoing link” identifies a related link starting from the information node by a number defined in the link table 10 (described later). The “incoming link” identifies a related link whose end point is the information node by a number defined in the link table 10 (described later).

  The outgoing link and incoming link elements are not essential elements in the configuration of the information holding unit 8. The element overlaps with an element included in the link table 10 to be described later, and can be obtained by searching the link table 10 with the information node number. However, in actual operation, as the amount of information registered in the repository 4 increases, the search time for the link table 10 increases, and it is expected that the performance of the entire software asset utilization apparatus 100 will be greatly affected. Therefore, it is effective to have the outgoing link and incoming link elements directly in the information holding table as a means for speeding up.

  The link table 10 is a table for defining and storing relational links established between a plurality of information nodes. The link table 10 includes one or more records having at least numbers, attributes, start point information, and end point information as elements. The number is an identifier for uniquely identifying the related link. The attribute indicates the type of relationship link. In the start point information, the information node that is the start point of the related link is specified by the number defined by the information holding unit 8. In the end point information, the information node that is the end point of the related link is specified by the number defined by the information holding unit 8.

  9 to 11 show an organization structure and a business structure of a certain virtual company using two relational links, that is, a peer relation and a detailed relation. Lines with arrows at both ends indicate equality. The equality relationship is set between information nodes having essentially the same meaning content in the same block. A line with an arrow on one side indicates a refinement relationship. In this example, the refinement relationship is set between information nodes located in blocks having different refinement levels.

  FIG. 12 shows an organization structure and a business structure of a certain virtual company by using a relational link of materialization relation. One line with an arrow indicates a materialization relationship. In this example, the materialization relationship is set between information nodes located in blocks having different abstraction levels.

  13 to 19 show an example in which the entire software assets related to the organization structure and business structure of a certain virtual company are registered in the repository 4. 7, FIG. 9 to FIG. 11 and FIG. 12 described above are obtained by extracting a part of FIG.

  A user uses the input device 1 via a predetermined user interface to input data related to information nodes, keywords, and related links to the processing device 3. Upon receiving these inputs, the information input unit 5 of the processing device 3 performs processing for storing data in the information holding unit 8, the keyword holding unit 9, the link table 10 and the like of the repository 4, respectively.

  As shown in FIGS. 13 to 19 described above, the repository 4 stores information nodes having a huge data size. Therefore, it is important to be able to access the information node accurately using keywords. Hereinafter, a detailed configuration of the information analysis unit 7 that performs processing related to keywords will be described with reference to FIG.

  The information analysis unit 7 includes a related link count unit 71, a relevance level calculation unit 72, a relevance level inspection unit 73, and an inspection setting holding unit 74. Each processing unit of the information analysis unit 7 performs various types of processing for detecting a keyword notation fluctuation. The processing of each processing unit will be described with reference to FIGS. FIG. 20 is a diagram conceptually showing an example of data held by the information holding unit 8 and the keyword holding unit 9. In FIG. 20, the information node with the keyword “generator” is surrounded by a line, the information node with the keyword “generator” is surrounded by a thick line, and the information node with the keyword “common part” is surrounded by a dotted line. I have to. 21 and 22 are diagrams showing each table corresponding to FIG. In the following description, the table held by the information holding unit 8 is described as an information holding table, the table held by the keyword holding unit 9 is described as a keyword management table, and the table held by the link table 10 is described as a link table. To do.

  The related link counting unit 71 refers to the link table and the keyword management table and counts the number of links of the keyword set in the start point information node and the keyword set in the end point information node of each link. FIG. 23 shows the count result of the related link count unit 71 for FIGS. 21 and 22. For example, the related link count unit 71 sets “5” as the number of links from the keyword “generator” set in the start point information node to the keyword “generator” set in the end point information node (No. 4, No. 4 in the link table). 5, No. 8, No. 9, No. 14). Similarly, for example, the related link count unit 71 sets “3” as the number of links from the keyword “generator” set to the start point information node to the keyword “generator” set to the end point information node (No. 1, No. 1 in the link table). .2, No. 7).

  The relevance calculating unit 72 refers to the count result calculated by the related link counting unit 71 (FIG. 23), and calculates the ratio of the keyword set in the end point information node for each keyword set in the start point information node. FIG. 24 shows the ratio calculated from the count result shown in FIG. For example, for the keyword “generator” set in the start point information node, the relevance calculation unit 72 sets the ratio of the keyword “generator” set in the end point information node to 45.6% (5 / (5 + 3 + 3)), and the end point information node The ratio of the keyword “generator” set to 27.2% (3 / (5 + 3 + 3)), the ratio of the keyword “common part” set to the end point information node to 27.2% (3 / (5 + 3 + 3)), And calculate. The relevance calculation unit 72 similarly calculates the keywords “generator” and “common component” set in the start point information node.

