JP6695847B2 - Software parts management system, computer - Google Patents

Description

  The present invention relates to a software component management system, computer and method.

  In order to improve the management of enterprises, the importance of data analysis for increasing the efficiency of work and optimizing work plans is increasing. An engineer called by a name such as a data scientist analyzes data using a data analysis library. If the data analysis process can be standardized, the data analysis process is implemented as an application in the system and is automated.

  In recent years, more than 4000 data analysis libraries have been made public even with so-called open source alone. There is known a conventional technique of recommending a data analysis library from a large number of data analysis library groups in order to support the work of engineers (Patent Documents 1 and 2).

  In Patent Document 1, when making a new work plan, the work process is sorted from the past results, the statistical value of the results is calculated, and the estimated value of the work time for each work process is output.

  In Patent Document 2, for each of a plurality of columns having similar attributes, a column in which common data is stored is determined as a synonymous column, and the similarity between the plurality of data can be calculated.

JP, 2013-069222, A JP, 2011-232879, A

  When technicians analyze data, they use various libraries that differ depending on the purpose of business and industry. For example, in the data analysis of the first industry, a library that detects delays by comparing a production plan with manufacturing results, a library that analyzes the error history of lines when delays occur, and a library that replans manufacturing plans. To use. In the data analysis for the second industry, a library that detects excess inventory by comparing the planned inventory quantity with the actual inventory, a library that analyzes the order delivery history when excess inventory is detected, and a library that re-plans the order plan And so on.

  The conventional technology cannot reuse the data analysis know-how cultivated in one industry in another industry. For example, the manufacturing plan data analysis know-how cultivated in the first industry cannot be utilized for inventory management data analysis in the second industry.

  The present invention has been made in view of the above problems, and an object thereof is to provide a software component management system, a computer, and a method capable of efficiently searching software components and improving workability.

  In order to solve the above problems, a software component management system according to one aspect of the present invention is a classification processing unit that classifies each software component into a predetermined component classification, execution result information and a classification process that indicates the execution result of each software component. Based on the classification result of each software component by the section, the execution order processing unit that detects the execution order of the component category to which each software component belongs recorded in the execution result information, and the part category to which the requested software component belongs Next, the recommended component parts that can be executed next are detected based on the classification result and the execution order, the software components belonging to the detected recommended component parts are detected, and the detected software components are used. And a recommendation processing unit for outputting to the request source.

  According to the present invention, since software components are classified into component categories and managed, a software component that can be executed next to a software component that has been requested to be used is detected for software components that span multiple industries and organizations. Can be output as an output, which improves the usability and workability of the user.

FIG. 1 is an overall schematic diagram of a library management system as an example of a “software component management system”. It is explanatory drawing which shows the outline of operation | movement of a library management system. It is a system configuration diagram of a library management system. It is an example of composition of a library database. It is an example of composition of a library execution record database. It is an example of composition of business data. It is an example of composition of a processing phase database. It is an example of composition of a processing flow database. It is a structural example of a similarity database. It is a flow chart of processing phase classification processing. It is an example of a table for managing the feature amount of the library. It is a flow chart of processing flow extraction processing. It is a flowchart of a recommendation process. It is an example of a screen of data analysis processing. FIG. 9 is a system configuration diagram of a library management system according to a second embodiment. It is an example of an analysis application edit screen. It is an example of a table that manages flow information. FIG. 9 is a system configuration diagram of a library management system according to a third embodiment. It is an example of a data analysis application screen.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings. The computer described in this embodiment includes a library execution function 314 and a library search and recommendation function 313.

  The computer, for example, executes a library execution processing unit 314, a library recommendation processing unit 313, a processing phase classification processing unit 310, a processing flow extraction processing unit 311, and a similarity calculation processing unit 312. The computer can use the library information 320, the library execution record information 321, and the business data 322. Hereinafter, the execution record information may be abbreviated as the execution record.

  The similarity calculation processing unit 312 calculates the library similarity from the library execution record 321. The processing phase classification processing unit 310 calculates a processing phase as an example of “parts classification” from the library information 320, the library execution record 321, and the business data 322. The processing flow extraction processing unit 311 calculates a processing flow from the processing phase information 323 and the library execution record 321.

  When the recommendation processing unit 313 receives a library execution request from the terminal 50 used by a technician such as a data scientist who is a user, the recommendation processing unit 313 executes the requested library and executes the library execution result and the library and processing phase. From the correspondence table and the processing flow, the next processing phase to be calculated by the user is calculated, and the library corresponding to the next processing phase is recommended.

  Even if the combination of Patent Document 1 and Patent Document 2 described above is possible, it is not possible to classify and manage the library as a high-level concept in the processing phase, so that different organizations such as industries, companies, factories, departments, etc. It is not possible to share the know-how of library utilization. Generally, it is difficult to determine the similarity between the libraries because the values of the data contained in the libraries are different when the organizations are different.

  On the other hand, according to the library management system of the present embodiment, for example, when a user analyzes data for business improvement, it is possible to recommend an appropriate library according to the purpose of data analysis. As a result, the user can efficiently use the library that matches the purpose of the data analysis from the plurality (usually a large number) of libraries, and can facilitate the data analysis work.

  The present embodiment will be described below with reference to a plurality of examples. It is obvious that the library management system according to this embodiment is not limited to the configuration described in the examples.

  In the first embodiment, the data analysis work is facilitated by recommending (suggesting) the library to be executed next (used) from the library management system to the user at the time of reactive data analysis using, for example, the Jupiter of Python. Turn into.

  In the second embodiment, for example, when a data analysis application is developed, the library management system recommends the library to be executed next to the user, thereby facilitating the development work of the data analysis application.

  In the third embodiment, for example, at the time of data analysis work using a dashboard (management tool) of a data analysis application, the user highlights the library to be executed next, thereby facilitating the data analysis work.

