JP7154113B2 - SYSTEM ASSET ANALYSIS DEVICE AND SYSTEM ASSET ANALYSIS METHOD - Google Patents

Description

The present invention relates to technology for analyzing system assets including software.

In a system including software, when a certain specification is changed, it is necessary to understand all the specifications that affect each other and to change them together. However, as software development becomes larger and more complex, it becomes more difficult to identify the extent of its influence. Therefore, there are increasing cases where the work of identifying the extent of influence affects development results such as quality and efficiency.

Therefore, conventionally, there have been provided methods for assisting in grasping the dependency relationships between design documents as development deliverables or description items therein.

Patent Literature 1 provides an influence range detection device and an influence range detection method capable of searching and specifying documents that have a correlation with documents to be changed.

In Patent Document 2, the file names of design documents (specifications, etc.) are used as identifiers to record the dependencies between files that have dependencies, and the files that may be affected by changing the target file are listed. ing.

Patent Literature 3 provides means for grasping dependencies using information inside a program and evaluating the degree of influence between program modules.

JP-A-2000-47861 JP 2009-265836 A JP 2010-282441 A

As the scale of systems including software increases, the cost of system construction and maintenance tends to increase, and the number of working personnel in teams such as maintenance and development tends to increase. Therefore, in the development and maintenance and modification work, the validity of the advance estimate of the man-hours and development amount, and the validity of the plan such as the division of work based on the estimate are factors that determine the success or failure of the development. When changing a certain specification, etc., it is necessary to fully understand the specifications, etc. that affect each other, and to change them accordingly. Here, design documents and implementations as development deliverables, and documents describing customer requirements, business specifications, etc. that are inputs thereof are referred to as system assets. As the system has undergone repeated maintenance over a long period of time, the cumulative number and scale of system assets that need to be understood increase, and it becomes difficult to identify the scope of impact. There are an increasing number of cases that affect the results of development such as

For example, if a human were to grasp the state of dependencies between deliverables in multiple upstream and downstream processes and try to find an appropriate range of influence, the analyst would spend a lot of time and effort on the analysis. Analysts (users) themselves need to understand the state of dependencies across functions and implementations, which requires a deep understanding of both functions and implementations. , is becoming more difficult.

SUMMARY OF THE INVENTION The present invention has been made in view of the circumstances described above, and an object thereof is to provide a technique capable of easily and appropriately analyzing related assets in system assets.

In order to achieve the above object, a system asset analysis device according to one aspect is a system asset analysis device for extracting assets related to an analysis request from a plurality of assets for software development, and determining a correspondence relationship between the related assets. a storage unit that stores association information indicating a property; a request reception unit that accepts an analysis request for an asset; a search unit that searches for a first asset that includes a term based on the analysis request; an asset identification unit that identifies a second asset related to the searched first asset; an asset analysis unit that identifies assets to be included in the extraction result based on the first asset and the second asset; a presenter for presenting results including the assets.

According to the present invention, related assets in system assets can be easily and appropriately analyzed.

FIG. 1 is an overall configuration diagram of a system resource analysis device according to the first embodiment. FIG. 2 is a functional configuration diagram of part of the system asset analysis device according to the first embodiment. FIG. 3 is a diagram for explaining the process and resources for system development according to the first embodiment. FIG. 4 is a configuration diagram of an asset correspondence table according to the first embodiment. FIG. 5 is a configuration diagram of a phrase appearance table according to the first embodiment. FIG. 6 is a flowchart of analysis processing by the system asset analysis device according to the first embodiment. FIG. 7 is a diagram for explaining asset correspondence information according to the first embodiment. FIG. 8 is a flowchart of related asset extraction processing according to the first embodiment. FIG. 9 shows an example of the analysis result by the natural language relationship analysis unit and the extraction result by the collation/integration unit according to the first embodiment. FIG. 10 is a configuration diagram of a similarity table according to the second embodiment. FIG. 11 is a configuration diagram of an asset correspondence table according to the third embodiment. FIG. 12 is a flowchart of asset correspondence table generation processing according to the third embodiment. FIG. 13 is a diagram for explaining assets extracted by the collation/integration unit according to the fourth embodiment. FIG. 14 is a diagram explaining an example of a display screen according to the fourth embodiment. FIG. 15 is a configuration diagram of an asset correspondence table according to the fifth embodiment. FIG. 16 is a configuration diagram of a semantic role table according to the fifth embodiment. FIG. 17 is a flowchart of analysis processing according to the fifth embodiment.

Some examples will be described with reference to the drawings. It should be noted that the examples described below do not limit the invention according to the claims, and that all of the elements and combinations thereof described in the examples are essential to the solution of the invention. is not limited.

In the following description, the information may be expressed in terms of "AAA table", but the information may be expressed in any data structure. That is, the "AAA table" can be called "AAA information" to indicate that the information is data structure independent.

First, the system resource analysis device according to the first embodiment will be explained.