The relevance calculating unit 72 calculates the relevance between the keywords using the calculated keyword ratio (FIG. 24). The degree of association calculation unit 72 calculates the degree of association between keywords using the following relational expression. The following formula is intended to increase the strength of ties between keywords.
Relevance between the keyword “A” and the keyword “B” = (the ratio of keywords having the keyword “A” as the starting point and the keyword “B” as the end point + the keywords having the keyword “B” as the starting point and the keyword “A” as the end point Ratio) / 2

  FIG. 25 shows the degree of association between keywords (keyword pairs) calculated from the keyword ratio shown in FIG. For example, the relevance calculation unit 72 sets the relevance of the keyword pair “generator”-“generator” to 51.1% ((27.2 + 75) / 2), and the relevance of the keyword pair “generator”-“common part” to 32. 4% ((27.2 + 37.5) / 2) and the relevance of the keyword pair “generator”-“common part” is calculated as 18.8% ((25 + 12.5) / 2).

  The degree-of-relevance calculating unit 72 may calculate the degree of association in consideration of, for example, the degree of similarity of the character strings of the keyword pairs instead of calculating a simple average. Specifically, when the character strings of the keyword pairs are similar, the relevance calculation unit 72 may perform the calculation process by weighting so that the relevance is higher than the average value. In other words, the relevance calculation unit 72 only needs to calculate the relevance of the keyword pair in consideration of the calculated keyword ratio (FIG. 24).

  The examination setting holding unit 74 holds a threshold value to be compared with the degree of association between keywords calculated by the degree of association calculation unit 72. The threshold value is set to a default value and may be configured so that the user can change the setting as appropriate.

  The degree-of-association checking unit 73 compares the degree of association between each keyword (keyword pair) calculated by the degree-of-association calculating unit 72 with the threshold value held by the examination setting holding unit 74. Then, the relevance checking unit 73 extracts keyword pairs having a relevance greater than or equal to a threshold value, and presents the extracted keyword pairs to the user.

  FIG. 26 is an example of an interface showing keyword pairs extracted from FIG. In this example, it is assumed that the threshold is 30%. The relevance checking unit 73 extracts and sorts keyword pairs having a relevance of 30% or more as a threshold, and displays an interface screen on which a treatment to be performed on each keyword pair can be selected. The interface screen is presented to the user by the output device 2.

  Examples of actions for keyword pairs include “correction”, “exclusion”, and “hold”. “Correction” means that correction is performed to unify two keywords into one. When “modify” is selected, each table in the repository 4 is rewritten. The rewriting process may be performed by the relevance checking unit 73 or may be performed by another processing unit. Here, the user may be able to select which keyword should be unified from the interface screen. When “exclude” is selected, the keyword pair is excluded from the processing target from the processing of each processing unit in the information analysis unit 7 thereafter. The exclusion setting keyword pair is held by the examination setting holding unit 74, and the relevance calculation unit 72 does not calculate the relevance of the keyword pair. When “hold” is selected, the process for the keyword pair is not performed. In this case, a countermeasure such as re-inspecting after resetting the threshold value can be considered.

In FIG. 26 described above, it is assumed that the keyword pair “generator” and “generator” needs to be corrected. Further, it is assumed that the keyword pair “generator” and “common part” need not be corrected. When the user selects “modify” for the former keyword pair and selects “exclude” for the latter keyword pair, the handling of the keyword notation is automatically terminated.

  Next, referring to the flowchart of FIG. 27, the flow of the examination process of the keyword notation fluctuation by the information analysis unit 7 will be described again.

  The related link count unit 71 refers to the link table and the keyword management table, and counts the number of links of the keyword set as the start point and the end point of each link (S1). The count result is, for example, as shown in FIG.

  The relevance calculating unit 72 refers to the count result, and calculates the ratio of keywords set in the end point (FIG. 24) for each keyword set in the start point (S2). The degree-of-association calculation unit 72 calculates the degree of association between the keywords (FIG. 25) using the calculated keyword ratio (FIG. 24) (S3).

The degree-of-association checking unit 73 compares the degree of association between the keywords (keyword pairs) calculated by the degree-of-association calculating unit 72 with the threshold held by the examination setting holding unit 74 (S4). Then, the relevance checking unit 73 extracts keyword pairs having a relevance greater than or equal to the threshold, sorts the extracted keyword pairs, and presents them to the user (S4, S5).