  In the following description, the information of this embodiment will be described using expressions such as “aaa table”, “aaa list”, “aaaDB (Database)”, (aaa is an arbitrary character string), but these information are not necessarily in that format. The information does not have to be stored in, and may be represented by a data structure other than a table, a list, a DB, a queue, or the like. Therefore, the “aaa table”, “aaa list”, “aaaDB”, etc. may be referred to as “aaa information” to indicate that they do not depend on the data structure.

  Moreover, when describing the contents of each information, the expressions “identification information”, “identifier”, “name”, “name”, and “ID (IDentification)” are used, but these can be replaced with each other. is there.

  Further, the information illustrated in the present embodiment may include fields (not shown), some data fields may be divided and registered in another DB, and some data fields may be registered. The field does not have to exist.

  In addition, the processing flow illustrated in the present embodiment may include processing steps (not shown), or some processing steps may not exist, and some processing steps are executed. The order may be changed.

  In the following description, the term “program” may be used as the subject, but the program executes the processing specified by the processor while using the memory and the communication port (communication control device). May be used as the subject. Further, the process disclosed with the program as the subject may be performed by a computer such as a management server or an information processing device. Further, part or all of the program may be realized by dedicated hardware.

  The various programs may be installed in each computer by a program distribution server or a computer-readable storage medium. In this case, the program distribution server includes a processor and a storage resource, and the storage resource further stores the distribution program and the program to be distributed. When the processor executes the distribution program, the processor of the program distribution server distributes the distribution target program to another computer.

  It should be noted that the present invention is not limited to the embodiments described below, but includes various modifications and equivalent configurations within the scope of the appended claims. For example, the above-described embodiments have been described in detail in order to explain the present invention in an easy-to-understand manner, and the present invention is not necessarily limited to those having all the configurations described. Further, part of the configuration of one embodiment may be replaced with the configuration of another embodiment. Further, the configuration of another embodiment may be added to the configuration of one embodiment. Further, a part of the configuration of each embodiment may be added / deleted / replaced with another configuration.

  Further, each of the above-mentioned configurations, functions, processing units, processing means, etc. may be realized by hardware, for example, by designing a part or all of them with an integrated circuit, and the processor realizes each function. It may be realized by software by interpreting and executing the program.

  Information such as a program, a table, and a file for realizing each function is recorded in a memory, a hard disk, a storage device such as SSD (Solid State Drive), or an IC card, an SD card, a DVD, a Blue Ray Disk, or another optical disk. It may be stored in the medium. Further, it is not necessary that all of them are stored in the same storage medium, and some programs, tables, and files may be stored in another storage medium, or they may be connected via a communication network (not shown). It may be stored in the storage medium of the device.

  Further, the control lines and information lines are shown to be necessary for explanation, and not all the control lines and information lines necessary for mounting are shown. In reality, it can be considered that almost all configurations are connected to each other.

  Further, the computer has an input / output device. Examples of the input / output device include a display, a keyboard, a pointer device, a tablet terminal, and a smartphone, but other devices may be used. Further, as an alternative to the input / output device, a serial interface or an Ethernet interface is used as the input / output device, a display computer having a display or a keyboard or a pointer device is connected to the interface, and display information is transmitted to the display computer, The display computer may perform display by receiving the input information from the display computer, or the input and display at the input / output device may be replaced by receiving the input. Note that the interface may be hereinafter referred to as I / F.

  Further, the computer has a communication I / F. As an example of the communication I / F, a LAN (Local Area Network) connection terminal, a SAN (Storage Area Network) connection terminal, and a wireless communication connection device are conceivable, but devices other than these may also be used. Good. The communication network may be a wired network or a wireless network, but either communication network may be used.

  A first embodiment will be described with reference to FIGS. FIG. 1 is an overall schematic diagram of a library management system according to the first embodiment.

  In the library management system 1 of FIG. 1, an outline of data and a processing flow inside the server 10 described later with reference to FIG. 3 is shown.

  The library management system 1 provides a user, such as a data scientist, with a service for executing a library and a service for recommending a library to be executed next, via a terminal 50.

  As will be described later, the library management system 1 includes a processing phase classification processing unit 310 as a “classification processing unit”, a process flow extraction processing unit 311, a similarity calculation processing unit 312, and a recommendation process as an “execution order processing unit”. A processing unit such as a unit 313 and a library execution function 314 is provided. Further, the library management system 1 includes an information storage unit such as a library database 320, a library execution record database 321, business data 322, a processing phase database 323, a processing flow database 324, and a similarity database 325, as will be described later.

  In FIG. 1, the processing units 310 to 314 and the information storage units 320 to 325 are illustrated as if they are provided in one computer, but the processing units and the information storage unit are distributed to a plurality of computers. May be arranged. The library management system 1 provided by this embodiment can be realized by the cooperation of the computers via a communication network or the like. For example, the DBs as the information storage unit may be provided in a computer or storage system different from the computer in which all or some of the processing units 310 to 314 are installed. The respective processing units 310 to 314 do not have to be mounted on the same computer, and may be provided on different computers (physical computer or virtual computer).

  In the following description, the database may be abbreviated as DB. Further, in the figure, the "section" of the "processing section" may be omitted and displayed. For example, the processing phase classification processing unit 310 may be referred to as the processing phase classification processing 310.

  In the library management system 1, as described above, the similarity calculation processing unit 312, the processing flow extraction processing unit 311, and the processing phase classification processing unit 310 respectively include a library DB 320, a library execution record DB 321, and business data 322. Information of the similarity DB 325, the processing phase DB 323, and the processing flow DB 324 is output by referring to necessary data among the above.

  The user can use the operation terminal 50 to send an execution request for a desired library to the library execution function 314 of the library management system 1. The library management system 1 executes the library requested by the terminal 50. The library execution result is registered in the library execution record DB 321.