FIG. 1 is an overall configuration diagram of a system resource analysis device according to the first embodiment.

The system asset analysis device 10 is configured by, for example, a general computer, and includes a CPU 100, a main storage device 101, an external storage device 102 as an example of a storage unit, a reading device 104, an input/output device 105, and a communication device. It comprises a device 106 and a bus 107 communicatively connecting them.

The CPU 100 executes various processes by executing programs stored in the main storage device 101 and the external storage device 102 . The main storage device 101 is, for example, a RAM (Random Access Memory), and stores programs executed by the CPU 100 and various information used by the CPU 100 . The external storage device 102 is configured by, for example, an HDD (Hard Disk Drive) or SSD (Solid State Drive), and stores various information and various programs.

A reading device 104 reads information from a portable storage medium 103 such as a CD-ROM or a DVD-ROM. In this embodiment, the system asset analysis program stored in the portable storage medium 103 is read by the reading device 104 and stored in the external storage device 102 . The input/output device 105 is a display (display device) that displays information, a keyboard, a mouse, or the like that receives input from a user. The input/output device 105 may be a device in which an input/output device is integrally configured, such as a display having a user interface function, or a device in which an input device and an output device are configured independently. There may be. The communication device 106 is a NIC (Network Interface Card) or the like, and connects the system asset analysis device 10 to a communication network.

In this embodiment, the system resource analysis program is executed by the system resource analysis device 10 alone. It may be executed by a connected server.

Next, the functional configuration of the system asset analysis device 10 will be described.

FIG. 2 is a functional configuration diagram of part of the system asset analysis device according to the first embodiment.

The external storage device 102 of the system asset analysis device 10 comprises a system asset storage section 121 and an analysis-related information storage section 122 .

The system asset storage unit 121 stores a plurality of assets (also called deliverables) in system development to be analyzed. Assets in system development include, for example, requirement specification information, requirement concept information, function specification information, design information, catalog item information, and source code for the system. The analysis-related information storage unit 122 stores various types of information necessary for analyzing system assets, such as information on external tools that can be used for analysis, an asset correspondence table 250 (see FIG. 4), and a phrase appearance table 260 (see FIG. 4), which will be described later. 5), etc. are stored.

The main storage device 101 stores computer programs such as a system asset analysis program, and the CPU 100 executes these programs to configure various processing function units. The processing function units configured by the CPU 100 executing the system asset analysis program in the main storage device 101 include an input reception unit 131 as an example of a request reception unit, a preprocessing unit 132, and an asset association information acquisition unit. 133, a natural language relationship analysis unit 134 as an example of a search unit, a collation/integration unit 135 as an example of an asset identification unit and an asset analysis unit, and an analysis result presentation unit 136 as an example of a presentation unit. . The processing of these functional units can be said to be processing executed by the CPU 100 .

The input reception unit 131 receives an analysis request for analyzing related assets from the user via the input/output device 105 . Analysis requests may be words, phrases, sentences, sentences used to analyze related assets. In this embodiment, a word based on an analysis request, for example, a word of an analysis request, a word included in a sentence of an analysis request, etc., is referred to as a "designated word". In this embodiment, using this specified word as a key, related assets are retrieved, analyzed, and the like. Note that the specified word may be one word or a plurality of words, that is, phrases or sentences.

The preprocessing unit 132 acquires data to be processed by the asset association information acquisition unit 133 and the natural language relationship analysis unit 134 from the system asset storage unit 121, and based on these data, the word/phrase occurrence table 260 and the asset A correspondence table 250 is generated and stored in the external storage device 102 .

The asset correspondence information acquisition unit 133 acquires the asset correspondence table 250 from the external storage device 102 . The natural language relationship analysis unit 134 acquires the word/phrase occurrence table 260 from the external storage device 102 , and searches for assets containing the specified word based on the analysis request received by the input receiving unit 131 based on the word/phrase appearance table 260 .

The collation/integration unit 135 collates/integrates the results obtained by the asset association information acquisition unit 133 and the natural language relationship analysis unit 134, identifies assets related to the analysis request, and analyzes the identified assets. result. The analysis result presenting unit 136 visualizes the analysis result by the collation/integration unit 135 so that the user can easily understand it, and displays it on the input/output device 105 .

Here, we will explain the concepts of upstream and downstream processes in system development, and the assets obtained as deliverables in each process.

FIG. 3 is a diagram for explaining the process and resources for system development according to the first embodiment.

The upstream process in system development refers to the process of generating requirements and specifications for the system expressed in a form that can be understood by the general public, such as documents and graphics using natural language. The deliverables (assets) obtained in the upstream process include a vague conceptual level and a concrete level such as detailed specifications. The upstream process is, for example, upstream process 1, which is the process of analyzing and extracting the customer's request concept, and the process of creating specific functional specifications, design information, catalog items, etc. to realize the request. It can be divided into an upstream process 2 and an upstream process 2. The deliverables (assets) of upstream process 1 are, for example, document information (document data) that enumerates and clarifies the concepts of customer requirements, and the deliverables (assets) of upstream process 2 are specific functional specifications, design documents, catalog items, etc.

