Information processing system and information processing method

Abstract

To estimate meaning of a data item of site data for creating a process model.SOLUTION: An information processing system: holds start / completion actual result data which is selected from site data related to process including a plurality of tasks, and in which a data item indicating predetermined meaning is designated, and meaning estimation object data selected from the site data; estimates a candidate of the data item indicating the predetermined meaning, from the data item of the meaning estimation object data, on the basis of a feature amount of the data item indicating the predetermined meaning of the start / completion actual result data, and the feature amount of each data item included in the meaning estimation object data; and estimates the data item indicating the predetermined meaning from the candidate, on the basis of a comparison result between the start / completion actual result data, and the meaning estimation object data when the candidate is the data item indicating the predetermined meaning.SELECTED DRAWING: Figure 2A

Description

[Technical Field]

[0001] The present invention relates to an information processing system and an information processing method. [Background technology]

[0002] In manufacturing, when a task is completed or a specific event occurs, on-site data is generated and collected. On-site workers analyze this on-site data to improve work efficiency.

[0003] When different departments manage each operation, it can be difficult to share information between them due to the fact that each department has its own system for managing operations. While each department tries to improve its own work efficiency by analyzing the on-site data it collects, collaboration between departments is essential to improve the work efficiency of the entire manufacturing process. Collaboration between departments can be promoted by presenting each department with not only the on-site data collected within its own department, but also on-site data generated throughout the entire manufacturing process in a mutually correlated manner.

[0004] The above is not unique to manufacturing sites, but is also true in industries that have a series of processes consisting of multiple operations, such as logistics, retail, or service industries.

[0005] Background art in this technical field includes Japanese Patent Application Laid-Open No. 2022-20927 (Patent Document 1) and International Publication No. 2018 / 025707 (Patent Document 2). Patent Document 1 discloses a process model creation unit that "creates a process model that indicates the relationship between the tasks that make up a process and the on-site data, in a process consisting of multiple steps, based on on-site data generated at a manufacturing site."

[0006] Furthermore, Patent Document 2 discloses that "the learning means learns a model that indicates regularity between the distribution of attribute values ​​according to the meanings of columns in a table and the meaning of the table, based on a table including the meanings of columns and learning data including the meanings of the table. The estimation means estimates the meaning of the table based on the distribution of attribute values ​​according to the meanings of columns in the input table and the model." [Prior art documents] [Patent documents]

[0007] [Patent Document 1] Japanese Patent Publication No. 2022-20927 [Patent Document 2] International Publication No. 2018 / 025707 Summary of the Invention [Problem to be solved by the invention]

[0008] The process model creation unit disclosed in Patent Document 1 creates a process model that defines the data structure of on-site data from the on-site data generated from operations included in the manufacturing process, but it is assumed that the meanings of the data items in the on-site data are known, and Patent Document 1 does not disclose a technology for creating a process model using on-site data whose data items have unknown meanings.

[0009] In order to create a process model that defines the data structure of on-site data generated in a manufacturing process from the on-site data, the technology disclosed in Patent Document 1 requires extracting information from the on-site data regarding when and where work was performed on the manufacturing object in the manufacturing process.

[0010] Here, with regard to on-site data generated during the manufacturing process, information relating to identifiers that identify the manufacturing object, information indicating timing such as time, and information relating to identifiers that identify the work may generally be recorded in multiple locations.

[0011] For example, in a table of work site data, the instruction ID that identifies the work object, the product ID that identifies the processing object, and the packaging ID that identifies the transport object may each be recorded in a separate column, the area ID that identifies the work location and the process ID that identifies the work process may each be recorded in a separate column, and the completion time that indicates the time the process was completed and the recording time that indicates the time the data was recorded may each be recorded in a separate column.

[0012] Patent Document 2 discloses a technology that "learns a model showing regularities between the meanings of table columns and the meanings of the table based on learning data including the meanings of the table columns and the meanings of the table, and estimates the meaning of the table based on the meanings and model of the input table columns," but does not disclose a technology that estimates the meaning of field data in order to create a process model from the field data.

[0013] Therefore, one aspect of the present invention estimates the meaning of data items of field data for creating a process model. [Means for solving the problem]

[0014] In order to solve the above problems, one aspect of the present invention employs the following configuration: An information processing system includes a processor and a memory, and the memory stores arrival / completion performance data selected from field data related to a process including a plurality of tasks and having data items indicating predetermined meanings designated therein; The processor holds meaning inference target data selected from the on-site data, feature amounts of data items that indicate the predetermined meaning of the delivery completion record data, and feature amounts of each data item included in the meaning inference target data, and the processor infers candidates for data items that indicate the predetermined meaning from the data items of the meaning inference target data based on the feature amounts of the data items that indicate the predetermined meaning of the delivery completion record data and the feature amounts of each data item included in the meaning inference target data, executes a comparison process that compares the delivery completion record data with the meaning inference target data when the candidates are data items that indicate the predetermined meaning, and infers the data item that indicates the predetermined meaning from the candidates based on the results of the comparison process. [Effects of the Invention]

[0015] According to one aspect of the present invention, it is possible to estimate the meaning of data items of field data for creating a process model.

[0016] Problems, configurations, and effects other than those described above will become apparent from the following description of the embodiments. [Brief explanation of the drawings]

[0017] [Figure 1] 1 is a block diagram showing an example of the configuration of a management system including a process model creating system according to a first embodiment. [Figure 2A] 1 is a block diagram illustrating an example of a functional configuration of a process model creating system according to a first embodiment. [Figure 2B] 1 is a block diagram illustrating an example of a hardware configuration of a process model creating system according to a first embodiment. [Figure 3] FIG. 10 is a sequence diagram illustrating an example of an operation procedure from creation to provision of data semantic information in the first embodiment. [Figure 4A] FIG. 2 is a diagram illustrating an example of a data configuration of on-site data in the first embodiment. [Figure 4B] FIG. 4 is a diagram illustrating an example of information included in the site data according to the first embodiment. [Figure 5]FIG. 2 is an explanatory diagram illustrating an example of a relationship between a manufacturing process and a business in the first embodiment. [Figure 6] FIG. 2 is an explanatory diagram showing an example of a process model related to the manufacturing process in the first embodiment. [Figure 7] FIG. 3 is a diagram illustrating an example of a data configuration of data semantic information according to the first embodiment. [Figure 8] 4 is a flowchart illustrating an example of a part of processing executed by the process model creating system according to the first embodiment. [Figure 9] 1 is a flowchart illustrating an example of a process flow creation process in the first embodiment. [Figure 10A] FIG. 2 is a diagram showing an example of a process flow before integration in the first embodiment. [Figure 10B] FIG. 2 is a diagram showing an example of a process flow before integration in the first embodiment. [Figure 10C] FIG. 1 is a diagram showing an example of an integrated process flow in Example 1. [Figure 11] 10 is a flowchart illustrating an example of a data meaning candidate estimation process according to the first embodiment. [Figure 12A] FIG. 2 is a diagram illustrating an example of the data configuration of a table that is a target of data meaning estimation in the first embodiment. [Figure 12B] FIG. 3 is a diagram illustrating an example of the data configuration of data meaning candidate information in the first embodiment. [Figure 13] 10 is a flowchart illustrating an example of a data meaning estimation process according to the first embodiment. [Figure 14] FIG. 10 is an explanatory diagram showing an example of a process for calculating the degree of compatibility with the process flow in the first embodiment. [Figure 15] FIG. 10 is an explanatory diagram showing an example of a process for calculating the degree of compatibility with the process flow in the first embodiment. [Figure 16] FIG. 10 is an explanatory diagram showing an example of a process for calculating the degree of compatibility with the process flow in the first embodiment. [Figure 17] FIG. 10 is an explanatory diagram showing an example of a process for calculating the degree of compatibility with the process flow in the first embodiment. [Figure 18A]FIG. 10 is a diagram showing an example of a screen configuration of a delivery completion record data setting screen in the first embodiment. [Figure 18B] FIG. 2 is a diagram showing an example of a screen configuration of a process flow display screen in the first embodiment. [Figure 19A] FIG. 10 is a diagram illustrating an example of a screen configuration of a data meaning estimation target setting screen in the first embodiment. [Figure 19B] FIG. 10 is a diagram showing an example of a screen configuration of a data meaning estimation result display screen in the first embodiment. [Figure 20] FIG. 11 is an explanatory diagram illustrating an example of a process for estimating the meaning of a table from table data meaning candidate information according to the second embodiment. [Figure 21] FIG. 11 is an explanatory diagram illustrating an example of a process of estimating data semantic information from a process model according to the third embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0018] Hereinafter, embodiments of the present invention will be described with reference to the drawings. Note that the embodiments described below do not limit the scope of the invention as claimed, and not all of the elements and combinations thereof described in the embodiments are necessarily essential to the solution of the invention.

[0019] In this embodiment, an example will be mainly described in which the process model creation system is applied to an information collection system that collects field data generated in each manufacturing process in a manufacturing factory and / or in the operations within the manufacturing process. However, the system can also be applied in the same way to information collection systems used in processes having multiple consecutive operations or steps in any type of industry, such as logistics, retail, or service.

[0020] In addition, in this embodiment, the data format of the site data is described as an example of a table consisting of one or more columns, but other data formats such as JSON (JavaScript (registered trademark) Object Notation) may also be used. [Example]

[0021] [network] 1 is a block diagram showing an example of the configuration of a management system including a process model creation system 110. The process model creation system 110 is connected to a network 120 such as the Internet. Also connected to the network 120 are one or more data generating devices 130, one or more on-site data storage units 140, and an information collection system 150.

[0022] In addition, an operation terminal 160 for operating the process model creation system 110 is connected to the process model creation system 110. The operation terminal 160 is a computer having an input device and a display device, and is, for example, a PC (Personal Computer), a smartphone, or a tablet terminal.

[0023] The data generating device 130 may be, for example, a barcode reader, a PC, or a server that acquires the work logs of workers, a machine that processes parts or assembles finished products, or a sensor that collects inspection information from RFID (Radio Frequency Identifier) ​​attached to parts or finished products. The data generating device 130 transmits the site data that it has collected or generated to the site data accumulation unit 140 via the network 120.