  When the recommendation processing unit 313 detects that the library has been executed, the recommendation processing unit 313 calculates a library candidate to be executed next (a library to be recommended), and outputs the calculated candidate to the operation terminal 50. As a result, the user can easily confirm the library to be executed next, so that the data analysis work can be facilitated and the work efficiency of the analysis work is improved. The outline of the operation will be described with reference to FIG.

  FIG. 2 shows an outline of the overall operation of the library management system 1. The library DB 320 includes a plurality (usually a large number) of libraries SA, SB, SC, SX, SY, SZ ,. ． ． I remember.

  The processing phase classification processing unit 310 processes each library into the processing phases P1, P2, P3 ,. ． ． Classify into. The processing phase is a conceptualization of processing contents such as “data collection”, “data analysis”, and “optimization”. Therefore, libraries having different concrete contents can be classified into the same processing phase.

  On the other hand, the similarity calculation processing unit 312 assigns each library to the groups Gα, Gβ ,. ． ． Classify separately. A group is a group that uses a library and distinguishes, for example, organizations such as companies, factories, research laboratories, government offices, and departments. Therefore, the library management system 1 classifies and manages each library by the processing phase and by the group using the classification table 326. The library classification table 326 can be configured by linking the processing phase database 323 and the similarity database 325.

  The processing flow extraction processing unit 311 extracts the execution order of the processing phases by analyzing the execution results stored in the library execution result database 321. That is, the processing flow extraction processing unit 311 calculates the processing flow 3240, which is the execution order of the processing phase, by reading the database 321 for the execution order of each library and replacing each library with the corresponding processing phase.

  When the recommendation processing unit 313 receives the execution request from the terminals 50-1 and 50-2, the recommendation processing unit 313 executes the requested library, updates the database 321, and extracts and proposes a library that can be executed next. To do.

  For example, when a user of a certain business type Gα causes the libraries SA and SB and the library management system 1 to execute using the terminal 50-1, the recommendation processing unit 313 extracts the next library SC based on the processing flow 3240. And propose to the terminal 50-1.

  Similarly, when a user of another business type Gβ causes the libraries SX and SY and the library management system 1 to execute using the terminal 50-2, the recommendation processing unit 313 determines the next library SZ based on the processing flow 3240. It is extracted and proposed to the terminal 50-2. More specifically, the recommendation processing unit 313 refers to the transitions of the processing phases P1, P2, and P3 extracted in the processing flow 3240, identifies the library SZ corresponding to the processing phase P3 in the industry Gβ in the library classification table 326, Propose the identified library SZ.

  In order to improve the accuracy of library proposal, it is preferable to execute a library having a predetermined value or more. However, in this embodiment, the predetermined value is not particularly limited.

  In the example of FIG. 2, the user of the business type Gα can use the rich execution record 321, so that the library SC to be executed next to the library SB can be easily obtained. On the other hand, the user of the business type Gβ different from the business type Gα executes the library SY next time by using the processing flow 3240 obtained from the business type Gα even if the user has a poor performance record. The library SZ should be easily available.

  FIG. 3 is a block diagram showing a configuration example of a library management system (which can also be called a computer system).

  The library management system 1 includes, for example, a server 10 and an operation terminal 50. The server 10 and the operation terminal 50 are connected via a communication network CN so as to be capable of bidirectional communication. It should be noted that the server 10 may be configured by one computer or may be configured by linking a plurality of computers. The computer on which the server 10 is placed may be a physical computer or a virtual computer.

  The server 10 has, for example, a processor 20, a memory 30, and a communication interface 40. In the figure, the processor is abbreviated as CPU, and the interface is abbreviated as I / F.

  The memory 30 stores computer programs 310 to 314 and data groups 320 to 325. The processor 20 realizes the function as the library management system 1 by executing the computer programs 310 to 314.

  The computer program includes, for example, a processing phase classification processing unit 310, a processing flow extraction processing unit 311, a similarity calculation processing unit 312, a recommendation processing unit 313, and a library execution processing unit 314. The data group includes, for example, a library DB 320, a library execution record DB 321, business data 322, a processing phase DB 323, a processing flow DB 324, and a similarity DB 325.

  The operation terminal 50 may be connected to the server 10. The operation terminal 50 is a computer that operates the server 10. The operation terminal 50 has an input / output device 51. The input / output device 51 is a device that inputs / outputs data according to a user operation. The input data is transmitted to the server 10 via the communication network CN. The input / output device 51 can also receive the data transmitted from the server 10 via the communication network CN and display the received data. Examples of the input / output device 51 include a keyboard, a touch panel, a mouse, a voice input device, a display, a printer, and a voice output device.

  Hereinafter, the data group will be described first, and then the computer program will be described.

  FIG. 4 shows an example of the library DB 320. The library DB 320 stores information about libraries that can be used by the server 10 and that can be used by any user during data analysis.

  The library DB 320 is information prepared in advance manually or by some program. The library DB 320 has, for example, an ID field 3201, a process name field 3202, a process detail field 3203, an input field 3204, and an output field 3205. Other databases described below are similar, but may include fields other than the illustrated fields, or one database may be divided into a plurality of databases.

  The ID field 3201 is information that uniquely identifies the library. The processing name field 3202 is information for recording the name of the library. The processing details field 3203 is information for recording the details of the processing of the library. The input field 3204 is a group of information indicating the meaning and format of data input to the library, and records information about zero or more arbitrary input data. The output field 3205 is a group of information indicating the meaning and format of the data output by the library, and records information about zero or more arbitrary output data.

  FIG. 5 shows an example of the library execution record DB 321. The library execution record DB 321 is information for recording the history of the library executed by the server 10, and is information prepared in advance manually or by some program.

  The library execution record DB 321 has, for example, an execution time field 3211, a library ID field 3212, an input field 3213, an output field 3214, and a user field 3215.