The downstream process in system development refers to the implementation of source code, etc., for realizing the required specifications of the upstream process using a computer, in other words, the process of generating data expressed in a format expressed in a computer language. Also, the downstream process may include a test process for testing the source code. Some source codes are written in natural language (Japanese, etc.) as supplementary explanations. The downstream process deliverables (assets) are source codes and the like. A part of the source code, such as a module, is also an asset.

Next, the asset correspondence table 250 will be explained.

FIG. 4 is a configuration diagram of an asset correspondence table according to the first embodiment.

The asset correspondence table 250 is a table format of asset correspondence information for managing assets having a correspondence relationship in system development in association with each other. Each entry in the asset correspondence table 250 includes an asset name (function name) column 251 and an asset name (specification item name) column 252 . The asset name (function name) column 251 stores the name (function name) of the function module (an example of the asset) in the source code corresponding to the entry. In the asset name (specification item name) column 252, an upstream process specification item (asset name) having a corresponding relationship with the downstream process module whose name is stored in the asset name (function name) column 251 corresponding to the entry. example) is stored. It should be noted that, in place of the specification item name, ID information indicating the description location of the specification, such as a chapter number, may be used.

This entry shows that the module in the asset name (function name) column 251 is created based on the specification item in the asset name (specification item name) column 252 . More specifically, the entry on the first line in FIG. 4 indicates that the function f1, which is the asset of the downstream process, is a module for realizing the specification item S1, which is the asset of the upstream process. Note that the specific example of the correspondence relationship between the function name f and the specification item name S in FIG. 4 is merely an example.

The asset correspondence table 250 may be stored in the analysis-related information storage unit 122 by inputting information created in advance by the user into the system asset analysis apparatus 10 from the outside. , or may be created in the system resource analysis apparatus 10 according to a predetermined algorithm.

Note that the asset correspondence table 250 in FIG. 4 provides columns corresponding to downstream process assets and upstream process assets, respectively, to indicate the correspondence relationship. In some cases, a column corresponding to the asset of each upstream process may be provided to associate the asset of the downstream process with the asset of each process of a plurality of upstream processes. Further, although the asset correspondence table 250 in FIG. 4 associates downstream process assets with upstream process assets, assets in the same hierarchy may be associated with each other. Also, the unit of assets to be associated may be the entire source code or specification, or a part thereof.

Next, the word/phrase occurrence table 260 will be described.

FIG. 5 is a configuration diagram of a phrase appearance table according to the first embodiment.

The word/phrase appearance table 260 is a table for managing the presence/absence of word appearance in each asset. , and whether or not the corresponding word appears in the corresponding asset is set in the cell at the intersection of each row name and column name. Specifically, if the corresponding word exists in the corresponding asset, "1" is set, meaning that it exists, and if it does not exist, it means that it does not exist. "0" is set. In the word/phrase appearance table 260 shown in FIG. 5, the word/phrase appearance table 260 manages the appearance of words in assets. may

Next, processing operations of the system asset analysis device 10 will be described.

FIG. 6 is a flowchart of analysis processing by the system asset analysis device according to the first embodiment.

First, through the input/output device 105, the input receiving unit 131 receives an input of an analysis request for analyzing related assets from the user (S101).

Next, the preprocessing unit 132 performs preprocessing such as character code conversion on a plurality of assets (source code, design document information, etc.) (S102).

Next, the asset association information acquisition unit 133 acquires asset association information (S103), and the natural language relationship analysis unit 134 executes processing for analyzing natural language relationships (S104). The processing of steps S103 and S104 may be performed in parallel. It should be noted that the process of step S103 may be omitted if the asset correspondence information is not used.

Next, the collation/integration unit 135 calculates the analysis result based on the asset correspondence information obtained in step S103 and the analysis result of the natural language relationship (including the first asset) obtained in step S104. The related asset extraction process (see FIG. 8) to be integrated is executed (S106).

Next, the analysis result presentation unit 136 presents the analysis result obtained by the related asset extraction process (S106). Specifically, the analysis result presenting unit 136 displays the analysis result using the input/output device 105 .

Here, the processing of step S103 will be described in detail.

FIG. 7 is a diagram for explaining asset correspondence information according to the first embodiment.

In step S103, the asset association information acquisition unit 133 generates asset association information shown in FIG. 7 based on the information in the asset association information table 250 of FIG. The asset association information is synonymous with the asset association information table 250 of FIG. 4, expresses each asset by a node, and expresses the correspondence between assets by links connecting the nodes.

Next, the processing of step S104 will be described in detail. In addition, it demonstrates with reference to FIG.