[0024] The site data accumulation unit 140 is, for example, a storage device such as a server or a memory, and accumulates the site data received from the data generation device 130. The site data indicates information relating to a process including a plurality of tasks.

[0025] The information collection system 150 collects information related to the implementation of a business. The information related to the implementation of a business includes, for example, an implementation record related to the implementation of the business. The implementation record related to the implementation of the business may include business information related to the implementation of the business and business-related information related to the business. For example, the business-related information is information indicating at least one of objects, people, equipment, procedures, etc. related to the implementation of the business.

[0026] The information collection system 150 holds a process model that reproduces in a digital space the business process of the manufacturing site where the data generating device 130 is located. The process model is a model that represents what workers, machines, parts, etc. are involved in what work procedures for each task included in the process. The process model includes metadata of the data generated by the data generating device 130 and stored in the site data storage unit 140, and the information collection system 150 can collect data from the site data storage unit 140 using the process model.

[0027] Note that some or all of business operator A that owns the process model creation system 110, business operator B that owns the information collection system 150, and business operators C1 and C2 that own the on-site data accumulation unit 140 may be the same. Also, some or all of the process model creation system 110, data generating device 130, on-site data accumulation unit 140, and information collection system 150 may be located in the same place. Also, some or all of the process model creation system 110, data generating device 130, on-site data accumulation unit 140, information collection system 150, and operation terminal 160 may be integrated into a single device.

[0028] [System Configuration] 2A is a block diagram showing an example of the functional configuration of the process model creation system 110. The process model creation system 110 includes, for example, a data setting unit 210, a process flow creation unit 220, a data meaning candidate estimation unit 230, a data meaning setting unit 240, and a process model creation unit 250, which are all functional units.

[0029] As described above, the process model creation system 110 is connected to the operation terminal 160, and further connected to the field data storage unit 140 via the network 120. The operation terminal 160 provides the user with an environment for inputting predetermined information into the data setting unit 210 and the data meaning setting unit 240.

[0030] The data setting unit 210 sets the delivery completion record data as site data selected from the site data stored in the site data storage unit 140. The process flow creation unit 220 creates a process flow from the delivery completion record data.

[0031] The data meaning candidate estimation unit 230 sets field data selected from the field data accumulated in the field data accumulation unit 140 as data to be estimated for data meaning estimation, and estimates column candidates corresponding to a specific meaning (data meaning) from among the columns (examples of data items) included in the data. The data meaning setting unit 240 determines which column corresponds to the specific meaning from the column candidates corresponding to the specific meaning in the data to be estimated for data meaning estimation, i.e., sets the data meaning. The process model creation unit 250 creates a process model using the process flow created by the process flow creation unit 220 and the data meaning set by the data meaning setting unit 240, and provides the process model to the information collection system 150.

[0032] 2B is a block diagram showing an example of the hardware configuration of the process model creation system 110. The process model creation system 110 is configured by a computer having, for example, a CPU (Central Processing Unit) 1001, a memory 1002, an auxiliary storage device 1003, an input device 1004, a display device 1005, and a communication device 1006.

[0033] The CPU 1001 is an example of a processor, and executes programs stored in the memory 1002. The memory 1002 includes a ROM (Read Only Memory), which is a non-volatile storage element, and a RAM (Random Access Memory), which is a volatile storage element. The ROM stores unchanging programs (e.g., a BIOS (Basic Input / Output System)). The RAM is a high-speed, volatile storage element such as a DRAM (Dynamic Random Access Memory), and temporarily stores programs executed by the CPU 1001 and data used when the programs are executed.

[0034] The auxiliary storage device 1003 is a large-capacity, non-volatile storage device such as a magnetic storage device (HDD (Hard Disk Drive)) or a flash memory (SSD (Solid State Drive)), and stores programs to be executed by the CPU 1001 and data to be used when the programs are executed. That is, the programs are read from the auxiliary storage device 1003, loaded into the memory 1002, and executed by the CPU 1001.

[0035] The input device 1004 is a device that receives input from the user, such as a keyboard or a mouse. The display device 1005 is a device that outputs the results of program execution in a format that can be viewed by the user, such as a display device or a printer.

[0036] The communication device 1006 is a network interface device that controls communication with other devices in accordance with a predetermined protocol, and may also include a serial interface such as a USB (Universal Serial Bus).

[0037] A part or all of the program executed by the CPU 1001 may be provided to the process model creation system 110 from a removable medium (such as a CD-ROM or flash memory) which is a non-transitory storage medium, or from an external computer equipped with a non-transitory storage device via a network, and may be stored in a non-volatile auxiliary storage device 1003 which is a non-transitory storage medium.

[0038] The process model creation system 110 is a computer system that is configured on one physical computer or on multiple logically or physically configured computers, and may operate in separate threads on the same computer, or may operate on a virtual computer built on multiple physical computer resources.

[0039] The CPU 1001 includes each functional unit included in the process model creation system 110. For example, the CPU 1001 functions as a data setting unit 210 by operating in accordance with a data setting program loaded into the memory 1002, and functions as a process flow creation unit 220 by operating in accordance with a process flow creation program loaded into the memory 1002. This also applies to the other functional units included in the process model creation system 110.

[0040] Note that some or all of the functions of the functional units included in the process model creating system 110 may be realized by hardware such as an ASIC (Application Specific Integrated Circuit) or an FPGA (Field-Programmable Gate Array).

[0041] The information held by each device included in the management system is not dependent on the data structure and may be represented in any data structure, for example, a list, a table, a database, or a queue, as appropriate.

[0042] [Operation procedure] 3 is a sequence diagram showing an example of an operation procedure from creation to provision of data semantic information. The operation terminal 160 accepts an instruction to acquire on-site data according to an input from a user (hereinafter simply referred to as a user) who uses the process model creation system 110, and transmits the instruction to the process model creation system 110 (S301). The on-site data acquisition instruction includes, for example, conditions related to the on-site data to be acquired (for example, the time period and location of the on-site data).

[0043] The data setting unit 210 of the process model creation system 110 acquires site data (table) from the site data storage unit 140 in accordance with the conditions related to the site data to be acquired, and presents the acquired data to the user by transmitting it to the operation terminal 160 and displaying it (S302).

[0044] The operation terminal 160 executes the process of selecting one or more tables from the tables presented in step S302 and the process of selecting columns in the tables that indicate predetermined meanings (for example, a column corresponding to a manufacturing object ID, a column corresponding to a task ID, and a column corresponding to a task execution time, which will be described later), in accordance with the user's input, and then transmits an arrival / completion record data setting instruction indicating the selected tables and the columns corresponding to the predetermined items to the process model creation system 110 (S303). Note that the column corresponding to the manufacturing object ID is an example of a tracking object identification item that indicates information that identifies a tracking object in a process. The column corresponding to the task ID is an example of a tracking point identification item that indicates information that identifies a tracking point in a process. The column corresponding to the task execution time is an example of a timing item that indicates the timing when the tracking object was present at the tracking point in the process.

[0045] The data setting unit 210 of the process model creating system 110 sets the arrival / completion record data in accordance with the table indicated by the arrival / completion record data setting instruction and the columns corresponding to the predetermined items (S304). The setting process of the arrival / completion record data will be described in detail later.

[0046] Next, the operation terminal 160 selects one or more tables from the field data (tables) presented in step S302, excluding the table selected in step S303, according to the user's input, generates a data meaning estimation instruction indicating the selected table, and transmits the instruction to the process model creation system 110 (S305).

[0047] The data setting unit 210 of the process model creation system 110 acquires the table indicated by the data meaning estimation instruction from the field data accumulation unit 140, and the data meaning candidate estimation unit 230 and the data meaning setting unit 240 estimate data meaning information for the acquired table based on the arrival / completion performance data information set in step S304, and present the estimation result to the user by transmitting it to the operation terminal 160 and displaying it (S306).

[0048] The operation terminal 160 checks the estimation result of the data semantic information presented in step S306 in accordance with the input of the user who uses the process model creation system 110, generates an instruction indicating whether or not to create a process model using the data semantic information indicated by the estimation result, and transmits the instruction to the process model creation system 110 (S307). Note that the operation terminal 160 may also modify the data semantic information indicated by the estimation result in accordance with the input of the user, and in this case, includes the modified data semantic information in the instruction indicating whether or not to create a process model.

[0049] When the process model creation unit 250 of the process model creation system 110 receives an instruction to create a process model, it creates a process model from the data semantic information indicated by the estimation result, and presents the process model creation result to the operation terminal 160 by displaying it (S308).

[0050] The operation terminal 160 confirms the process model presented in step S307 in accordance with an instruction from the user who uses the process model creation system 110, and transmits to the process model creation system 110 an instruction indicating whether or not to provide the presented process model to the information collection system 150 (S309). Note that the operation terminal 160 may also modify the process model in accordance with the input from the user, and in this case, the modified process model is included in the instruction indicating whether or not to provide the process model to the information collection system 150.

[0051] When the process model creating unit 250 of the process model creating system 110 receives an instruction to provide the process model to the information collecting system 150, the process model creating unit 250 provides the process model to the information collecting system 150 (S310).

[0052] As long as the process of estimating data semantic information for the field data acquired from the field data storage unit 140, the process of creating a process model from the estimated data semantic information, and the process of providing the created process model to the information collection system 150 are executed, the execution order of the processes of steps S301 to S310 may be partially reversed.

[0053] [Field data content] FIG. 4A is a diagram showing an example of the data structure of the site data. The site data storage unit 140 stores multiple tables of the site data, and table 410 is one of the multiple tables, and the name of table 410 is assumed to be "Table A." As shown in table 410 in FIG. 4A, each table of the site data includes one or more columns. Table 410 includes columns col1 to col7. However, data meanings may not be assigned to some or all of the columns in each table of the site data (i.e., it may be unclear what information the column indicates).

[0054] 4B is a diagram showing an example of information included in the site data. Each table of the site data may, for example, hold some or all of the types of information shown in table 420, or may not hold any of the types of information, or may hold information by main category that is not shown in table 420. Note that table 420 is described to explain an example of information that may be included in the site data in this embodiment, and therefore does not need to be stored in the site data accumulation unit 140.