  The execution time field 3211 is information indicating the time when the library was executed. The library ID field 3212 is information that uniquely identifies the executed library. The input field 3213 is information indicating the location of the data actually input to the library. The output field 3214 is information indicating the data output by the library. The user field 3215 is information that uniquely identifies the user who executed the library. The user field 3215 can be associated with a user information management table (not shown) that manages user information. The user information management table can manage, for example, the user ID, the user name, the group name to which the user belongs (company name, department name, etc.), the skill level of the user, the skill level, the network address of the operation terminal 50, and the like.

  FIG. 6 shows an example of the business data 322. The work data 322 is information on the work that the user wants to improve. The table illustrated in FIG. 6 is a part of business data. The business data 322 is information prepared in advance manually or by some program.

  The business data 322 illustrated in FIG. 6 shows a part of the information regarding the line of the manufacturing factory. The business data 322 has, for example, a line ID field 3221, a part number field 3222, a scheduled input time field 3223, a scheduled discharge time field 3224, a scheduled time record field 3225, and a discharged time record field 3226.

  The line ID field 3221 is information that uniquely identifies a production line. The part number field 3222 is information that uniquely identifies the part manufactured on the production line. The scheduled injection time field 3223 is information indicating the scheduled time at which the production of parts will start on the line. The scheduled discharge time field 3224 is information indicating a scheduled time at which the manufacturing of the parts is completed on the line. The input time record field 3225 is information indicating the time when the actual manufacture of the parts is started on the line. The discharge time record field 3226 is information indicating the actual time when the manufacturing of the parts on the line is completed.

  FIG. 7 shows an example of the processing phase DB 323. The processing phase DB 323 is information that holds the correspondence relationship of which processing phase the library is classified into, and is generated by the processing phase classification processing 211.

  The processing phase DB 323 has, for example, a processing phase field 3231 and a library ID field 3232. The processing phase field 3231 is information that uniquely identifies the processing phase determined by the processing phase classification processing unit 310. The library ID field 3232 is information that identifies a library group classified into processing phases, and stores the IDs of one or more libraries.

  FIG. 8 shows an example of the processing flow DB 324. The processing flow DB 324 is information that holds in what order components (libraries) with different processing steps are used, and is generated by the processing flow extraction processing 223.

  The processing flow DB 324 has, for example, a transition source field 3241, a transition destination field 3242, a condition field 3243, and an ID field 3244.

  The transition source field 3241 is information indicating which processing phase the executed library belongs to. The transition destination field 3242 is information indicating which processing phase the library belonging to is executed next to the executed library. The condition field 3243 indicates information about what value the execution result of the library has when the transition source processing phase transits to the transition destination processing phase. The ID field 3244 is information that uniquely identifies the processing flow registered in the processing flow DB 324.

  An image of the processing flow indicated by the information of the processing flow DB 324 described in this embodiment is shown as a processing flow 3240 in FIG. In the process flow 3240, the phase names such as “Phase P1” are simply displayed as “P1” and the conditions of “true” and “false” are simply omitted as “T” and “F”.

  FIG. 9 shows an example of the similarity DB 325. The similarity DB 325 is information that holds the correspondence relationship to which group the library is classified, and is generated by the similarity calculation processing unit 312.

  The similarity DB 325 has, for example, a group field 3251 and a library ID field 3252. The group field 3251 is information that uniquely identifies a similar processing group determined by the similarity calculation processing unit 312. The library ID field 3252 is information specifying a library group that is determined to be similar to similar processing, and stores IDs of one or more libraries.

  Next, an example of a computer program executed by the library management system 1 will be described.

  The processing phase classification processing unit 310 is a processing for classifying a plurality of existing libraries into processing phases corresponding to the purpose of data analysis by the user.

  FIG. 10 is a flowchart showing a procedure example of the processing phase classification processing unit 310 executed by the server 10.

  The processing phase classification processing unit 310 acquires the information registered in the library DB 320 and the library execution record DB 321, and the information registered in the business data 322 (S11). The processing phase classification processing unit 310 extracts the feature amount of the library from the data acquired in step S11 (S12).

  The processing phase classification processing unit 310 classifies a combination of close feature quantities among the feature quantities extracted in step S12 as a process phase (S13). Finally, the processing phase classification processing unit 310 outputs the results classified in step S13 as the processing phase DB 323 (S14).

  The processing phase classification processing unit 310 needs to be executed once or more before the processing flow extraction processing unit 311 and the recommendation processing unit 313. Further, the processing phase classification processing unit 310 may be executed regularly, for example, every day at noon or every Sunday at midnight, or in the case of detecting that the information in the library DB 320 has been changed. The program may be automatically executed by some program, or may be manually executed.

  FIG. 11 shows an example of the library feature amount 327 extracted by the processing phase classification processing unit 310 from the library DB 320 and the library execution record DB 321.

  The library feature amount 327 includes, for example, an ID field 3271, an input format field 3272, an input source field 3273, an input data characteristic field 3274, an execution frequency field 3275, an output destination field 3276, and an output data field 3277. Have.

  The feature amount 327 of the library may have entries corresponding to all the entries included in the library DB 320, but the present invention is not limited to this, and a part of the entries may be saved, or a part of the duplicated data may be saved. May be included.

  The ID field 3271 is identifier information that uniquely identifies the library and corresponds to the ID field 3201 of the library DB 320. The input format field 3272 is information indicating the format of input data and is extracted from the input field 3204 of the library DB 320.

  The input source field 3273 is information indicating where the data input to the library is connected, and is extracted from the input field 3213 of the library execution record DB 321. The input data characteristic field 3274 is information indicating the update frequency of the data entry as the characteristic of the data actually input to the library, and is extracted from the business data 322.

  The execution frequency field 3275 is information indicating the frequency with which the library is executed, and is extracted from the execution time field 3211 of the library execution record DB 321. The output destination field 3276 is information indicating where the output of the library is connected, and is extracted from the input field 3213 of the library execution record DB 321. The output data field 3277 is information indicating new data calculated by the library, and is extracted from the output field 3205 of the library DB 320 and the output field 3214 of the library execution record DB 321.