  Next, effects of the software asset utilization apparatus 100 according to the present embodiment will be described. As described above, the information analysis unit 7 analyzes the relationship of the relationship links between information nodes, and detects the occurrence probability of keyword notation according to the analysis result. The related link is a link indicating a relationship between information nodes. Therefore, by considering the relationship between the information nodes, it is possible to realize a keyword notation fluctuation with higher accuracy than simple character string comparison.

  In addition, the software asset utilization apparatus 100 detects keyword pairs that are likely to be swayed and provides measures for them. By performing the above-described “correction” treatment, it is possible to mechanically unify keywords. Further, by performing the above-described “exclusion” treatment, it is possible to avoid re-displaying a keyword pair that does not need to be corrected, and to prevent a keyword pair from being erroneously corrected. Further, since the keyword pairs extracted by the relevance checking unit 73 are sorted, the user can recognize the keyword pairs that have a high possibility of correction.

<Other embodiments>
In addition to the algorithm described in the first embodiment described above, it is possible to further improve the accuracy of checking the notation fluctuation by considering the attribute of the related link. Two examples for improving the inspection accuracy will be described below.

  First, a first modification will be described. As described above, attributes (substructure, detailing, materialization, and association) are assigned to the relationship link. Here, the relationship link having the lower structure attribute is a link indicating the parent-child relationship of the information nodes, and indicates that the relationship between the information nodes is very close. Therefore, the relation link count unit 71 performs the count process by paying attention to the relation link having the lower structure attribute.

  FIG. 28 is a diagram conceptually illustrating an example of data held by the information holding unit 8 and the keyword holding unit 9. In FIG. 28, the relational link indicating the lower structure attribute is indicated by a thick arrow. The related link count unit 71 performs the count process by setting the count number related to the related link of the lower structure attribute to twice the other count number.

29 to 32 show processing results (count results, calculation ratios, relevance levels, and extracted keyword pairs when the count value of the related link of the lower structure attribute is 2 and the count value of the other related links is 1. Corresponding to FIGS. 23 to 26). For example, the related link count unit 71 sets “6” as the number of links from the keyword “generator” set as the start point to the keyword “generator” set as the end point (No. 4 and No. 5 in the link table (FIG. 22)). No. 8, No. 9 (count value = 2), No. 14) are calculated (FIG. 29). The processing is the same as in the first embodiment except for the count processing of the related link count unit 71.

  As shown in FIG. 32, the degree of association between the keyword pair “generator”-“generator” is higher than that in FIG. 26, and the degree of association between the keyword pair “generator”-“common parts” is lower than that in FIG. The inspection accuracy is improved.

  Subsequently, a second modification will be described. In the second example, a count value is assigned for each attribute of the related link. Specifically, the count value of the relational link having the subordinate structure attribute is 3, the count value of the relational link having the detailed attribute = 2, the count value of the relational link having the specific attribute = 2, and the relation link having the related attribute The count value = 1. The count value is set in consideration of the strength of connection between information nodes. The related link count unit 71 performs a count process based on the count value.

  FIGS. 33 to 36 show the processing results (the count results, the calculation ratio, the relevance, and the extracted keyword pairs are shown and correspond to FIGS. 23 to 26) when the above-described count value allocation is used. . For example, the related link count unit 71 sets “12” as the number of links from the keyword “generator” set as the start point to the keyword “generator” set as the end point (No. 4 in the link table (FIG. 22) (count value = 3), No. 5 (count value = 3), No. 8 (count value = 2), No. 9 (count value = 2), No. 14 (count value = 2)) are calculated (FIG. 32). . The processing is the same as in the first embodiment except for the count processing of the related link count unit 71.

  As shown in FIG. 36, the degree of association between the keyword pair “generator”-“generator” is higher than that in FIG. 26, and the degree of association between the keyword pair “generator”-“common parts” is lower than that in FIG. The inspection accuracy is improved.

  The above-described embodiments are merely examples relating to application of technical ideas obtained by the present inventors. That is, the technical idea is not limited to the above-described embodiment, and various changes can be made.