The natural language relationship analysis unit 134 performs processing with reference to the phrase appearance table 260 . For example, when the specified word based on the analysis request is "age", the natural language relationship analysis unit 134 sets "1" in the column 261 having the column name "age" in the phrase appearance table 260. Identifies the name of the record (row) that contains the In this example, the natural language relationship analysis unit 134 identifies S1, S2, S3, S4, f1, and f2 as record names. Here, these names are names of assets that include "age" in the description. That is, it means that the word "age" is included in S1, S2, S3, S4, f1 and f2. Next, the natural language relationship analysis unit 134 determines that S1, S2, S3, S4, f1, and f2 are assets strongly related to the specified word "age". If the specified word is "withdrawal", the column 262 having "withdrawal" as the column name is referred to by the same process, and the assets strongly related to "withdrawal" are analyzed.

Next, the related asset extraction process (step S105 in FIG. 6) will be described in detail.

FIG. 8 is a flowchart of related asset extraction processing according to the first embodiment. In the description of the flowchart, a specific example in which the asset correspondence table 250 is in the state shown in FIG. 4, the word/phrase occurrence table 260 is in the state shown in FIG. 5, and the specified word is age will be described as appropriate. FIG. 9 shows an example of the analysis result by the natural language relationship analysis unit and the extraction result by the collation/integration unit according to the first embodiment.

First, the collation/integration unit 135 inputs the acquisition result of the asset correspondence information acquisition unit 133 and the analysis result obtained by the natural language relationship analysis unit 134 (S201).

Next, the matching/integration unit 135 acquires the function names of the assets that are not included in the analysis results obtained by the natural language relationship analysis unit 134 (S202). For example, when the natural language relationship analysis unit 134 outputs the analysis results (S1, S2, S3, S4, f1, f2) shown in FIG. Get the function name (f3, f7 in this example).

Next, the collation/integration unit 135 refers to the asset correspondence table 250 and extracts each record including the function name acquired in step S202 (S203). In a specific example, the collation/integration unit 135 extracts records containing the function names f3 and f7. According to these records, the function with the function name f3 corresponds to the specification item of S5, and the function with the function name f7 corresponds to the specification item of S4.

Next, the matching/integration unit 135 determines whether or not each of the specification item names associated with the function name is included in the analysis result of the natural language relationship analysis unit 134 (S204). In a specific example, the collation/integration unit 135 compares the specification item names S4 and S5 with the analysis results (S1, S2, S3, S4, f1, and f2) for determination.

As a result, if the specification item name is found in the analysis result (S204: YES), the collation/integration unit 135 adds the specification item name corresponding to this specification item name to the lower end of the analysis result by the natural language relationship analysis unit 134. A function name (an example of the second asset) is added (S205), this analysis result is output to the analysis result presentation unit 136 (S206), and the process ends. In the specific example, since the specification item name S4 is included in the analysis result, f7 corresponding to S4 is added to the analysis result as shown in FIG. 9B and output as the extraction result. Here, the extraction result shown in FIG. 9B is an asset group that is related (strongly related) to the specified word "age".

On the other hand, if the specification item name is not found in the analysis result (S204: NO), the process is terminated. In the specific example, since the specification item name S5 is not included in the analysis results, f3 corresponding to S5 is not added to the analysis results.

According to the related asset extraction process described above, as shown in a specific example, the extraction result obtained by adding the function name f7 to the analysis result (FIG. 9A) by the natural language relationship analysis unit 134 is the designated word " It is obtained as a group of assets that are related (strongly related) to “age”.

For example, if the description of the processing contents of a function in a source file is written in a natural language in the specifications, but is not written in that natural language as a comment in the source file, then that natural language is used as the designated word. Then, the analysis result by the natural language relationship analysis unit 134 does not include the source file. On the other hand, according to the related asset extraction process described above, such a source file (corresponding to f7 in the specific example) can be included in the extraction result based on the correspondence relationship between the specification and the source file. Therefore, assets (assets such as functions such as f7 in the specific example) that need to be grasped as the scope of influence of repair work or the like related to the designated word can be included in the extraction results without omission.

Next, a system asset analysis apparatus according to a second embodiment will be described. It should be noted that the second embodiment will be described with a focus on the differences from the first embodiment, and for the sake of convenience, will be described with reference to the drawings relating to the first embodiment.

In the system asset analysis apparatus according to the first embodiment, when the natural language relationship analysis unit 134 detects that the asset includes the same word as the reference word, the asset is "related (strongly related)" to the reference word. )”, the natural language relationship analysis unit 134 in the system asset analysis apparatus according to the second embodiment detects that the asset has a predetermined relationship even if it is not the same word (phrase, sentence, etc.) as the reference word. An asset is also associated with a specified word if it contains a word (eg, similar words (phrases, sentences, etc.)).

The natural language relationship analysis unit 134 according to the second embodiment refers to the word/phrase appearance table 260 in the same manner as the natural language relationship analysis unit 134 of the first embodiment, but the natural language relationship analysis unit 134 uses the word/phrase Characteristically, based on the appearance table 260, a similarity table 270 (see FIG. 10) is created and referred to.