[0055] The information types indicated by table 420 include, for example, a product identifier, a product type identifier, a work identifier, a work type identifier, work date and time, recording date and time, a part identifier, a work-related information identifier, and a work result. A product identifier is information that uniquely identifies a product. A product type identifier is information that uniquely identifies a product type. A work identifier is information that uniquely identifies a work. A work type identifier is information that uniquely identifies a work type. Table 420 describes the data type of each type of information.

[0056] The work date and time is an example of information indicating when the work was performed. Furthermore, the information indicating when the work was performed may include multiple types of dates and times, such as the start date and time and the end date and time of the work. The recording date and time is an example of information indicating when the data was recorded.

[0057] The part identifier is information that uniquely identifies the part used to create the product. If the part and the product are identified as the same material before and after the work, the value of the product identifier may be the same as the value of the part identifier. Also, if a product is created using multiple parts, multiple part identifiers may be registered in one record (or one cell) of the site data table. Also, the work-related information identifier is information that uniquely identifies information related to the work. The work result is information that indicates the result of the work.

[0058] 4A, col1 corresponds to the product identifier, col2 corresponds to the product type identifier, col3 corresponds to the work identifier, col4 corresponds to the work type identifier, col5 corresponds to the work date and time, col6 corresponds to the work result, and col7 corresponds to the part identifier. In the following explanation, we will take as an example a case where time information is used as information indicating the timing of the work included in the work site data.

[0059] [Process Configuration] 5 is an explanatory diagram showing an example of the relationship between a manufacturing process and operations. A manufacturing process includes one or more operations that produce a product. In the example of FIG. 5, process 510 includes, for example, operations 520a and 520b.

[0060] [Process model configuration] 6 is an explanatory diagram showing an example of a process model related to a manufacturing process. The process model 610 is generated by associating on-site data generated from various devices (data generating device 130) related to each task included in the process, and is a model that shows what workers, machines, parts, etc. are involved in each task included in the process, and in what work procedures and how they are related.

[0061] The process model 610 is represented by a directed graph including, for example, a business node (a node labeled "Business" in the diagram), a material node (a node labeled "Material" in the diagram), and a business-related information node (a node labeled "Related Information" in the diagram).

[0062] A business node is a node associated with a business, and nodes 620, 621, and 622 are all examples of business nodes. A material node is a node associated with a material. Material nodes include part nodes, which are nodes associated with materials needed to perform a business, and product nodes, which are associated with products produced as a result of performing a business using parts as materials. Nodes 630 and 650 are both examples of material nodes, with node 630 corresponding to a part node and node 650 corresponding to a product node.

[0063] For example, suppose that the product associated with node 650, which is a product node, is a material (part) used in a subsequent task associated with node 621, and the material associated with node 630, which is a part node, is a product produced in a previous task associated with node 622, which is a task node. In this case, node 620, which is a task node, and node 622, which is a task node, are associated via node 630, which is a part node, and node 620, which is a task node, and node 621, which is a task node, are associated via node 650, which is a product node. In other words, the attributes of the material associated with node 630, which is a part node, and the material associated with node 650, which is a product node, have the same meaning, and if there is no need to particularly distinguish between parts and products, both may be treated as simple material nodes.

[0064] A task-related information node is a node associated with a worker who performs the task, a machine used to perform the task, and / or a task procedure that defines the procedure for performing the task; node 640 is an example of a task-related information node.

[0065] In addition, material nodes including part nodes and product nodes, and business-related information nodes 640 associated with workers (Man), equipment (Machine), and / or work procedures (Method), etc., are sometimes referred to as 4M information or 4M nodes.

[0066] The business-related information is not limited to 4M information, but may be, for example, 5M information including information on measurements performed in the business, or 1M to 3M information that does not include one or more Ms in the 4M information, or may include information on the quality of the products for which the business was performed. Furthermore, the business-related information may include information on the environment in which the business was performed and / or information on the management status.

[0067] For example, an edge starting from a material node is connected to a business node, an edge starting from a business node is connected to a material node, and an edge starting from a business-related information node is connected to a business node. Also, for example, edges starting from multiple material nodes may each be connected to one business node, edges starting from multiple business-related information nodes may each be connected to one business node, or multiple edges may exist starting from one business node (each of the multiple edges is connected to a different material node).

[0068] [Data Semantics] FIG. 7 is a diagram showing an example of the data structure of data semantic information. Data semantic information 720 shows a column corresponding to a manufacturing object ID, a column corresponding to a task ID, and a column corresponding to a task execution time in each of one or more tables of the site data. The manufacturing object ID is information that identifies the manufacturing object. The manufacturing object is synonymous with material, and includes parts and products. The task ID is information that identifies the tasks that make up the manufacturing process. Task nodes included in the process model are distinguished by task ID. The task execution time is information that indicates the timing when a task is performed on a manufacturing object. Any time during the execution of a task, including the task start (commencement) time and the task completion time, is an example of the task execution time.

[0069] In the example of Figure 7, data semantic information 720 indicates that in "Table A" of the site data (i.e., table 410 shown in Figure 4A), col1 corresponds to the manufacturing object ID, col3 corresponds to the business ID, and col5 corresponds to the business execution time.

[0070] Note that some tables of the site data may not include all or some of the columns corresponding to the manufacturing object ID, the task ID, and the task execution time. In the following explanation, information about one table in the data semantic information 720 is also referred to as table data semantic information. For example, the information stored in cell 721 in the data semantic information 720 in FIG. 7 is table data semantic information about "Table A."

[0071] [Data Semantic Inference Processing] 8 is a flowchart showing an example of a part of the processing executed by the process model creating system 110. The processing shown in FIG. 8 corresponds to the processing from step S303 to step S306 shown in FIG.

[0072] The data setting unit 210 receives an instruction to set actual delivery completion data from the operation terminal 160, and sets actual delivery completion data in accordance with the received instruction to set actual delivery completion data (S801). As described above, the instruction to set actual delivery completion data includes information indicating one or more tables, and information indicating, in each of the one or more tables, the columns corresponding to the production object ID, the column corresponding to the task ID, and the column corresponding to the task execution time.

[0073] Therefore, in step S801, the data setting unit 210 refers to the arrival / completion actual data setting instruction, acquires one or more tables from the site data accumulation unit 140, and assigns information indicating the columns corresponding to the manufacturing object ID, the job ID, and the job execution time to each of the acquired one or more tables, thereby setting each of the one or more tables as arrival / completion actual data.

[0074] Here, the delivery completion record refers to the record that occurs from the start of a task to its completion, and the delivery completion record data is a table that includes at least the columns of the manufacturing object ID, the task ID, and the task execution time.

[0075] In step S801, for example, col1 of table 410 of the site data shown in FIG. 4A is determined to be the manufacturing object ID, col3 is determined to be the task ID, and col5 is determined to be the task time information, and table 410 is set as arrival and completion performance data.

[0076] The process flow creation unit 220 creates a process flow (S802) based on the delivery completion record data set by the data setting unit 210. Details of the process flow creation processing in step S802 will be described later with reference to FIG.

[0077] Here, a process flow refers to a sequence of business nodes and material nodes that constitute a process model. In other words, a process flow represents information on the path that a material (an example of a tracing target) takes through a business (an example of a tracing location). A process flow differs from a process model in that a process flow does not include business-related information nodes, whereas a process model may include business-related information nodes. In Figure 6, the sequence of node 622, which is a business node, node 630, which is a material node, node 620, which is a business node, and node 650, which is a material node, is an example of a process flow.

[0078] The data setting unit 210 receives a data meaning estimation instruction from the operation terminal 160, and acquires one or more tables of site data (different from the arrival / completion record data) indicated by the data meaning estimation instruction from the site data accumulation unit 140. The data meaning candidate estimation unit 230 estimates which column is a candidate for the manufacturing object ID, which column is a candidate for the task ID, and which column is a candidate for task time information for each of the acquired one or more tables, based on the arrival / completion record data set by the data setting unit 210 (S803). Details of the data meaning candidate estimation process in step S803 will be described later with reference to FIG. 11 etc.

[0079] Information indicating the column candidates corresponding to the manufacturing object ID, the column candidates corresponding to the task ID, and the column candidates corresponding to the task time information estimated for each table of the site data is also called data meaning candidate information. Information related to one table in the data meaning candidate information is also called table data meaning candidate information.

[0080] The data meaning setting unit 240 infers the data meaning from the data meaning candidate information based on the data meaning candidate information inferred by the data meaning candidate inference unit 230 and the process flow created by the process flow creation unit 220 (S804). Details of the data meaning inference process in step S804 will be described later with reference to FIG. 13 etc.

[0081] [Process flow creation details] 9 is a flowchart showing an example of the process flow creation processing in step S802. The process flow creation unit 220 executes the processing of steps S902 to S905 for each of all production object IDs included in the delivery completion record data (S901). That is, in step S901, the process flow creation unit 220 selects one unselected production object ID from the production IDs included in the delivery completion record data, and executes the processing of steps S902 to S905 for the selected production object ID, repeating this series of processing until it has been executed for all production object IDs.

[0082] The process flow creation unit 220 obtains, from the arrival / completion record data, a record including the selected manufacturing object ID, i.e., information indicating the work performed for the manufacturing object identified by the selected manufacturing object ID, as the manufacturing arrival / completion record for the manufacturing object ID (S902).

[0083] The process flow creation unit 220 sorts the production start / completion records obtained in the most recent step S902 using the business time information as a key (S903), and obtains the business IDs while maintaining the order in the records indicated by the sorted production start / completion records, thereby obtaining the permutation of the business IDs (i.e., information indicating the order in which the business operations were performed on the production object indicated by the production object ID) (S904).

[0084] The process flow creation unit 220 creates a process flow for the currently selected manufacturing object ID by arranging each of the business nodes distinguished by the business ID indicated by the permutation of the business IDs acquired in the most recent step S904 in the order indicated by the permutation, and connecting each of the business nodes via a material node (which does not distinguish materials regardless of the manufacturing object of the currently selected manufacturing object ID) (S905).