  In this embodiment, these fields 3271 to 277 are illustrated as the feature amount 327 of the library. The feature amount 327 of the library is not limited to the fields shown in FIG. 11, and some of the illustrated fields may not exist or may include unillustrated fields. The fields not shown may be specified by some program.

  The process flow extraction processing unit 311 is a process of calculating information indicating how the processing phase has changed in the data analysis by the user.

  FIG. 12 is a flowchart showing a procedure example of the processing flow extraction processing unit 311 executed by the server 10.

  The processing flow extraction processing unit 311 acquires the library execution record DB 321 and the processing phase DB 323 (S21). The processing flow extraction processing unit 311 replaces the library ID field 3212 of the library execution record DB 321 with the processing phase field 3231 based on the data acquired in step S21 (S22).

  The processing flow extraction processing unit 311 extracts a processing flow (S23). For example, the processing flow extraction processing unit 311 determines an entry near the execution time field 3211 of the library execution record DB 321 as a continuous processing phase, and extracts the continuous processing phase as a transition source field 3241 and a transition destination field 3242 of the processing flow. To do. Further, the processing flow extraction processing unit 311 extracts the output field 3214 of the library execution record DB 321 when the processing flow is continuous as a condition field 3243 of the processing flow DB 324. Finally, the processing flow extraction processing unit 311 outputs the processing flow DB 324 extracted in step S23.

  In step S21, for example, it is possible to preferentially extract the processing flow of an expert or a skilled user. For example, in step S21, the entry acquired from the library execution result database 321 may be limited to the user ID acquired by some program and executed. Alternatively, in step S21, data may be acquired by limiting the range of values in the execution time field 3211 for the purpose of excluding past trends and prioritizing the latest trends. A predetermined entry may be extracted from the library execution record DB 321 according to the conditions other than the user attribute and the execution time of the library described above.

  In step S23, the value input to the condition field 3243 may be exactly the same as that of the output field 3214, may be registered by narrowing down to a characteristic output format, or the result of performing some statistical processing may be registered. Alternatively, a value corresponding to a table indicating some feature amount may be registered.

  The processing flow extraction processing unit 311 needs to be executed once or more before the recommendation processing unit 313. Further, the processing flow extraction processing unit 311 may be executed periodically, such as every day at noon or every Sunday at midnight, or may detect that the information in the library execution record DB 321 has been changed. In this case, it may be automatically executed by some program or may be manually executed.

  The similarity calculation processing unit 312 is a process for calculating the similarity for determining whether the libraries are frequently used in the same group (company, department, etc.).

  The similarity calculation processing unit 312 acquires the library execution record DB 321, calculates the similarity of a library group frequently used by the same user from the user ID field 3215 as a similar library, and outputs the similarity DB 325.

  The similarity calculation processing unit 312 may calculate the similarity based on the input / output relationship of data between the input field 3213 and the library as a library of similar classification for a component group having a connection relationship for input / output. Alternatively, the similarity calculation processing unit 312 may acquire the library DB 320 and calculate the similarity using natural language analysis from the information described in the process name field 3202 and the process detail field 3203.

  The similarity calculation processing unit 312 needs to be executed once or more before the recommendation processing unit 313. In addition, the similarity calculation processing unit 312 may be executed regularly, for example, every day at noon or every Sunday at midnight, or that the information in the library DB 320 or the library execution record DB 321 is changed. In the case of detection, it may be automatically executed by some program or may be executed manually.

  The recommendation processing unit 313 is a process of recommending a library to be executed next when the user executes the library.

  FIG. 13 is a flowchart showing a procedure example of the recommendation processing unit 313 executed by the server 10.

  When the recommendation processing unit 313 detects the completion of library execution in the library execution processing unit 314 as a trigger event (S31), the recommendation processing unit 313 retrieves data from the library execution record DB 321, the processing phase DB 323, the processing flow DB 324, and the similarity DB 325. Each is acquired (S32).

  The recommendation processing unit 313 calculates one or more libraries to be recommended, that is, a library estimated to be used next by a user who has executed a certain library, based on the data acquired in step S32 (S33). Details of step S33 will be described later. Finally, the recommendation processing unit 313 outputs the recommended library calculated in step S33 to the terminal 50 that has requested execution of the library (S34).

  A detailed example of step S33 for calculating the recommended library will be described. The recommendation processing unit 313 determines the processing phase of the executed library by acquiring the processing phase field 3231 of the processing phase DB 323 corresponding to the library ID field 3212 of the library execution record DB 321 (S331). ..

  When a plurality of users share the library management system 1, the recommendation processing unit 313 may be executed by focusing on the data of a specific user using the information in the user ID field 3215 of the library execution record DB 321.

  The recommendation processing unit 313 compares the processing phase information specified in step S331 with the transition source field 3241 of the processing flow DB 324 to obtain a matching entry, and further, the information of the output 3214 of the library execution record DB 321 and the processing flow. The condition field 3243 of the DB 324 is compared, and the transition destination field 3242 of the matching entry is determined as the next processing phase (S332).

  Here, when a plurality of transition destination fields 3242 are acquired, the recommendation processing unit 313 may determine that there are a plurality of recommendation target libraries, and may proceed to the recommendation target calculation step S333, or may be entered several times in the past. The transition destination field 3242 may be prioritized and then output with reference to the transition of the processing phase obtained from.

  The recommendation processing unit 313 acquires the ID of the library stored in the library ID field 3252 of the processing phase DB 323 corresponding to the transition destination field 3242 determined in step S332 (S333).

  When there are a plurality of library IDs, the recommendation processing unit 313 acquires the data of the library ID field 3252 of the similarity DB 325 in which the library ID field 3212 of the latest entry of the library execution record DB 321 exists, and the library ID field 3252 is obtained. Priority may be set such that the library IDs existing in both of the library ID fields 3212 are calculated as high priority candidates. Alternatively, for example, the prioritization may be performed by a program such as increasing the priority of a component having a large number of executions. Alternatively, the output may be limited to one library having the highest calculated priority.