Next, the data structure of the similarity table 270 and the flow of creating and referencing the similarity table 270 will be described.

FIG. 10 is a configuration diagram of a similarity table according to the second embodiment.

The similarity table 270 is a table for managing the degree of similarity between a certain designated word (“ackai” in the example of FIG. 10) and one or more words, and stores entries corresponding to each word. The similarity table 270 in FIG. 10 manages the degree of similarity between designated words and words. A record (row) of the similarity table 270 includes a word name column 271 and a similarity column 272 . The word name column 271 stores words that may be designated words in the analysis request, words that are extracted from assets, words that are column names of the word/phrase appearance table 260, and the like. The similarity column 272 stores a similarity indicating the degree of similarity between the word stored in the word name column 271 and a specific designated word. The degree of similarity may be, for example, an edit distance or another similarity index. Furthermore, for example, the specified word is written in katakana, the word is written in kanji, and the edit distance between the specified word and the word in katakana notation is used as the similarity. The edit distance may be expressed in roman letters, or in English words.

In the similarity table 270, the word name column 271 of a specific record (for example, the bottommost record) stores a specific specified word (“ackai” in FIG. 10) targeted by this similarity table 270. This record allows confirmation of a specific designated word.

According to the degree of similarity table 270, it can be seen that the degree of similarity of the word “withdrawal” to the specified word is “3”. In this example, the degree of similarity means that the smaller the numerical value of the degree of similarity, the more similar. Therefore, according to the similarity table 270 of FIG. 10, the numerical value of the similarity of "withdrawal" is the smallest, so it can be seen that it is most similar to the specified word "ackai".

The natural language relationship analysis unit 134 creates a similarity table 270 indicating similarities between the specified word input by the input receiving unit 131 and a plurality of words. Note that if the similarity table 270 corresponding to the specified word already exists, the similarity table 270 does not have to be created.

The natural language relationship analysis unit 134 refers to the similarity table 270 corresponding to the specified word and acquires words (an example of related words) similar to the specified word. For example, if the specified word is "ackai" and the similarity table 270 shown in FIG. ” is obtained.

After that, the natural language relationship analysis unit 134 analyzes the natural language relationship of "withdrawal" in the same way as for the specified word "ackai". Specifically, the natural language relationship analysis unit 134 refers to the word/phrase appearance table 260 shown in FIG. Let the name (S2, S3, f3) be the analysis result. That is, the analysis result that the word "withdrawal" is included in S2, S3, and f3 is generated. In this embodiment, the natural language relationship analysis unit 134 sets the analysis result by combining the analysis result of the specified word and the analysis result of the related word.

According to this example, even if "Ackai" is specified in the analysis request as an error in "Tekkai", analysis, etc. is performed using "Tekkai" or "Retract", and the analysis results are not returned. Since it is generated, it is possible to allow notation fluctuations such as writing errors in the analysis request and appropriately extract the assets desired by the user.

In addition, even if the asset contains the erroneous word "ackai" and the specified word is "withdrawal", the same process as above will be used to properly extract assets that include "ackai". can do.

Next, a system resource analysis device according to a third embodiment will be described. It should be noted that the third embodiment will be described with a focus on the points that differ from the first embodiment, and for the sake of convenience, will be described with reference to the drawings relating to the first embodiment.

The system asset analysis apparatus according to the third embodiment improves creation efficiency by providing the preprocessing unit 132 according to the first embodiment with a function for automatically creating an asset correspondence table. . Hereinafter, the pre-processing unit according to the third embodiment will be described as a pre-processing unit 132 for convenience.

The preprocessing unit 132 of the system asset analysis device calculates feature quantities that quantify the appearance frequency and positional features of the natural language described in the assets, and evaluates the similarity between assets based on the calculated feature quantities. Then, assets that are similar to each other are identified and collected as subsets, and an asset correspondence table 280 (see FIG. 11) is generated based on the subsets. Here, the preprocessing unit 132 corresponds to an association information generation unit.

FIG. 11 is a configuration diagram of an asset correspondence table according to the third embodiment.

The asset correspondence table 280 is a table format of asset correspondence information that associates and manages assets that have a correspondence relationship in system development. Each entry in the asset correspondence table 280 includes an asset name (function name) column 281 and an asset name (similar asset name) column 282 . The asset name (function name) column 281 stores the name (function name) of the function module (an example of the asset) in the source code corresponding to the entry. The asset name (similar asset name) column 282 stores one or more asset names that have a correspondence relationship with the downstream process module whose name is stored in the asset name (function name) column 281 corresponding to the entry. In the example shown in FIG. 11, the asset name (similar asset name) column 282 stores only the asset name of the module in the downstream process, that is, only the function name. (example) may be stored.

The processing operation of the asset correspondence table generation processing by the preprocessing unit 132 will be described.

FIG. 12 is a flowchart of asset correspondence table generation processing according to the third embodiment.