[0085] The process flow creation unit 220 creates a process flow that integrates the process flows for all the manufacturing object IDs included in the arrival / completion record data (S906), and ends the process flow creation process.

[0086] Note that when each of a plurality of tables is set as actual delivery / completion data, the process flow creation unit 220 executes the processing of steps S901 to S905 for each of the plurality of tables, and integrates the process flows for all of the manufacturing object IDs included in at least one of the plurality of tables in step S906. In other words, even when a plurality of tables are set as actual delivery / completion data, a single integrated process flow is created, just like when one table is set as actual delivery / completion data.

[0087] The process of integrating process flows in step S906 will be described with reference to Figures 10A, 10B, and 10C. Figures 10A and 10B are diagrams showing an example of the process flows before integration. Figure 10C is a diagram showing an example of the integrated process flow.

[0088] Fig. 10A shows a process flow 1010 with a manufacturing object ID of "mat_1". Fig. 10B shows a process flow 1020 with a manufacturing object ID of "mat_2". The process flow creation unit 220 created the process flows 1010 and 1020 in the processing of steps S901 to S905.

[0089] At this time, in step S906, the process flow creation unit 220 creates process flow 1030 by integrating process flow 1010 and process flow 1020. Specifically, for example, the process flow creation unit 220 may create an integrated process flow (here, process flow 1030) that includes all edges (edges are identified by combinations of start points and end points) and nodes of the process flows to be integrated (here, process flows 1010 and 1020), or may create an integrated process flow that includes edges and nodes that are included in a predetermined percentage or more of the process flows to be integrated. Note that when the process flows to be integrated include multiple process flows that are independent of each other (e.g., multiple process flows that do not include common edges or nodes), although each of the multiple independent process flows is a sequence of business nodes and material nodes, it is not possible to connect these sequences with edges when creating the integrated process flow. Therefore, the created integrated process flow includes multiple sequences of business nodes and material nodes that are not connected to each other by edges.

[0090] [Candidate Estimation Details] 11 is a flowchart showing an example of the data meaning candidate estimation process in step S803. The data meaning candidate estimation unit 230 refers to the delivery completion record data set in the data setting unit 210, calculates the feature amount of the column corresponding to the production object ID using data and the like in the cells included in the column corresponding to the production object ID, calculates the feature amount of the column corresponding to the business ID using data and the like in the cells included in the column corresponding to the business time information, and calculates the feature amount of the column corresponding to the business time information using data and the like in the cells included in the column corresponding to the business time information (S1101).

[0091] The data meaning candidate estimation unit 230 calculates features using at least one of the following: a statistical quantity such as the average value and / or variance of the character string lengths of the data included in the column for which features are to be calculated; the data type of the data included in the column for which features are to be calculated; the name of the column for which features are to be calculated; the name of the table that includes the column for which features are to be calculated; and a value that indicates the meaning of the data included in the column for which features are to be calculated as a character string.

[0092] Furthermore, it is desirable to use the same feature calculation algorithm to calculate the feature amounts of the column corresponding to the manufacturing object ID, the feature amounts of the column corresponding to the business ID, and the feature amounts of the column corresponding to the business time information.

[0093] The data meaning candidate estimation unit 230 executes the processes of steps S1103 to S1106 for each of one or more tables (i.e., tables that are data meaning estimation targets and are different from the arrival / completion actual data) acquired from the site data in accordance with the data meaning estimation instruction by the data setting unit 210 (S1102). That is, in step S1102, the data meaning candidate estimation unit 230 selects one unselected table from the tables that are data meaning estimation targets, and executes the processes of steps S1102 to S1106 for the selected table, and repeats this series of processes until it has been executed for all tables that are data meaning estimation targets.

[0094] The data meaning candidate estimation unit 230 executes the processes of steps S1104 to S1105 for each column included in the table selected in the most recent step S1102 (S1103). That is, in step S1103, the data meaning candidate estimation unit 230 selects one unselected column from the columns included in the table selected in the most recent step S1102, and executes the processes of steps S1104 to S1105 for the selected column, repeating this series of processes until it has been executed for all columns.

[0095] The data meaning candidate estimation unit 230 calculates the feature quantity of the column in the table selected in the most recent step S1102 using data, etc., from each cell of the column selected in the most recent step S1103. Note that the feature quantity calculation algorithm in step S1101 and the feature quantity calculation algorithm in step S1104 are the same.

[0096] The data meaning candidate estimation unit 230 calculates the similarity between the feature of the column corresponding to the manufacturing object ID of the delivery completion record data calculated in step S1101, the feature of the column corresponding to the task ID, and the feature of the column corresponding to the task execution time and the feature of the column extracted in the most recent step S1104 (S1105).

[0097] A known method may be used to calculate the similarity between feature quantities. Specifically, for example, when the average character string length of data included in a column is calculated as a feature quantity, the data meaning candidate estimation unit 230 calculates the similarity based on the ratio of the average character string lengths. Also, for example, the data meaning candidate estimation unit 230 may calculate the similarity using a machine learning method.

[0098] The data meaning candidate estimation unit 230 identifies columns from the selected table that are determined to have a high similarity based on predetermined conditions in at least one of the following: the similarity to the column corresponding to the manufacturing object ID calculated in step S1105; the similarity to the column corresponding to the business ID; and the similarity to the column corresponding to the business execution time. The data meaning candidate estimation unit 230 then creates table data meaning candidate information using the identified columns (S1106).

[0099] Specifically, the data meaning candidate estimation unit 230 determines, among the columns of the selected table, a column that has a high similarity to a column corresponding to a manufacturing object ID based on a predetermined condition as a manufacturing object ID candidate in the table data meaning candidate information for the selected table. The data meaning candidate estimation unit 230 also determines, among the columns of the selected table, a column that has a high similarity to a column corresponding to a business ID based on a predetermined condition as a business ID candidate in the table data meaning candidate information for the selected table. The data meaning candidate estimation unit 230 also determines, among the columns of the selected table, a column that has a high similarity to a column corresponding to a business execution time based on a predetermined condition as a business execution time candidate in the table data meaning candidate information for the selected table.

[0100] If the similarity calculated in step S1105 is determined as a quantitative numerical value, the data meaning candidate estimation unit 230 determines in step S1106 that the similarity is high if the similarity is equal to or greater than a predetermined threshold.

[0101] The data meaning candidate estimation unit 230 creates data meaning candidate information including table data meaning candidate information for each table estimated in step S1106 (S1107), and ends the data meaning candidate estimation process.

[0102] 12A is a diagram showing an example of the data configuration of a table that is a target of data meaning estimation. Table 1210 is an example of a table (site data) that is a target of data meaning estimation, and its name is assumed to be "Table B." Table 1210 includes columns col_a to col_f, respectively.

[0103] 12B is a diagram showing an example of the data configuration of the data meaning candidate information created in step S1107. The data meaning candidate information 1220 indicates the table name of the data meaning estimation target and the table data meaning candidate information related to that table. For example, the information stored in cell 1221 in the data meaning candidate information 1220 in FIG. 12B is the table data meaning candidate information related to "Table B."

[0104] For example, assume that "Table A" of the site data (i.e., table 410 shown in FIG. 4A) is set as actual arrival / completion data, and col1 of "Table A" corresponds to the manufacturing object ID, col3 corresponds to the task ID, and col5 corresponds to the task execution time.

[0105] In this case, the example of data meaning candidate information 1220 in FIG. 12B indicates that it has been determined in step S1206 that in “Table B” (i.e., table 1210 shown in FIG. 12A), which is the target of data meaning estimation, col_a and col_f are manufacturing object ID candidates, col_b and col_c are business ID candidates, and col_d and col_e are business execution time candidates.

[0106] In addition, if the similarity is calculated as a quantitative value in step S1105, the data meaning candidate information 1220 may further hold information indicating the similarity corresponding to the column of candidate manufacturing object IDs, the similarity corresponding to the column of candidate business IDs, and the similarity corresponding to the column of candidate business execution times.

[0107] Note that a single table for which data meaning is inferred may contain multiple columns that correspond to any of the candidate manufacturing object IDs, candidate business IDs, and candidate business execution times; a single table for which data meaning is inferred may not contain any columns that correspond to any of the candidate manufacturing object IDs, candidate business IDs, and candidate business execution times; or a single column in a table for which data meaning is inferred may correspond to two or more of the candidate manufacturing object IDs, candidate business IDs, and candidate business execution times.

[0108] [Final judgment] 13 is a flowchart showing an example of the data meaning inference process in step S804. The data meaning setting unit 240 executes the processes of steps S1302 to S1304 for each table that is a target of data meaning inference (S1301). That is, in step S1301, the data meaning setting unit 240 selects one unselected table from the tables that are targets of data meaning inference, and executes the processes of steps S1302 to S1304 for the selected table. This series of processes is repeated until it has been executed for all tables that are targets of data meaning inference.

[0109] However, tables for which data meaning estimation is to be performed that contain only one of manufacturing object ID candidates, business ID candidates, or business execution time candidates, and tables for which data meaning estimation is to be performed that contain none of manufacturing object ID candidates, business ID candidates, or business execution time candidates, are excluded from the processing of steps S1301 to S1304 (i.e., are not selected in step S1301).

[0110] The data meaning setting unit 240 identifies table data meaning candidate information related to the selected table for data meaning estimation from the data meaning candidate information 1220, generates all possible combinations of the columns of manufacturing object ID candidates, task ID candidates, and task execution time candidates based on the table data meaning candidate information, and executes the process of step S1303 for each of the generated combinations (S1302). That is, in step S1302, the data meaning setting unit 240 selects one unselected combination from the generated combination patterns, and executes the process of step S1303 for the selected combination, and repeats this series of processes until it has executed for all patterns of the combinations.

[0111] As described above, in the example of data meaning candidate information 1220 in FIG. 12B, according to the table data meaning candidate information for “Table B,” col_a and col_f of “Table B” are manufacturing object ID candidates, col_b and col_c are business ID candidates, and col_d and col_e are business execution time candidates.