  The library execution processing unit 314 is a process of executing a library. Upon receiving the library execution request (use request) from the operation terminal 50, the library execution processing unit 314 executes the requested library and transmits the execution result to the operation terminal 50.

  When the library is executed, the library execution processing unit 314 outputs information on the execution of the library as a library execution log. The output library execution log is registered in the library execution record DB 321 automatically or manually by some program. The library execution processing unit 314 may be executed by a program (not shown) existing in the server 10 or may be executed without communication with the operation terminal 50.

  FIG. 14 is a screen configuration diagram showing a display example of recommendation results in the present embodiment. The display screen 60 includes a plurality of display areas such as an input field 61, an output field 62, and a recommendation field 63, for example.

  The input field 61 is a field in which the user inputs an analysis function such as a library name for data analysis. The output field 62 is a field for displaying the library input to the input field 61 and the execution result of the analysis process. The recommendation field 63 is a field for displaying a library (function) candidate to be executed next.

  For example, the recommendation field 63 is fixedly displayed at the bottom of the screen 60. The input field 61 and the output field 62 are scrolled according to the usage situation of the user, and the new input field 61 and the new output field 62 are displayed. As a result, the user can execute sequential data analysis while referring to the library recommended by the library management system 1.

  According to the present embodiment, the user can select the optimum library group according to the processing flow from the library groups registered in the library DB 320 when performing data analysis using the library management system 1. , Data analysis work can be simplified.

  Further, for example, by using the processing flow DB 324 extracted based on a specific user attribute such as an expert or an expert, even a user who is a beginner of data analysis can refer to the past process flows by the expert or the expert. Data analysis can be performed and it can be useful for training data analysts.

  Further, it may be combined with a program for receiving and executing the library recommended by the operation terminal 50 by the library management system 1 according to the present embodiment. This allows the data analysis itself to be semi-automated or automated.

  When the library management system 1 according to the present embodiment does not exist, the library execution history in a certain group is not obtained until the same user has a history of using both the library groups of a certain group and another group. It can also be used by other groups. It should be noted that this description is not a description of the conventional art but a description of a comparative example for explaining the effect of the present embodiment.

  According to the library management system 1 of the present embodiment, it is possible to extract information that exceeds the classification by the similarity DB 325, that is, the processing phase DB 323 and the processing flow DB 324, and utilize the library execution history (library execution record). Therefore, in this embodiment, even if the same user has no record of using the library groups of both a certain group and another group, the library execution history of a certain group can be used in other groups. It is possible and convenience is improved.

  For example, if a certain group is a library related to factory operation monitoring and another group is a library related to retailing, users who use the library across both groups (data analysis It is considered difficult to obtain the person's history. However, according to the present embodiment, the library execution history can be utilized across groups such as industries, companies, and departments, so that the efficiency of data analysis work can be improved.

  A second embodiment will be described with reference to FIGS. Each of the following embodiments including this embodiment corresponds to a modification of the first embodiment, and therefore the differences from the first embodiment will be mainly described.

  In this embodiment, when the user develops the data analysis application, the library management system 1a recommends the data analysis library to be used next to the user, thereby facilitating the development of the data analysis application.

  FIG. 15 shows a configuration example of the library management system 1a according to this embodiment. The configuration is the same as that of the server 10 described in FIG. 3 except that the server 10 has an analysis application development processing unit 315.

  FIG. 16 shows an example of an output screen displayed on the operation terminal 50 according to the present embodiment. The edit screen 70 is a flow editor screen that describes the flow of analysis processing. The edit screen 70 has, for example, a list field 71 in which a list of libraries as analysis processing components is displayed, and a flow field 72 that describes the execution order of the libraries (flow of analysis processing).

  The result of the recommendation process in this embodiment is displayed, for example, as a recommended result field 73 illustrated on the flow field 72. The user operates the adoption button 74 to apply the recommended library. As a result, according to this embodiment, the work of describing the data analysis processing flow can be facilitated.

  The processing of the recommendation processing unit 313 in this embodiment will be described with reference to FIG. The recommendation processing unit 313 of this embodiment acquires a change on the flow field 72 of the edit screen 70 as a trigger event (S31).

  The recommendation processing unit 313 acquires data from the flow information 328, the processing phase DB 323, the processing flow DB 324, and the similarity DB 325 as information of the analysis application displayed on the editing screen 70 (S32), Based on the data, one or a plurality of recommended target libraries estimated to be used next by the user who executed the library is calculated (S33). Finally, the recommendation processing unit 313 transmits the recommended library to the operation terminal 50 (S34).

  FIG. 17 shows an example of the flow information 328. The flow information 328 is information holding the content displayed in the flow field 72 of the analysis process of the edit screen 70. The flow information 328 is generated by the user operating the operation terminal 50.

  The flow information 328 has, for example, a library ID field 3281, a label field 3282, an input library ID field 3283, an X coordinate field 3284, and a Y coordinate field 3285.

  The library ID field 3281 is information that uniquely identifies which library the processing component icon arranged in the flow field 72 represents. The label field 3282 is information about a label used when displaying a library in the flow field 72.

  The input library ID field 3283 is information holding the input / output relationship of the data expressed in the flow field 72. The library ID field 3283 is information that identifies a processing component input to the entry, and stores zero or more elements.

  The X coordinate field 3284 is information for storing the X coordinate when the processing component icon is displayed on the flow field 72. The Y coordinate field 3285 is information that stores the Y coordinate when the processing component icon is displayed on the flow field 72.

  In step S331 of FIG. 13, the recommendation processing unit 313 compares the library ID field 3281 of the flow information 328 with the library ID field 3232 of the processing phase DB 323, and uses the value of the corresponding processing phase field 3231 as the library ID field 3281. Overwrite. Similarly, the recommendation processing unit 313 overwrites the input library ID field 3283 with the value of the corresponding processing phase field 3231 and outputs the flow information 328.