First, the preprocessing unit 132 receives a plurality of assets (functions, specifications, etc.) to be analyzed from the system asset storage unit 121 (S401), and creates a word/phrase appearance table indicating the appearance frequency of words/phrases in the assets ( S402). Here, in the word/phrase appearance table created in step S402, the value of each cell in the word/phrase appearance table 260 shown in FIG. ) is stored.

Next, the preprocessing unit 132 determines one asset to be processed from among the assets to be analyzed, acquires one row of the phrase occurrence table corresponding to this asset to be processed (S403), and Based on the frequency of appearance of each word of , the importance of each word in the asset is calculated, and the feature amount of the asset is calculated (S404). Here, the importance of each word may be a tf-idf value in the known TF-IDF method. Also, the asset feature amount is, for example, a vector having the importance of each word as a component. Note that the asset feature amount is not limited to this.

Next, the preprocessing unit 132 determines whether there is an asset for which the feature amount has not been calculated (S405), and if there is an asset for which the feature amount has not been calculated (S405: Yes), the process proceeds to step S403. Proceeding to perform the process of calculating the feature amount for the assets that have not yet been calculated.

On the other hand, if there is no asset for which the feature amount has not been calculated, that is, if the feature amount has been calculated for all the assets to be analyzed (S405: No), the preprocessing unit 132 All two asset combinations are derived (S406).

Next, the preprocessing unit 132 acquires a set of assets in the derived combination of the two assets as the following processing target (S407). Next, the preprocessing unit 132 calculates the degree of similarity between the set of assets to be processed (S408). Here, the degree of similarity between assets may be, for example, the cosine between vectors, which is the feature quantity of each asset.

Next, the preprocessing unit 132 determines whether the calculated similarity is higher than a predetermined reference similarity (S409), and if the calculated similarity is higher than the predetermined reference similarity (S409: If Yes), these assets are considered to be similar to each other, so this group of assets is added to the asset correspondence table 280 (S410), and the process proceeds to step S411. On the other hand, if the calculated similarity is not higher than the predetermined reference similarity (S409: No), it is considered that the assets in this group are not similar to each other. Proceed to S411.

In step S411, the preprocessing unit 132 determines whether or not there is an unprocessed combination of assets, and if there is an unprocessed combination of assets (S411: Yes), The process advances to step S407 in order to execute the same process as above. On the other hand, if there is no unprocessed asset combination (S411: No), it means that it has been determined whether or not all asset combinations are similar. End the process.

According to this asset correspondence table generation process, the asset correspondence table 280 as shown in FIG. 11 can be generated regardless of the settings input by the user.

Next, a system resource analysis apparatus according to a fourth embodiment will be described. In addition, the system asset analysis apparatus according to the fourth embodiment displays analysis results limited to the assets required by the user by the collation/integration unit 135 according to the first to third embodiments. It is intended to improve the processing efficiency and the visibility of the user. For convenience, the description will be made with reference to the drawings relating to the first embodiment.

The input reception unit 131 in the system asset analysis apparatus according to the fourth embodiment further receives from the user the asset type to be output. The collation/integration unit 135 extracts only assets of the asset type accepted by the input acceptance unit 131 from the analysis results. Here, the input reception unit 131 corresponds to the type designation reception unit.

A range of assets extracted by the collation/integration unit 135 will be described.

FIG. 13 is a diagram for explaining assets extracted by the collation/integration unit according to the fourth embodiment. FIG. 13A shows assets when the asset type to be output is not accepted, and FIG. 13B is a diagram explaining assets when "function" is accepted as the asset type.

When the input receiving unit 131 does not receive the asset type to be output, specifications S1 to S4 and functions f1, f2, and f7 are extracted as assets, as shown in FIG. When "function" is accepted as the asset type, only the functions f1, f2, and f7 are extracted without displaying the specifications, as shown in FIG. 13(B). As a result, it is possible to avoid adding unnecessary information such as the specification name to the analysis result for the user searching for the related function.

Next, an example of an analysis result display screen displayed by the analysis result presentation unit 136 will be described.

FIG. 14 is a diagram for explaining an example of a display screen according to the fourth embodiment. Note that the screen display example of FIG. 14 is an example different from the detection result of FIG.

For example, when the input receiving unit 131 does not receive the target asset type, the result of the collation/integration unit 135 does not limit the asset type. A diagram surrounded by a solid line 500 (a diagram including upstream process assets and downstream process assets) is displayed as a network diagram. Further, when the input receiving unit 131 receives "specifications" as the asset type to be targeted, the result of the collation/integration unit 135 is limited to assets with only specifications. , the analysis result presentation unit 136 displays a diagram surrounded by a dashed line 502 as a network diagram of assets. Further, when the input receiving unit 131 receives "function" as the asset type to be targeted, the result of the collation/integration unit 135 is limited to assets with functions only. The result presentation unit 136 displays a diagram surrounded by a dashed line 501 as a network diagram of assets.