[0112] At this time, in step S1302, the data meaning setting unit 240 generates all patterns of combinations consisting of a column of candidate manufacturing object IDs, a column of candidate business IDs, and a column of candidate business execution times from the table data meaning candidate information for "Table B."

[0113] The combinations of all patterns may include combinations in which the column of the candidate manufacturing object ID is not selected, combinations in which the column of the candidate business ID is not selected, and combinations in which the candidate business execution time is not selected. However, in the combinations, it is desirable that at least two of the column of the candidate manufacturing object ID, the column of the candidate business ID, and the column of the candidate business execution time are selected, and hereinafter, it is assumed that at least two of these are selected.

[0114] Specifically, for example, all possible combinations of (production object ID candidates, task ID candidates, task execution time candidates) for "Table B" are (col_a,col_b,col_d), (col_a,col_b,col_e), (col_a,col_c,col_d), (col_a,col_c,col_e), (col_f,col_b,col_d), (col_f,col_b,col_e), (col_f,col_c,col_d), (col_ f,col_c,col_e), (0,col_b,col_d), (0,col_b,col_e), (0,col_c,col_d), (0,col_c,col_e), (col_a,0,col_d), (col_a,0,col_e), (col_f,0col_d), (col_f,0,col_e), (col_a,col_b,0), (col_a,col_c,0), (col_f,col_b,0), and (col_f,col_c,0). Note that "0" included in the combination of (manufacturing object ID candidate, task ID candidate, task execution time candidate) indicates that the corresponding candidate column is not selected.

[0115] The data meaning setting unit 240 assumes that the combination selected in the most recent step S1302 for the selected table of the data meaning estimation target is correct (i.e., the columns of the manufacturing object ID candidate, task ID candidate, and task execution time candidate indicated by the selected combination correspond to the manufacturing object ID, task ID, and manufacturing execution time, respectively), and calculates the compatibility between the selected table of the data meaning estimation target (site data) and the process flow created by the process flow creation unit 220 under this assumption (S1303).

[0116] The data meaning setting unit 240 associates information indicating the combination whose degree of compatibility determined in step S1303 is high based on predetermined conditions with information indicating the table that is the target of data meaning inference, and sets this as table data semantic information (S1304), and ends the data meaning inference process. Specifically, in step S1304, for example, the data meaning setting unit 240 associates information indicating the combination with the highest degree of compatibility with information indicating the table that is the target of data meaning inference, but if the highest degree of compatibility is equal to or less than a predetermined value, the data meaning inference process may end without performing the association.

[0117] An example of the process of calculating the degree of conformance with the process flow in step S1303 will be described below with reference to Figures 14 to 17. In the examples of Figures 14 to 17, it is assumed that arrival and completion record data 1420 is set by the data setting unit 210, and that a process flow 1410 is created from the arrival and completion record data 1420 by the process flow creation unit 220. In the arrival and completion record data 1420, it is assumed that col_1 is set as the manufacturing object ID, col_2 is set as the task ID, and col_3 is set as the task execution time.

[0118] 14 is an explanatory diagram showing an example of the process of calculating the degree of conformance with the process flow in step S1303. Table 1430 is the site data (named "Table C") that is the target of data meaning estimation, and table data meaning candidate information for "Table C" is set in the data meaning candidate information 1220.

[0119] The table data meaning candidate information indicates that col_a and col_d are manufacturing object ID candidates, col_b and col_e are task ID candidates, and col_c and col_f are task execution time candidates in "Table C." Also, it is assumed that the data meaning setting unit 240 is selecting (col_a, col_b, col_c) as a combination of (manufacturing object ID candidate, task ID candidate, task execution time candidate) in step S1302.

[0120] A process for calculating the degree of conformance with the process flow 1410 under the above assumptions will be described. The data meaning setting unit 240 compares table 1430, which assumes that col_a is the manufacturing object ID, col_b is the task ID, and col_c is the task execution time, with arrival / completion performance data 1420 (process flow 1410).

[0121] Specifically, for example, the data meaning setting unit 240 determines whether the combination of the manufacturing object ID, task ID, and task execution time indicated by each record in the table 1430 under the assumption is consistent with the combination of the manufacturing object ID, task ID, and task execution time indicated by the arrival / completion performance data 1420, and with the process flow 1410.

[0122] 14, the first record of table 1430 indicates information that a task identified by a task ID "task A" was performed on a manufacturing object identified by a manufacturing object ID "mat_1" at "2024 / 01 / 01 10:10:00." The second record of table 1430 indicates information that a task identified by a task ID "task A" was performed on a manufacturing object identified by a manufacturing object ID "mat_1" at "2024 / 01 / 01 10:20:00."

[0123] The first record of the delivery completion data 1420 indicates that a task identified by a task ID "task A" was performed on a production object identified by a production object ID "mat_1" at "2024 / 01 / 01 10:00:00." The process flow 1410 indicates that "task B" is the subsequent task of "task A," and the second record of the delivery completion data indicates that a task identified by a task ID "task B" (which is the subsequent task of "task A" as described above) was performed on a production object identified by a production object ID "mat_1" at "2024 / 01 / 01 11:00:00." In other words, the arrival / completion performance data 1420 and the process flow 1410 indicate that "Task A" for "mat_1" can be performed after "2024 / 01 / 01 10:00:00" and before "2024 / 01 / 01 11:00:00" (the time when the subsequent task is performed).

[0124] Therefore, the above information indicated by the first record of table 1430 and the above information indicated by the second record are both consistent with the arrival / completion actual data 1420 and the process flow 1410. Details will be described later, but the degree of conformance with the process flow 1410 is calculated, for example, based on the proportion of records that are consistent with the arrival / completion actual data 1420 (process flow 1410) out of all records in table 1430, and the higher this proportion, the higher the conformance.

[0125] 14, it is assumed that the data meaning setting unit 240 determines in step S1302 that when (col_a, col_b, col_c) is selected as the combination of (manufacturing object ID candidate, task ID candidate, task execution time candidate), it has determined that this combination has the highest degree of compatibility with the process flow 1410. At this time, the data meaning setting unit 240 estimates table data semantic information for "Table C" according to the combination (i.e., generates information indicating that in "Table C," col_a corresponds to the manufacturing object ID, col_b corresponds to the task ID, and col_c corresponds to the task execution time), and stores this in the data semantic information 1450.

[0126] 15 is an explanatory diagram showing an example of the process of calculating the degree of conformance with the process flow in step S1303. Table 1510 is the site data (named "Table D") that is the target of data meaning estimation, and table data meaning candidate information for "Table D" is set in data meaning candidate information 1220.

[0127] The table data meaning candidate information indicates that in "Table D," col_a is a manufacturing object ID candidate, col_b is a task ID candidate, and col_c is a task execution time candidate. Also, it is assumed that the data meaning setting unit 240 is selecting (col_a, col_b, 0) as the combination of (manufacturing object ID candidate, task ID candidate, task execution time candidate) in step S1302 (i.e., only the task execution time candidate column is not selected in this combination).

[0128] The process of calculating the degree of conformance with the process flow 1410 under the above assumptions will be described. The data meaning setting unit 240 compares the table 1510, which assumes that col_a is the manufacturing object ID and col_b is the business ID, with the arrival / completion record data 1420.

[0129] Specifically, for example, the data meaning setting unit 240 determines whether the combination of the production object ID and the task execution time indicated by each record in table 1510 under this assumption is consistent with the combination of the production object ID and the task execution time indicated by the arrival / completion record data 1420, and with the process flow 1410. In other words, in the combination being selected in step S1302, the task execution time is not taken into consideration when determining consistency, because the column of the task execution time candidate is not selected.

[0130] 15, the time of task execution is not taken into consideration when determining consistency, so the first record in table 1510 indicates information that a task identified by a task ID "task B" was performed on a production object identified by a production object ID "mat_1." This information is consistent with the information indicated by the second record in arrival / completion performance data 1420, which indicates that a task identified by a task ID "task B" was performed on a production object identified by a production object ID "mat_1."

[0131] 15, it is assumed that the data meaning setting unit 240 selects (col_a, col_b, col_c) as the combination of (manufacturing object ID candidate, task ID candidate, task execution time candidate) in step S1302. In this case, the first record in table 1510 indicates information that a task identified by a task ID of "task B" was performed on a manufacturing object identified by a manufacturing object ID of "mat_1" at "2024 / 01 / 05 17:00:00".

[0132] However, the second record in the actual arrival and completion data 1420 indicates that "Task C" (the successor task to "Task B" in process flow 1410) to be performed for "mat_1" will start at "2024 / 01 / 01 11:00:00", and therefore the first record in table 1510, which indicates that "Task B" for "mat_1" started at "2024 / 01 / 05 17:00:00", after the start of "Task C", the successor task to "Task B", is not consistent with the actual arrival and completion data 1420 (process flow 1410).

[0133] In this way, in a combination of (production object ID candidate, task ID candidate, task execution time candidate), the proportion of records that are consistent with the arrival completion actual data 1420 (process flow 1410) out of all records in table 1510 is more likely to be higher when only two of the production object ID candidate column, task ID candidate column, and task execution time candidate columns are selected than when the production object ID candidate column, task ID candidate column, and task execution time candidate column are selected.

[0134] However, the more columns of candidate production object IDs, candidate task IDs, and candidate task execution times are selected, the less likely it is that the records in the table that are the target of data meaning estimation will match the arrival / completion record data 1420 (process flow 1410). However, if there is a match, it is considered that the likelihood that the combination is correct is high. Therefore, the degree of conformance with process flow 1410 will be higher, for example, the more columns of candidate production object IDs, candidate task IDs, and candidate task execution times are selected, and it is desirable that the degree of conformance be higher the higher the proportion.

[0135] 15, it is assumed that the data meaning setting unit 240 determines in step S1302 that when (col_a, col_b, 0) is selected as the combination of (manufacturing object ID candidate, task ID candidate, task execution time candidate), it has determined that this combination has the highest degree of compatibility with the process flow 1410. At this time, the data meaning setting unit 240 estimates table data semantic information for "Table D" according to this combination (i.e., generates information indicating that in "Table D," col_a corresponds to the manufacturing object ID, col_b corresponds to the task ID, and there is no column corresponding to the task execution time), and stores this information in the data semantic information 1450.