  The recommendation processing unit 313 uses the transition source field 3241 of the process flow DB 324 and the input library ID field 3283 of the flow information 328 as corresponding information, the transition destination field 3242 of the process flow DB 324 and the library ID field 3281 of the flow information 328. By comparing as corresponding information, a matching flow is calculated (s332). The recommendation processing unit 313 detects an entry in which the transition source field 3241 corresponding to the library ID field 3281 exists but the corresponding transition destination field 3242 does not exist, and outputs it as the next processing phase to be recommended (S332). .. Then, the recommendation processing unit 313 specifies and outputs the library to be recommended, as in the first embodiment (S333).

  For example, the entry of the library ID “2” in FIG. 17 corresponds to the phase P2 according to FIG. According to FIG. 8, the phase P2 and the phase P3 exist at the transition destination of the entry of the process flow DB 324 whose phase P2 is the transition source. Since the library corresponding to the phase P3 is not registered in the flow information 328 illustrated in FIG. 17, it is specified as a candidate for the processing phase to be executed next.

  According to FIG. 7, the libraries corresponding to the processing phase P3 are the library IDs “3” and “6”. The recommendation processing unit 313 determines a library having a library ID “3” belonging to the same group Gα as the library ID “2” as a target to be preferentially recommended based on the similarity shown in FIG. 9, and outputs it as a recommendation target.

  This embodiment, which is configured in this way, also exhibits the same effects as the first embodiment. Further, according to the present embodiment, the user who is developing the data analysis application via the operation terminal 50 is registered in the library DB 320 when developing the data analysis application using the library management system 1a. The optimum library group according to the processing flow can be efficiently selected from the libraries, and the data analysis work can be simplified.

  Furthermore, by extracting a recommended library using the processing flow DB 324 by an expert or expert, even a beginner of data analysis application development can develop an application by referring to the process flow of the expert or expert. it can.

  A third embodiment will be described with reference to FIGS. 18 and 19. In this embodiment, when the user analyzes the data using the data analysis application, the library management system 1b recommends the function relating to the data analysis library to be used next to the user.

  FIG. 18 shows a configuration example of the library management system 1b according to this embodiment. The configuration is the same as that of the server 10 described in FIG. 3 except that the server 10 has an analysis application execution processing unit 316.

  FIG. 19 shows an example of an output screen displayed on the operation terminal 50 according to the present embodiment. The display screen 80 is a screen of the data analysis application. FIG. 19 shows, for example, an analysis application having a KPI monitoring function area 81, a cause analysis 1 function area 82, a cause analysis 2 function area 83, an optimization 1 function area 84, and an optimization 2 function area 85. The example of a screen is shown.

  For example, the user can check the occurrence of a problem by looking at the KPI monitoring function area 81 and check the cause of the problem by using the cause analysis 1 function area 82 and the cause analysis 2 function area 83. .. Further, the user can make a plan for solving the problem by checking the optimization 1 functional area 84 and the optimization 2 functional area 85.

  In the present embodiment, when a problem occurrence in the KPI monitoring function area 81 is detected, a function (library) to be used next is specified, the specified library is highlighted, and the user is prompted to use the function. As a result, in this embodiment, the analysis work using the data analysis application can be facilitated. For example, on the display screen 80 illustrated in FIG. 19, the cause analysis 1 functional area 82 is highlighted, and the user can easily determine where on the display screen 80 to operate.

  The flow of processing of the recommendation processing unit 313 in this embodiment will be supplemented. When the user activates the analysis application from the operation terminal 50, the analysis application execution processing unit 316 is executed on the server 10.

  The analysis application execution processing unit 316 activates the library execution processing unit 314 in order to execute the library used in the analysis application. As a result, the recommendation processing unit 313 described in the first embodiment is executed, and the library group to be executed next is output.

  When the recommendation processing unit 313 of the present embodiment calculates the library group to be executed next, it determines the correspondence with the library used in the analysis application, and the function of the analysis application including the library group to be executed next. Identify (library) and highlight the display for the identified function. For example, in addition to displaying the frame thick, the background can be highlighted by means such as changing the background color by animation. Since the user can easily recognize the highlighted function (library) as the library to be used next, the data analysis work using the analysis application can be facilitated.

  For example, in the screen illustrated in FIG. 19, the library with the library ID “1” is used in the KPI monitoring function area 81, and the library with the library ID “2” is used in the cause analysis 1 function area 82. The library having the library ID “3” is used in the function 1 functional area 84, the library having the library ID “4” is used in the cause analysis 2 function area 82, and the library ID “5” is used in the optimization 2 function area 84. ] Library is being used.

  In this case, from FIG. 7, the KPI monitoring function area 81 corresponds to the phase P1, the cause analysis 1 function area 82 and the cause analysis 2 function area 83 correspond to the phase P2, and the optimization 1 function area 84 and the optimization 2 It can be seen that the functional area 85 corresponds to the phase P3. Thereby, also in the present embodiment, the function (library) to be executed next can be specified.

  According to the present embodiment, the user who is analyzing the data using the data analysis application via the operation terminal 50, when performing the data analysis using the library management system 1b, a plurality of users installed in the analysis application. It becomes possible to select the most suitable function according to the processing flow from the function (library), and the data analysis work can be simplified.

  It should be noted that the present invention is not limited to the above-described embodiments, but includes various modifications. For example, the above embodiments are described in detail for the purpose of explaining the present invention in an easy-to-understand manner, and are not necessarily limited to those having all the configurations described.

  Further, each of the above-described configurations, functions, processing units, processing means, and the like may be realized in hardware by designing a part or all of them with, for example, an integrated circuit. Further, the above-described respective configurations, functions, processing units, processing means, and the like may be realized in a cloud system by designing a part or all of them, for example, with a virtual machine.