Arbitrary processing in the software asset utilization apparatus 100 can also be realized by causing a CPU (Central Processing Unit) to execute a computer program. In this case, the computer program can be stored and provided to the computer using various types of non-transitory computer readable media. Non-transitory computer readable media include various types of tangible storage media. Examples of non-transitory computer-readable media include magnetic recording media (for example, flexible disks, magnetic tapes, hard disk drives), magneto-optical recording media (for example, magneto-optical disks), CD-ROMs (Read Only Memory), CD-Rs, CD-R / W, semiconductor memory (for example, mask ROM, PROM (Programmable ROM), EPROM (Erasable PROM), flash ROM, RAM (Random Access Memory)). The program may also be supplied to the computer by various types of transitory computer readable media. Examples of transitory computer readable media include electrical signals, optical signals, and electromagnetic waves. The temporary computer-readable medium can supply the program to the computer via a wired communication path such as an electric wire and an optical fiber, or a wireless communication path.

Finally, the outline of the software asset utilization apparatus 100 according to the present invention will be described again with reference to FIG. The repository 4 is an asset management in which information nodes, which are individual information units constituting software assets, are arranged in blocks divided according to abstraction and detail, and the relationship between information nodes is set by relationship links. Holds the matrix. Specifically, the repository 4 manages various data related to the asset management matrix using the information holding unit 8, the keyword holding unit 9, the link table 10, and the like.

  The information analysis unit 7 counts the keywords set in the start information node and the end information node of each related link in the asset management matrix, calculates the relevance of each keyword pair according to the count, The occurrence probability of the notation fluctuation of each keyword pair is calculated according to the relevance.

  The software asset utilization apparatus 100 can detect the fluctuation of the keyword notation more accurately than the simple character string comparison by performing the analysis using the relational link.

A part or all of the above embodiment can be described as in the following supplementary notes, but is not limited thereto.

<Appendix 1>
For each block divided according to abstraction and detail, information nodes that are individual information units constituting software assets are arranged, and an asset management matrix in which the relationship between each information node is set by a relationship link is held. A repository,
The keywords set in the information node at the start point and the end point of each of the related links in the asset management matrix are counted, and the relevance level of each keyword pair is calculated according to the count, An information analysis unit that calculates the occurrence probability of the fluctuation of each keyword pair according to,
Software asset utilization device comprising

<Appendix 2>
The information analysis unit
A relation link counting unit that counts keywords set in the information node at the start point and the information node at the end point of each of the relation links in the asset management matrix;
Based on the count result of the related link count unit, the ratio of keywords set in the information node at the end point is calculated for each information node at the start point, and each keyword pair included in the asset management matrix is calculated based on the ratio. A relevance calculator that calculates the relevance of
A relevance checking unit that extracts a keyword pair having a high possibility of notation fluctuation by comparing the relevance of each keyword pair calculated by the relevance calculation unit with a threshold value;
The software asset utilization apparatus according to appendix 1, comprising:

<Appendix 3>
The relationship link includes at least a directed substructure attribute link indicating the relationship between the upper and lower structures,
The software asset utilization device according to appendix 2, wherein the relation link count unit counts the count number of the lower structure attribute link more than the relation link having other attributes.

<Appendix 4>
The software asset utilization device according to attachment 2, wherein the relation link count unit changes a count number according to an attribute of the relation link.

<Appendix 5>
The software asset utilization apparatus according to any one of appendix 2 to appendix 4, wherein the repository is updated so as to integrate the notation of the keyword pairs extracted by the relevance checking unit.

<Appendix 6>
In any one of appendices 2 to 5, wherein the relevance check unit generates an interface screen that displays the keyword pairs having relevance greater than the threshold sorted in descending order of relevance. The software asset utilization device described.

<Appendix 7>
The software information utilization apparatus according to any one of Supplementary Note 1 to Supplementary Note 6, wherein the information analysis unit excludes a keyword pair designated in advance from the relevancy calculation processing.

<Appendix 8>
For an asset management matrix in which information nodes, which are individual information units that make up a software asset, are placed on blocks divided according to abstraction and detail, and the relationship between each information node is set by relationship links.
The keywords set in the information node at the start point and the end point of each of the related links in the asset management matrix are counted, and the relevance level of each keyword pair is calculated according to the count, A software asset utilization method, comprising: an information analysis step of calculating a probability of occurrence of notation fluctuation of each keyword pair according to

<Appendix 9>
In the information analysis step,
A relation link counting step of counting keywords set in the information node at the start point and the information node at the end point of each of the relation links in the asset management matrix;
Based on the count result in the relation link counting step, the ratio of keywords set in the information node at the end point is calculated for each information node at the start point, and each keyword included in the asset management matrix is calculated based on the ratio Relevance calculating step for calculating the relevance of the pair;
A relevance level inspecting step of extracting a keyword pair having a high possibility of notation fluctuation by comparing the relevance level of each keyword pair calculated in the relevance level calculating step and a threshold;
The software asset utilization method according to appendix 8, comprising:

<Appendix 10>
The relationship link includes at least a directed substructure attribute link indicating the relationship between the upper and lower structures,
The software asset utilization method according to appendix 9, wherein in the relation link counting step, the number of counts of the lower structure attribute links is counted more than that of the relation links having other attributes.