[0136] 16 is an explanatory diagram showing an example of the process of calculating the degree of conformance with the process flow in step S1303. Table 1610 is the site data (named "Table E") that is the target of data meaning estimation, and table data meaning candidate information for "Table E" is set in data meaning candidate information 1220.

[0137] The table data meaning candidate information indicates that in "Table E," col_a is a manufacturing object ID candidate, col_b is a task ID candidate, and col_c is a task execution time candidate. Also, it is assumed that the data meaning setting unit 240 is selecting (col_a, 0, col_c) as the combination of (manufacturing object ID candidate, task ID candidate, task execution time candidate) in step S1302 (i.e., only the task ID candidate column is not selected in this combination).

[0138] A process for calculating the degree of conformance with the process flow 1410 under the above assumptions will be described. The data meaning setting unit 240 compares the table 1610, which assumes that col_a is the manufacturing object ID and col_c is the task execution time, with the arrival / completion record data 1420.

[0139] Specifically, for example, the data meaning setting unit 240 determines whether the combination of the manufacturing object ID and the business ID indicated by each record in the table 1610 under this assumption is consistent with the combination of the manufacturing object ID and the business ID indicated by the arrival / completion record data 1420, and with the process flow 1410. In other words, since the business ID column is not selected in the combination being selected in step S1302, the business ID is not taken into consideration in determining the consistency.

[0140] In the example of Figure 16, the business ID is not taken into consideration when determining consistency, so the first record in table 1610 indicates information that some business was performed on a manufacturing object identified by the manufacturing object ID "mat_1" at "2024 / 01 / 01 10:10:00".

[0141] Here, the first record of the delivery completion data 1420 indicates information that some task was performed at "2024 / 01 / 01 10:00:00" on a production object identified by the production object ID "mat_1." The second record of the delivery completion data indicates information that some task was performed at "2024 / 01 / 01 11:00:00" on a production object identified by the production object ID "mat_1." In other words, this indicates that task A for "mat_1" can be performed after "2024 / 01 / 01 10:00:00" and before "2024 / 01 / 01 11:00:00." Therefore, the above information indicated by the first record in table 1610 is consistent with the delivery completion data 1420.

[0142] 16, it is assumed that the data meaning setting unit 240 selects (col_a, col_b, col_c) as the combination of (manufacturing object ID candidate, task ID candidate, task execution time candidate) in step S1302. In this case, the first record in table 1610 indicates information that a task identified by a task ID of "task X" was performed on a manufacturing object identified by a manufacturing object ID of "mat_1" at "2024 / 01 / 01 11:00:00".

[0143] However, this information is inconsistent with the information indicated in the first record of the arrival completion data 1420, which indicates that a task identified by a task ID "Task A" was performed on a manufacturing object identified by a manufacturing object ID "mat_1" at "2024 / 01 / 01 10:00:00", and the information indicated in the second record of the arrival completion data 1420, which indicates that a task identified by a task ID "Task B" (the successor task to "Task A" in process flow 1410) was performed on a manufacturing object identified by a manufacturing object ID "mat_1" at "2024 / 01 / 01 11:00:00" (i.e., the tasks are inconsistent).

[0144] 16, it is assumed that the data meaning setting unit 240 determines in step S1302 that when (col_a, 0, col_c) is selected as the combination of (manufacturing object ID candidate, task ID candidate, task execution time candidate), it has determined that this combination has the highest degree of compatibility with the process flow 1410. At this time, the data meaning setting unit 240 estimates table data semantic information for "Table E" according to the combination (i.e., generates information indicating that in "Table E", col_a corresponds to the manufacturing object ID, there is no column corresponding to the task ID, and col_c corresponds to the task execution time), and stores this in the data semantic information 1450.

[0145] 17 is an explanatory diagram showing an example of the process of calculating the degree of conformance with the process flow in step S1303. Table 1710 is the site data (named "Table X") that is the target of data meaning estimation, and table data meaning candidate information for "Table X" is set in data meaning candidate information 1220.

[0146] The table data meaning candidate information indicates that in "Table X," col_a is a manufacturing object ID candidate, col_b is a task ID candidate, and col_c is a task execution time candidate. Also, it is assumed that the data meaning setting unit 240 is selecting (0, col_b, col_c) as the combination of (manufacturing object ID candidate, task ID candidate, task execution time candidate) in step S1302 (i.e., only the manufacturing object ID candidate column is not selected in this combination).

[0147] The following describes the process of calculating the degree of conformance with the process flow 1410 under the above assumptions. The data meaning setting unit 240 compares the table 1710, which assumes that col_b is the task ID and col_c is the task execution time, with the arrival / completion record data 1420.

[0148] Specifically, for example, the data meaning setting unit 240 determines whether the combination of task ID and task execution time indicated by each record in table 1710 under this assumption is consistent with the combination of task ID and task execution time indicated by arrival / completion record data 1420, and with the process flow 1410. In other words, since the production object ID column is not selected in the combination being selected in step S1302, the production object ID is not taken into consideration in determining consistency.

[0149] In the example of Figure 17, the manufacturing object ID is not taken into consideration when determining consistency, so the first record in table 1710 indicates information that a task identified by the task ID "Task A" was performed at "2024 / 01 / 01 10:00:00", and the second record in table 1710 indicates information that a task identified by the task ID "Task A" was performed on a manufacturing object at "2024 / 01 / 01 10:01:00".

[0150] Furthermore, the first record of the delivery completion actual data 1420 indicates that a task identified by a task ID "Task A" was performed at "2024 / 01 / 01 10:00:00." Furthermore, the process flow 1410 indicates that "Task B" is the subsequent task of "Task A," and the second record of the delivery completion actual data indicates that a task identified by a task ID "Task B" (which, as described above, is the subsequent task of "Task A") was performed at "2024 / 01 / 01 11:00:00." In other words, the delivery completion actual data 1420 and the process flow 1410 indicate that "Task A" can be performed after "2024 / 01 / 01 10:00:00" and before "2024 / 01 / 01 11:00:00" (the time when the subsequent task is performed).

[0151] Therefore, the above information indicated by the first record of the table 1710 and the above information indicated by the second record are both consistent with the arrival / completion performance data 1420 and the process flow 1410 .

[0152] 17, it is assumed that the data meaning setting unit 240 selected (col_a, col_b, col_c) as the combination of (manufacturing object ID candidate, task ID candidate, task execution time candidate) in step S1302. In this case, the first record of table 1610 indicates information that a task identified by a task ID "task A" was performed on a manufacturing object identified by a manufacturing object ID "metal" at "2024 / 01 / 01 10:00:00", and the second record of table 1610 indicates information that a task identified by a task ID "task A" was performed on a manufacturing object identified by a manufacturing object ID "metal" at "2024 / 01 / 01 10:01:00".

[0153] However, this information is inconsistent with the information indicated in the first record of the arrival completion data 1420, which indicates that a task identified by a task ID "Task A" was performed on a manufacturing object identified by a manufacturing object ID "mat_1" at 10:00:00 on 2024 / 01 / 01, and the information indicated in the second record of the arrival completion data 1420, which indicates that a task identified by a task ID "Task B" (the successor task to "Task A" in process flow 1410) was performed on a manufacturing object identified by a manufacturing object ID "mat_1" at 11:00:00 on 2024 / 01 / 01 (the manufacturing objects do not match).

[0154] 17, it is assumed that the data meaning setting unit 240 determines in step S1302 that when (0, col_b, col_c) is selected as the combination of (production object ID candidate, task ID candidate, task execution time candidate), it has determined that this combination has the highest degree of compatibility with the process flow 1410. At this time, the data meaning setting unit 240 estimates table data semantic information for "table X" according to the combination (i.e., generates information indicating that in "table X," there is no column corresponding to the production object ID, col_b corresponds to the task ID, and col_c corresponds to the task execution time), and stores this in the data semantic information 1450.

[0155] A specific example of a method for calculating the degree of conformance with a process flow as a quantitative value will be described below. The data meaning setting unit 240 may calculate the degree of conformance with a process flow as a quantitative value, for example, using the following (Equation 1).

[0156]

number

[0157] In (Equation 1), L indicates the degree of conformance with the process flow, S indicates the score, and W indicates the weight for the score. Also, i indicates which column was selected from the data meaning candidates. Also, i_t, i_a, and i_p (where "_" indicates a subscript) respectively indicate the columns of the task execution time candidates, task ID candidates, and manufacturing object ID candidates selected from the table targeted for data meaning estimation.

[0158] That is, L i_t,i_a,i_p indicates the compatibility with the process flow in the column combination pattern selected in step S1302, and S i_t,i_a,i_p indicates the score given based on the consistency with the process flow in the column combination pattern, and w i_t,i_a,i_p indicates the weight multiplied by the score given based on the consistency with the process flow in the column combination pattern. Note that i_t=0, i_a=0, and i_p=0 indicate that the task execution time candidate column, task ID candidate column, and manufacturing object ID candidate column are not selected, respectively.

[0159] In the example of calculating the degree of conformance in Fig. 14, (col_a, col_b, col_c) is selected as the combination of (manufacturing object ID candidate, task ID candidate, task execution time candidate) from "Table C" of the data semantic inference target, so i_t = col_c, i_a = col_b, i_p = col_a. Also, in the example of calculating the degree of conformance in Fig. 15, (col_a, col_b, 0) is selected as the combination of (manufacturing object ID candidate, task ID candidate, task execution time candidate) from "Table D" of the data semantic inference target, so i_t = 0, i_a = col_b, i_p = col_a.

[0160] S i_t,i_a,i_p is a predetermined value that varies depending on whether i_t, i_a, and i_p are 0 or not. For example, when all of i_t, i_a, and i_p are not 0, S i_t,i_a,i_p = S1, and when i_a is 0 and i_t and i_p are not 0, then S i_t,i_a,i_p = S2, and when i_p is 0 and i_a and i_t are not 0, then S i_t,i_a,i_p = S3, and when i_t is 0 and i_a and i_p are not 0, then S i_t,i_a,i_p = S4, the relationship is set such that S1 > S2 > S3 > S4. It is desirable that S1 is larger than any of S2, S3, and S4, but the magnitude relationship between S2, S3, and S4 may be arbitrary, or at least some of S2, S3, and S4 may be the same value.