According to the system asset analysis device according to the present embodiment, only assets of the selected type are displayed, so it is possible to avoid the display of unnecessary asset information for the user. You can grasp the necessary information.

Next, a system resource analysis apparatus according to a fifth embodiment will be described. In addition, the system asset analysis apparatus according to the fifth embodiment, in each asset associated as a related asset in any one of the natural language relationship analysis units 134 according to the first to fourth embodiments, each asset Based on the semantic role of the words in the system, the same or similar related words are analyzed, and the assets searched using the related words obtained by the analysis based on the semantic role are also added to the analysis results. . Here, the natural language relationship analysis unit 134 corresponds to a related word detection unit and an additional search unit.

The analysis-related information storage unit 122 of the system asset analysis device stores an asset correspondence table 250A and a semantic role table 290 as an example of semantic role information.

FIG. 15 is a configuration diagram of an asset correspondence table according to the fifth embodiment.

The asset correspondence table 250A has the same configuration as the asset correspondence table 250 of the first embodiment, but the values in the table are values for explanation of specific examples in the fifth embodiment.

FIG. 16 is a configuration diagram of a semantic role table according to the fifth embodiment.

The semantic role table 290 stores multiple entries corresponding to words in multiple assets. An entry in the meaning role table 290 includes an asset name column 291 , a type column 292 , a No column 293 and a meaning (point of view, attribute) column 294 .

The asset name column 291 stores the asset name that includes the word corresponding to the entry. The type column 292 stores the type of word corresponding to the entry. For example, if the word corresponding to the entry is a specification description, the type is specification, and if the word corresponding to the entry is a design document description, the type is design document, and the word corresponding to the entry is If it is a description (comment) in natural language in the source code, the type is code description, and if the word corresponding to the entry is a word in the programming language of the source code, the type is code. The No column 293 stores a number (sentence number) specifying a sentence in the asset containing the word corresponding to the entry. That is, the values in the No column 293 are the same for words existing in the same sentence in the same asset.

The meaning (point of view, attribute) column 294 stores information about the word corresponding to the entry and the semantic role of the word in the sentence. In this embodiment, the meaning (point of view, attribute) column 294 includes a condition column 295 , an effect column 296 , and others 297 . The condition column 295 stores words that correspond to the entry and that have a semantic role as conditions in sentences. The effect column 296 stores the details of the effect when the conditions are satisfied. The other column 297 stores, for example, descriptions of comments in the source code.

The semantic role table 290 may be created in advance by the user, or may be created by analyzing assets using a system asset analysis device.

FIG. 17 is a flowchart of analysis processing according to the fifth embodiment. The steps similar to those in FIG. 6 are denoted by the same reference numerals, and overlapping descriptions are omitted. A specific example will be described assuming that the asset correspondence table 250A is in the state shown in FIG. 15 and the semantic role table 290 is in the state shown in FIG.

In step S501, the natural language relevance analysis unit 134 identifies the semantic role of the specified word in the asset in which the specified word is detected, based on the semantic role table 290, and A word (an example of a related word) is identified from assets associated with the asset in which the designated word is detected. For example, if the specified word is "payment type", the asset S2 is obtained by searching, and the asset f20 is specified as an asset related to the asset S2 from the asset correspondence table 250A. Since the semantic role is the condition for the "payment type" in the asset S2, there is a high possibility that "WRK-sikyu-C-1" (entry 290-4), which is the condition for the semantic role in the asset f20, is synonymous. Extracted as related words. It should be noted that although words having the same semantic role may be determined as related words, other contents may be added for the purpose of further improving accuracy. For example, if an expression similar to the specified word is included, if other conditional words existing in the same sentence are the same or similar, or if the condition is satisfied, the condition corresponding to the specified word At least one of whether the effect is the same or similar to the effect when satisfying Considered in a concrete example, in this example, ``shikyu'' and ``supply'' (shikyu) are similar, and another condition ``14'' (entries 290-2, 290-5 ) are similar, and the effect when each condition is satisfied is "discount" (Waribiki) and "wari" (Wari), and the edit distance of katakana is close and similar. Type" and "WRK-Sikyu-C-1" are highly likely to be similar, and can be determined to be suitable as related terms.

Next, the natural language relevance analysis unit 134 searches for assets using the related terms identified in step S501 (S502). The assets detected by this search correspond to the third assets. For example, when "WRK-Sikyu-C-1" is specified as a related term, the natural language relevance analysis unit 134 searches for assets using "WRK-Sikyu-C-1", and the result , the asset f21 is detected, and the asset f21 is added to the result (S2, f20) already extracted by the collation/integration unit 135 . As a result, the analysis result presentation unit 136 presents S2, f20, and f21 as the analysis results.

According to the system asset analysis apparatus according to the fifth embodiment, it is possible to identify the related words corresponding to the specified word by considering the semantic role between the corresponding assets, so that the related assets can be extracted more appropriately. can.

In the fifth embodiment, the related words related to the specified word are identified by using the semantic role in the sentence. The categories are not limited to "conditions" and "effects", and may be other categories.