  Further, each of the above-described configurations, functions, and the like may be realized by software by a processor interpreting and executing a program that realizes each function. Information such as a program, a table, and a file that realizes each function can be placed in a memory, a hard disk, a recording device such as an SSD (Solid State Drive), or a recording medium such as an IC card, an SD card, and a DVD.

  Further, the control lines and information lines are shown as being considered necessary for explanation, and not all the control lines and information lines in the product are necessarily shown. In practice, it may be considered that almost all configurations are connected to each other.

  Further, each constituent element of the present invention can be arbitrarily selected, and an invention having a selected configuration is also included in the present invention. Furthermore, the configurations described in the claims can be combined in addition to the combinations specified in the claims.

Furthermore, the above-described embodiment can be expressed as a computer program as follows.
"A computer program that causes a computer to function as a software component management system that searches for software components,
A classification processing unit that classifies each software component into a predetermined component category,
Based on the execution result information indicating the execution results of each software component and the classification result of each software component by the classification processing unit, the execution order of the part classification to which each software component recorded in the execution result information belongs is detected. An execution order processing unit that
Based on the classification result and the execution order, a recommendation target component classification that can be executed next to the component classification to which the usage-requested software component belongs is detected, and the detected recommendation target component classification is detected. A recommendation processing unit that detects the software component to which it belongs and that outputs the detected software component to the usage request source;
A computer program for realizing. "

  The computer program described above can be fixed and distributed in a computer-readable recording medium such as a semiconductor memory, an optical disk, or a hard disk. Further, the above computer program can be transferred between separate devices physically separated by using a private line or a public line.

  1, 1a and 1b: library management system, 50: operating terminal, 310: processing phase classification processing unit, 311: processing flow extraction processing unit, 312: similarity calculation processing unit, 313: recommendation processing unit, 314: library execution function , 320: library DB, 321: library execution record DB, 322: business data, 323: processing phase DB, 324: processing flow DB, 325: similarity DB

Claims (13)

A software component management system for retrieving software components,
A classification processing unit that classifies each software component into a predetermined component category,
Based on the execution result information indicating the execution results of each software component and the classification result of each software component by the classification processing unit, the execution order of the part classification to which each software component recorded in the execution result information belongs is detected. An execution order processing unit that
Based on the classification result and the execution order, a recommendation target component classification that can be executed next to the component classification to which the usage-requested software component belongs is detected, and the detected recommendation target component classification is detected. A recommendation processing unit that detects the software component to which it belongs and that outputs the detected software component to the usage request source;
A software component management system equipped with. The classification processing unit, based on at least one of an input / output relationship for each software component, a characteristic of each software component, and a characteristic of data input / output to / from each software component, Classify which software category a software component belongs to,
The software component management system according to claim 1. The recommendation processing unit uses the characteristics of the data output by the usage-requested software component as an explanatory variable to calculate in which of the execution sequences the usage-requested software component is located,
The software component management system according to claim 2. The recommendation processing unit outputs, to the issuer of the usage request, of the detected software components, a software component that shares a predetermined attribute with the software component requested to be used is a higher-ranking candidate.
The software component management system according to claim 1. The execution order information includes a transition source part category to which the transition source software part belongs, a transition destination part category to which the transition destination software part belongs, and a transition condition for transitioning from the transition source part category to the transition destination part category. Has been recorded,
The execution order processing unit uses, as an explanatory variable, the characteristics of the data output by the software component belonging to the transition source component division when there are transitions from one transition source component division to different transition destination component divisions. By using, to calculate the probability of transition to the different transition destination part classification,
The software component management system according to claim 1. The classification processing unit is configured to register a new software component, update any existing software component, or update a setting for using the existing software component. When one occurs, the software components are classified into the predetermined component classification,
The software component management system according to claim 1. The execution order processing unit is configured to perform the predetermined execution based on predetermined execution result information executed by a predetermined user in the execution result information and a classification result of each software component by the classification processing unit. Detects the execution order of the parts category to which each software part belongs recorded in the execution record information of
The software component management system according to claim 1. The predetermined attribute detects a part category to which a software part with a history of being used in the past by the usage request source of the usage-requested software part belongs, as the part category to be recommended,
The software component management system according to claim 1. The software component is a data analysis library used for business improvement,
The software component management system according to claim 1. A computer that functions as a software component management system that searches for software components,
An arithmetic unit,
A storage unit for storing a predetermined computer program executed by the arithmetic unit,
The arithmetic unit, by executing the predetermined computer program,
A classification processing unit that classifies each software component into a predetermined component category,
Based on the execution result information indicating the execution results of each software component and the classification result of each software component by the classification processing unit, the execution order of the part classification to which each software component recorded in the execution result information belongs is detected. An execution order processing unit that
Based on the classification result and the execution order, a recommendation target component classification that can be executed next to the component classification to which the usage-requested software component belongs is detected, and the detected recommendation target component classification is detected. A recommendation processing unit that detects the software component to which it belongs and that outputs the detected software component to the usage request source;
A calculator that realizes. The execution order information includes a transition source part category to which the transition source software part belongs, a transition destination part category to which the transition destination software part belongs, and a transition condition for transition from the transition source part category to the transition destination part category. Has been recorded,
The execution order processing unit uses, as an explanatory variable, the characteristics of the data output by the software component belonging to the transition source component division when there are transitions from one transition source component division to different transition destination component divisions. By using, to calculate the probability of transition to the different transition destination parts classification,
The computer according to claim 10. The execution order processing unit is configured to perform the predetermined execution based on predetermined execution performance information executed by a predetermined user in the execution performance information and a classification result of each software component by the classification processing unit. Detects the execution order of the part category to which each software part belongs recorded in the execution record information of
The computer according to claim 10. The predetermined attribute detects a part category to which a software part with a history of being used in the past by the usage request source of the usage-requested software part belongs, as the part category to be recommended,
The computer according to claim 10.

Images