[0161] w i_t,i_a,i_p is the ratio of the number of records that match the arrival / completion performance data (process flow) to all records in the table for which data meaning is to be inferred, under the condition that i_t, i_a, and i_p are selected. As mentioned above, columns that correspond to values ​​of 0 among i_t, i_a, and i_p are not taken into consideration when determining the consistency.

[0162] In the example of FIG. 14, even when i_t=col_c, i_a=col_b, and i_p=col_a (i.e., when all are not 0), the records included in the "table C" that is the data meaning estimation target are consistent with the delivery completion record data 1420. i_t,i_a,i_p In addition, S i_t,i_a,i_p Since is assigned a high score of S1, the combination of i_t=col_c, i_a=col_b, and i_p=col_a will have a high fitness value.

[0163] In the example of Figure 15, if i_t=col_c, i_a=col_b, and i_p=col_a, then S i_t,i_a,i_p is S1 and has a high score, but the records contained in "Table D" that is the data semantic estimation target are not consistent with the arrival completion data 1420. i_t,i_a,i_p On the other hand, in the example of Figure 15, if i_t=0, i_a=col_b, and i_p=col_a, then S i_t,i_a,i_p Although the score is low at S4, the records included in the data meaning estimation target "Table D" are consistent with the arrival completion record data 1420. i_t,i_a,i_p In this case, the combination of i_t=0, i_a=col_b, i_p=col_a may have a higher fitness value than the combination of i_t=col_c, i_a=col_b, i_p=col_a.

[0164] Furthermore, when the data meaning candidate estimation unit 230 creates the data meaning candidate information 1220, if the similarity of each column is recorded as a quantitative value in the data meaning candidate information 1220, the quantitative value may be taken into consideration in determining the degree of compatibility with the process flow. For example, the variable x may be the quantitative value of the similarity with the data meaning, and the function f may be a function for taking the quantitative value into consideration (for example, f may be an increasing function of x), and the value obtained by adding the value of the function f into which the variable x is substituted to the variable L may be determined as the degree of compatibility.

[0165] As shown in the examples in Figures 14 to 17, the data meaning setting unit 240 calculates the compatibility with the process flow for each column combination that serves as table data meaning candidate information for a table of field data that is the target of table data meaning estimation, and sets the column combination pattern that is determined to have the highest compatibility (and is greater than or equal to a predetermined value) as the table data meaning information for that table.

[0166] [Screen display] 18A is a diagram showing an example of the screen configuration of the arrival completion record data setting screen. Arrival completion record data 1810 is displayed on, for example, the operation terminal 160 in order to accept settings related to the arrival completion record data from the user in step S801.

[0167] The delivery completion performance data 1810 includes an area 1820 for accepting settings of access information to the site data storage unit 140, an area 1830 for displaying a list of site data stored in the site data storage unit 140, and an area 1840 for displaying details of site data selected from the site data list displayed in area 1830.

[0168] Area 1820 is an area for accepting settings such as host information for the on-site data storage unit 140 and port information for accessing the database.

[0169] Area 1830 displays, for example, a list of site data stored in site data accumulation unit 140 and information indicating whether each displayed site data is set as actual delivery completion data. Area 1830 also has an interface for displaying details of a table selected from the list in area 1840. Area 1830 also has an interface for receiving an instruction to create a process flow from the set actual delivery completion data when the setting of actual delivery completion data has been completed.

[0170] Area 1840 displays details of the table selected in area 1830. Area 1840 displays, for example, a list of table columns and actual records. Area 1840 also has an interface for accepting settings such as which column is the production object ID, which column is the task ID, and which column is the task execution time when the displayed table is set as arrival completion actual data. Area 1840 also has an interface that reflects the column settings made by the user, an interface for accepting settings that the displayed table is not arrival completion actual data, and the like.

[0171] 18B is a diagram showing an example of the screen configuration of a process flow display screen 1850. A process flow display screen 1850 displays the process flow created in step S802 from the delivery completion record data set in step S801.

[0172] 19A is a diagram showing an example of the screen configuration of a data meaning inference target setting screen 1910. The data meaning inference target setting screen 1910 is displayed on, for example, the operation terminal 160 in order to receive settings from the user in step S803 regarding the field data to be inferred as the data meaning information.

[0173] The data meaning estimation target setting screen 1910 includes, for example, an area 1920 for accepting the setting of access information to the site data storage unit 140, an area 1930 for displaying a list of site data stored in the site data storage unit 140, and an area 1940 for displaying details of site data selected from the site data list displayed in area 1930.

[0174] Area 1920 is an area for accepting settings such as host information for the site data storage unit 140 and port information for accessing the database.

[0175] Area 1930 displays, for example, a list of site data stored in site data accumulation unit 140 and information indicating whether each of the displayed site data is set as a target for data semantic information estimation. Area 1930 also has an interface for displaying details of a table selected from the list in area 1940. Area 1930 also has an interface for receiving an instruction to perform data semantic information estimation for site data set as a target for data semantic information estimation.

[0176] Area 1940 displays details of the table selected in area 1930. Area 1940 displays, for example, a list of table columns or actual records. Area 1940 may also have an interface for accepting settings of at least one of a column corresponding to a manufacturing object ID, a column corresponding to a business ID, and a column corresponding to a business execution time when the column is known for the table set as the data meaning estimation target. Area 1940 may also have an interface that reflects column settings made by the user.

[0177] In addition, for tables in which at least one of a column corresponding to the manufacturing object ID, a column corresponding to the business ID, and a column corresponding to the business execution time is set, the table may be excluded from the target for estimating data semantic information (i.e., the at least one meaning is treated as correct), and the set information may be used as training data in a machine learning method when estimating table data semantic candidate information for other data that is the target for estimating data semantic information.

[0178] 19B is a diagram showing an example of the screen configuration of a data meaning estimation result display screen 1950. The data meaning estimation result display screen 1950 is a screen for displaying the estimation result in the data meaning estimation process in step S804, and is displayed on the operation terminal 160, for example.

[0179] For example, the estimation results of data semantic information for each piece of field data that was the subject of data semantic information estimation are displayed on the data semantic estimation result display screen 1950. Also, for example, the data semantic setting unit 240 provides the estimation results of data semantic information to the process model creation unit 250, and the process model creation unit 250 creates a process model using the estimation results of data semantic information.

[0180] In this case, the process model may be displayed on the data semantic estimation result display screen 1950. In the process model, field data (including the table that was the target of data semantic estimation) is linked to the business-related information node connected to the business node, and the above-mentioned estimation result of the data semantic information is used to link the field data. Furthermore, the data semantic estimation result display screen 1950 may display information indicating field data that could not be linked to the process model as a result of estimating the data semantic information.

[0181] As described above, the process model creation system 110 of this embodiment compares the column feature values ​​of the table to be subjected to data meaning estimation with the column feature values ​​of the arrival / completion performance data to generate data meaning candidate information, thereby narrowing down data meaning candidates with a small amount of calculation and high accuracy.

[0182] Furthermore, the process model creation system 110 of this embodiment selects a combination of columns from the data meaning candidate information, and calculates the degree of compatibility based on the consistency between the table to be subjected to data meaning estimation, assuming that the data meaning indicated by the combination is correct, and the arrival / completion record data and the process flow created from the arrival / completion record data, thereby making it possible to estimate the data meaning with a small amount of calculation and with a high degree of accuracy.

[0183] Furthermore, if the user inputs the arrival / completion performance data and the settings of the table for which data meaning is to be inferred, the process model creation system 110 can infer the data meaning of the table for which data meaning is to be inferred, which reduces the workload on the user. [Example]

[0184] In this embodiment, an example of processing will be described in which the data meaning setting unit 240 estimates the meaning of a table from table data meaning candidate information. Below, explanations of points similar to those in the first embodiment will be omitted as appropriate.

[0185] 20 is an explanatory diagram showing an example of a process for estimating the meaning of a table from table data meaning candidate information. In the example of Fig. 20, it is assumed that arrival / completion record data 2010 is set by the data setting unit 210, and that process flows 2020 and 2030 are created from the arrival / completion record data 2010 by the process flow creation unit 220. In the arrival / completion record data 2010, col_1 is set as the manufacturing object ID, col_2 is set as the task ID, and col_3 is set as the task execution time.

[0186] Table 2040 is on-site data that is the subject of data meaning estimation (its name is "Table F"), and table data meaning candidate information for "Table F" is set in the data meaning candidate information 1220.

[0187] The table data semantic candidate information indicates that in "Table F," col_1 and col_2 are production object ID candidates, there are no task ID candidates, and col_3 is a task execution time candidate. In this embodiment, when selecting a combination consisting of a production object ID candidate column, a task ID candidate column, and a task execution time candidate column in step S1302, two production object ID candidate columns can be selected for the production object ID candidate. Note that one of the two production object ID candidate columns can be selected as the production object ID candidate column for the input part, and the other column can be selected as the production object ID candidate column for the product to be manufactured from that part.

[0188] Here, for example, the data meaning setting unit 240 may infer that "Table F" is parts input result data when the data in each of the two columns selected from the manufacturing object ID candidates in "Table F" (table 2040) can both be confirmed in the data in col1 corresponding to the manufacturing object ID in the arrival / completion result data 2010. Here, the parts input result data indicates the results of input materials (input parts) used when work is performed on the manufacturing object in the manufacturing process.

[0189] In the example of Figure 20, it is determined from the arrival / completion record data 2010 that the task execution times for "mat_1" of "Task A," "Task B," and "Task C" included in the process flow 2020 corresponding to "mat_1" indicated in the cell of col_1 in the first record of "Table F" are all earlier than the task execution times for "mat_2" of "Task P," "Task Q," and "Task R" that constitute the process flow 2030 corresponding to "mat_2" indicated in the cell of col_2 in the second record of "Table F." In this case, it may be estimated that the part "mat_1" was used as an input material when the task was performed on the product "mat_2."