Other semantic role categories include, for example:
Example 1: Meaning by nominative, experiential, instrumental, objective, source, target, local, temporal (Fillmore's deep case (Charles J. Fillmore: Principles of case grammar, Sanseido, 1975)) Roles Example 2: Semantic roles (discourse relations between discourse units) when viewed as a discourse structure.
- TEMPORAL (time) ... "before", "after", etc. - CONTINGENCY (relationship possibility) ... cause ("reason", "result"), cognitive cause ("basis"), condition, epistemological Conditions (e.g. "related")
-COMPARISON
- EXPANSION --- example, selection, paraphrase, exception, enumeration, etc. Example 3: 5W1H (When, Where, Who, What, Why, How)

It should be noted that the present invention is not limited to the above-described embodiments, and can be modified appropriately without departing from the scope of the present invention.

For example, in the above embodiments, part or all of the processing performed by the CPU may be performed by a dedicated hardware circuit. Also, the programs in the above embodiments may be installed from program sources. The program source may be a program distribution server or storage media (eg, portable storage media).

DESCRIPTION OF SYMBOLS 10... System resource analysis apparatus 100... CPU, 101... Main storage device, 102... External storage device, 103... Portable storage medium, 104... Reader, 105... Input/output device, 106... Communication device, 121... System resource Storage unit 122 Analysis-related information storage unit 131 Input reception unit 132 Preprocessing unit 133 Asset association information acquisition unit 134 Natural language relationship analysis unit 135 Matching/integration unit 136 Analysis result presentation part

Claims (7)

A system asset analysis device for extracting assets related to an analysis request from a plurality of software development assets,
a storage unit that stores correspondence information indicating a correspondence relationship between related assets;
a request reception unit that receives an analysis request for assets;
a search unit that searches for a first asset that includes a term based on the analysis request;
an asset identification unit that identifies a second asset related to the first asset searched by the search unit based on the association information;
an asset analysis unit that identifies assets to be included in an extraction result based on the first asset and the second asset;
a presentation unit that presents results including the assets identified by the asset analysis unit;
with
The storage unit further stores semantic role information indicating semantic roles in assets of the plurality of words included in the plurality of assets,
The search unit, based on the semantic role information, detects words in the second asset that have the same or similar semantic role as words based on the analysis request in the first asset as related words, and Search for a third asset containing
The asset analysis unit is a system asset analysis device that identifies assets to be included in the extraction result based on the first asset, the second asset, and the third asset. The assets include a downstream process asset written in a computer language and an upstream process asset written in a natural language from which the downstream process asset was created;
2. The system asset analysis apparatus according to claim 1, wherein said association information includes information associating said upstream process assets with said downstream process assets. 3. The system according to claim 1, further comprising an association information generating unit that determines a correspondence relationship between the plurality of assets based on the appearance frequency of words in the assets and generates the association information. Asset analysis device. 4. The information generating unit according to claim 3, wherein the information generation unit detects feature quantities of the plurality of assets based on appearance frequencies of the plurality of words of the assets, and determines correspondence relationships between the assets based on the feature quantities. system asset analyzer. further comprising a type designation reception unit that receives designation of the type of asset to be presented as an extraction result by the presentation unit;
5. The system asset analysis device according to any one of claims 1 to 4, wherein the asset analysis unit specifies assets of the type received by the type designation reception unit as assets to be included in analysis results. A system asset analysis device for extracting assets related to an analysis request from a plurality of software development assets,
a storage unit that stores correspondence information indicating a correspondence relationship between related assets and semantic role information indicating semantic roles in assets of a plurality of words included in the plurality of assets;
a request reception unit that receives an analysis request for assets;
a search unit that searches for a first asset that includes a term based on the analysis request;
an asset identification unit that identifies a second asset related to the first asset searched by the search unit based on the association information;
a related term detection unit that detects, as a related term, a term of the second asset having the same or similar semantic role as the term based on the analysis request in the first asset, based on the semantic role information;
an additional search unit that searches for a third asset that includes the related term;
an asset analysis unit that identifies assets to be included in an extraction result based on the first asset and the third asset;
a presentation unit that presents results including the assets identified by the asset analysis unit;
A system asset analyzer comprising: A system asset analysis method by a system asset analysis device for extracting assets related to an analysis request from a plurality of software development assets,
Receive analysis requests for assets,
searching for a first asset containing terms based on the analysis request;
identifying a second asset related to the searched first asset based on the correspondence information indicating the correspondence relationship of the related assets stored in the storage unit;
Based on the semantic role information indicating the semantic role in assets of a plurality of words included in a plurality of assets, the second asset having the same or similar semantic role as the word based on the analysis request in the first asset Detecting a word in an asset as a related word, searching for a third asset containing the related word,
identifying assets to be included in analysis results based on the first asset, the second asset, and the third asset;
A system asset analysis method that presents analysis results that include identified assets.

Images