<Appendix 11>
The software asset utilization method according to appendix 9, wherein in the relation link counting step, the count number is changed in accordance with an attribute of the relation link.

<Appendix 12>
12. The software asset utilization method according to any one of appendix 9 to appendix 11, wherein the repository is updated so as to integrate the notation of the keyword pair extracted in the relevance level checking step.

<Appendix 13>
13. The software asset utilization method according to any one of appendix 9 to appendix 12, wherein in the information analysis step, a keyword pair designated in advance is excluded from the relevancy calculation processing.

<Appendix 14>
A program for causing a computer to execute the software asset utilization method according to any one of appendix 8 to appendix 13.

DESCRIPTION OF SYMBOLS 100 Software asset utilization apparatus 1 Input device 2 Output device 3 Processing apparatus 4 Repository 5 Information input part 6 Information extraction part 7 Information analysis part 71 Relation link count part 72 Relevance degree calculation part 73 Relevance degree inspection part 74 Inspection setting holding part 8 Information Holding unit 9 Keyword holding unit 10 Link table

Claims (10)

For each block divided according to abstraction and detail, information nodes that are individual information units constituting software assets are arranged, and an asset management matrix in which the relationship between each information node is set by a relationship link is held. A repository,
Count the number of keyword links set in the information node at the start point and the end point of each of the related links in the asset management matrix, and calculate the relevance of each keyword pair according to the count, An information analysis unit that calculates the probability of occurrence of the shake of each keyword pair according to the degree of association;
Software asset utilization device comprising The information analysis unit
A relation link count unit that counts the number of links of keywords set in the information node at the start point and the information node at the end point of each of the relation links in the asset management matrix;
Based on the count result of the related link count unit, the ratio of keywords set in the information node at the end point is calculated for each information node at the start point, and each keyword pair included in the asset management matrix is calculated based on the ratio. A relevance calculator that calculates the relevance of
A relevance checking unit that extracts a keyword pair having a high possibility of notation fluctuation by comparing the relevance of each keyword pair calculated by the relevance calculation unit with a threshold value;
The software asset utilization apparatus according to claim 1, comprising: The relationship link includes at least a directed substructure attribute link indicating the relationship between the upper and lower structures,
3. The software asset utilization apparatus according to claim 2, wherein the relation link count unit counts the count number of the lower structure attribute link more than the relation link having other attributes.   The software asset utilization apparatus according to claim 2, wherein the relation link count unit changes a count number according to an attribute of the relation link.   5. The software asset utilization apparatus according to claim 2, wherein the repository is updated so as to integrate the notation of the keyword pairs extracted by the relevance test unit.   The said relevance degree test | inspection part produces | generates the interface screen which sorted and displayed the keyword pair with a relevance degree larger than the said threshold value in order of the relevance degree, The one of Claims 2-5 characterized by the above-mentioned. Software asset utilization device described in 1.   The software information utilization apparatus according to claim 1, wherein the information analysis unit excludes a keyword pair designated in advance from a relevance calculation process. For the asset management matrix in which information nodes, which are individual information units constituting software assets, are arranged in blocks divided according to abstraction and detail, and the relationship between each information node is set by relationship links.
Count the number of keyword links set in the information node at the start point and the end point of each of the related links in the asset management matrix, and calculate the relevance of each keyword pair according to the count, A software asset utilization method comprising an information analysis step of calculating a probability of occurrence of notation fluctuation of each keyword pair according to the degree of association. In the information analysis step,
A relation link counting step of counting the number of keyword links set in the information node at the start point and the end point of each of the relation links in the asset management matrix;
Based on the count result in the relation link counting step, the ratio of keywords set in the information node at the end point is calculated for each information node at the start point, and each keyword included in the asset management matrix is calculated based on the ratio Relevance calculating step for calculating the relevance of the pair;
A relevance level inspecting step of extracting a keyword pair having a high possibility of notation fluctuation by comparing the relevance level of each keyword pair calculated in the relevance level calculating step and a threshold;
The software asset utilization method according to claim 8.   A program for causing a computer to execute the software asset utilization method according to claim 8 or 9.

Images