[0190] The data meaning setting unit 240 may provide information indicating the estimated part input record data together with the data meaning information to the process model creation unit 250. Furthermore, the process model creation unit 250 may create a process model using the provided part input record data. Furthermore, the process model creation system 110 may display information about the process model including information about the part input record data on the data meaning estimation result display screen 1950. [Example]

[0191] In this embodiment, an example of processing will be described in which the data meaning setting unit 240 estimates data meaning estimation information based on the process model created by the process model creation unit 250. Hereinafter, explanations of points similar to those in the first embodiment will be omitted as appropriate.

[0192] 21 is an explanatory diagram showing an example of processing for inferring data semantic information from a process model. A process model 2110 is created by the process model creation unit 250. In the process model 2110, a task-related information node 2112 is connected to a task node 2111 indicating "task B." Information linked to the task-related information node 2112 includes a table 2121 (whose name is "table a") and a table 2122 (whose name is "table b").

[0193] In addition, in the process model 2110, a task-related information node 2114 is connected to a task node 2113 indicating "task D." Information linked to the task-related information node 2114 includes a table 2123 (whose name is "table c").

[0194] Here, it is assumed that the data meaning setting unit 240 provides the estimated table data semantic information of "table a" and "table b" to the process model creation unit 250, and it is estimated that the table data semantic information will be linked to the business-related information node 2112 connected to the business node 2111 of "business B" in the process model 2110.

[0195] Here, for example, if the data meaning setting unit 240 determines that there is a correlation between the data included in "table a" and the data included in "table b", it determines that the table data semantic information of "table a" and "table b" provided to the process model creation unit 250 is appropriate, and if it determines that there is no correlation, it determines that the table data semantic information of "table a" and "table b" provided to the process model creation unit 250 is inappropriate.

[0196] Also, for example, if the data meaning setting unit 240 determines that the correlation between the data contained in "Table a" and the data contained in "Table c" is higher than the correlation between the data contained in "Table a" and the data contained in "Table b", then "Table a" and "Table c" should be linked to the same business-related information.

[0197] Therefore, in this case, the data meaning setting section 240 may determine that the table data semantic information of "table a", "table b", and "table c" provided to the process model creating section 250 is inappropriate.

[0198] The data meaning setting unit 240 may use, as the correlation between the data, for example, an index indicating the cross-correlation between time-series data included in the data, or an index indicating the similarity of character strings in table names. The data meaning setting unit 240 may also determine whether the table data semantic information is appropriate or inappropriate depending on whether the index indicating the correlation between the data is equal to or greater than a predetermined threshold.

[0199] Furthermore, the data meaning setting unit 240 may redo the estimation of data semantic information for a table for which it has determined that the table data semantic information is inappropriate. Specifically, for example, the data meaning setting unit 240 may change the values ​​of the parameters used to calculate the degree of conformance (for example, by changing the values ​​of S1, S2, S3, and S4 from their initial values ​​according to a user input) and then redo the data meaning estimation process corresponding to the table data semantic information determined to be inappropriate, or may present the table data semantic information determined to be inappropriate and the corresponding table to the operation terminal 160, set at least a portion of the information included in the table data semantic information according to a user input (assuming that the at least a portion of the information is correct), and then redo the data meaning candidate estimation process and the data meaning estimation process for the table.

[0200] For example, if three or more tables are linked to one business-related information node, the data meaning setting unit 240 may cluster the three or more tables according to a predetermined algorithm, and determine that the table data semantic information of the table belonging to the cluster with the largest number of tables is appropriate, and determine that the table data semantic information of the table belonging to a cluster other than the cluster with the largest number of tables is inappropriate.

[0201] As described above, according to the embodiment, the process model creation system 110 can set data semantic information for on-site data, create a process model based on the set semantic information, and provide the process model to the information collection system 150. As a result, it is possible to automate the task of creating a process model to be used in the information collection system 150, and it is expected to have the effect of making it easier to collect and analyze on-site data in the information collection system 150.

[0202] In the above-described embodiment, an example has been described in which the process model creation system 110 infers the data meaning of field data generated in a manufacturing process at a manufacturing site, but the process model creation system 110 can be applied not only to manufacturing sites but also to any business type that is made up of multiple tasks and generates field data. For example, in the logistics or transportation industry, there are multiple steps that make up the transportation process from receiving goods to sorting, storage, packaging, and delivery, and semantic data estimation for field data related to the transportation process can be performed in the same way as for field data at a manufacturing site.

[0203] The present invention is not limited to the above-described embodiments, but includes various modifications. For example, the above-described embodiments have been described in detail to clearly explain the present invention, and the present invention is not necessarily limited to those including all of the described configurations. Furthermore, it is possible to replace part of the configuration of one embodiment with the configuration of another embodiment, or to add the configuration of another embodiment to the configuration of one embodiment. Furthermore, it is possible to add, delete, or replace part of the configuration of each embodiment with other configurations.

[0204] Furthermore, the above-described configurations, functions, processing units, processing means, etc. may be partially or entirely implemented in hardware, for example, by designing them as integrated circuits. The above-described configurations, functions, etc. may also be implemented in software, with a processor interpreting and executing a program that implements each function. Information such as the programs, tables, and files that implement each function can be stored in a memory, a recording device such as a hard disk or SSD (Solid State Drive), or a recording medium such as an IC card, SD card, or DVD.

[0205] In addition, the control lines and information lines shown are those that are considered necessary for the explanation, and do not necessarily show all the control lines and information lines in the product. In reality, it can be assumed that almost all components are interconnected. [Explanation of symbols]

[0206] 110 process model creation system, 130 data generation device, 140 field data accumulation unit, 150 information collection system, 160 operation terminal, 210 data setting unit, 220 process flow creation unit, 230 data meaning candidate estimation unit, 240 data meaning setting unit, 250 process model creation unit, 610 process model, 720 data meaning information, 1001 CPU, 1002 memory, 1003 auxiliary storage device, 1220 data meaning candidate information

Claims

1. An information processing system, a processor and a memory, The memory includes: Arrival and completion performance data selected from field data related to a process including a plurality of tasks and having data items indicating predetermined meanings designated therein; Semantic estimation target data selected from the on-site data; A feature value of a data item indicating the predetermined meaning of the delivery completion record data; a feature amount of each data item included in the semantic estimation target data; The processor: Inferring candidates for data items that indicate the predetermined meaning from the data items of the meaning estimation target data based on feature amounts of data items that indicate the predetermined meaning of the delivery completion record data and feature amounts of each data item included in the meaning estimation target data; execute a comparison process to compare the delivery completion record data with the meaning estimation target data when the candidate is a data item indicating the predetermined meaning; An information processing system that infers a data item that indicates the predetermined meaning from the candidates based on a result of the comparison process.

2. 2. The information processing system according to claim 1, The processor creates a process model that models the relationships between tasks included in the process based on the arrival / completion performance data and the estimation results of the data items.

3. 2. The information processing system according to claim 1, the data item indicating the predetermined meaning includes a tracking object identification item indicating information for identifying a tracking object in the process, a tracking point identification item indicating information for identifying a tracking point in the process, and a timing item indicating a timing at which the tracking object was present at the tracking point in the process; The processor: creating a process flow showing information on the tracking object following the tracking point based on the tracking object identification item, the tracking point identification item, and the timing item of the delivery completion record data; In the comparison process, the data to be subjected to meaning estimation is compared with the process flow when the candidate is a data item indicating the predetermined meaning.

4. 4. The information processing system according to claim 3, The processor: In the comparison process, a degree of conformance between the meaning estimation target data and the process flow is calculated when the candidate is a data item indicating the predetermined meaning; An information processing system that estimates a candidate whose calculated degree of suitability is determined to be high based on a predetermined condition as a data item that indicates the predetermined meaning.

5. 5. The information processing system according to claim 4, The processor calculates the degree of compatibility based on an index indicating the degree to which a combination of the tracking object identification item, the tracking location identification item, and the timing item indicated by the candidate in the meaning estimation target data matches a combination of the tracking object identification item, the tracking location identification item, and the timing item in the arrival / completion actual data.

6. 6. The information processing system according to claim 5, The processor calculates the degree of compatibility based on the index and a score that is predetermined corresponding to a combination of the tracking target identification item, the tracking location identification item, and the timing item indicated by the candidate.

7. 5. The information processing system according to claim 4, When the processor determines that a combination including the timing item and a plurality of the tracking object identification items indicated by the candidate in the meaning estimation target data matches a combination including the timing item and the tracking object identification items in the arrival / completion actual data, the processor infers that the meaning estimation target data is parts input actual data.

8. 3. The information processing system according to claim 2, The processor: based on the estimation result, associating the on-site data including the semantic estimation target data with the process model; An information processing system that determines whether an estimation result of a data item indicating the predetermined meaning in the meaning estimation target data is appropriate based on a correlation between on-site data associated with the process model.

9. 9. The information processing system according to claim 8, The processor: Regarding the semantic estimation target data for which the estimation result is determined to be inappropriate, Accepting input of information indicating a part of the data item indicating the predetermined meaning; an information processing system that, assuming that the part of the information is correct, executes again the process of estimating the candidate, the comparison process, and the process of estimating the data item indicating the predetermined meaning from the candidate.

10. An information processing method by an information processing system, the information processing system includes a processor and a memory; The memory includes: Arrival and completion performance data selected from field data related to a process including a plurality of tasks and having data items indicating predetermined meanings designated therein; Semantic estimation target data selected from the on-site data; A feature value of a data item indicating the predetermined meaning of the delivery completion record data; a feature amount of each data item included in the semantic estimation target data; The information processing method includes: the processor estimates candidates for data items that indicate the predetermined meaning from the data items of the meaning estimation target data based on a feature amount of a data item that indicates the predetermined meaning of the arrival completion record data and a feature amount of each data item included in the meaning estimation target data; the processor executes a comparison process of comparing the arrival / completion record data with the meaning estimation target data when the candidate is a data item indicating the predetermined meaning, The information processing system, wherein the processor infers a data item that indicates the predetermined meaning from the candidates based on a result of the comparison process.

Images