Information processing device, information processing method and information processing program

Abstract

To provide an information processing device, information processing method and information processing program for improving accuracy of classifying and organizing natural language used in business processes.SOLUTION: An information processing device includes: an acquisition unit for acquiring a document; an extraction unit that extracts, from the document, at least any of natural language, an item name with which the natural language is extracted, a business process and a document type; a reading unit that searches a document information table to determine whether information identical to and / or similar to at least any of the natural language, the item name, the business process and the document type is present in the document information table, and reads information from the document information table on the basis of a result of the determination; a registration unit that registers, in a data set, data of the natural language, the item name, the business process, the document type, a process classification and items having a dependency-correlation and / or an attribute-correlation with the item; and a document information storage unit for storing the document information table.SELECTED DRAWING: Figure 1

Description

[Technical Field]

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

[0002] In business processes such as design and development, when data is aggregated and passed to tools or systems to carry out business operations, the aggregated data is usually stored in fragments and written in various formats, requiring engineers and tool experts to process each piece of data. By structuring this data, it is possible to associate and utilize a series of data. In this case, it is necessary to organize and re-register the relationships between multiple pieces of data. Means for organizing and integrating multiple pieces of data have been disclosed (see, for example, Patent Documents 1 to 3).

[0003] Patent Document 1 discloses an information processing device that integrates two or more quality function deployment tables. Patent Document 2 discloses an information processing device that displays correlation diagrams. Patent Document 3 discloses an information processing device that makes it easy to create tables such as deployment tables related to quality function deployment. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] JP 2014-112340 A [Patent Document 2] JP 2019-219787 A [Patent Document 3] JP 2016-081185 A Summary of the Invention [Problem to be solved by the invention]

[0005] In order to consolidate data stored in fragments and written in various formats, it was necessary to translate and normalize the data by classifying and organizing the natural language used in business processes.To achieve this, normalization has been performed by translating using a dictionary, using AI or machine learning to remove variations by utilizing distributed representations, correcting omissions of subjects or words within the text, etc., but sufficient accuracy has still not been achieved.

[0006] In view of the above problems, the present invention aims to provide an information processing device, an information processing method, and an information processing program that can improve the accuracy of classifying and organizing natural language used in business processes. [Means for solving the problem]

[0007] A first aspect of the present invention is an information processing device comprising: an acquisition unit that acquires a specified document; an extraction unit that extracts at least one of natural language from the specified document, the name of an item in the specified document from which the natural language has been extracted, a business process to which the specified document belongs, and a type of the specified document; a reading unit that searches a document information table to determine whether information identical and / or similar to at least one of the received natural language, the name of the item, the business process, and the type of the specified document is present in the document information table, and based on the determination result, reads out information listed in the document information table about the process category to which the item belongs and items that have a dependency correlation and / or attribute correlation with the item; a registration unit that registers data about the natural language, the name of the item, the business process, the type of the specified document, the process category to which the item belongs, and items that have a dependency correlation and / or attribute correlation with the item in a dataset; and a document information storage unit that stores the document information table.

[0008] In a first aspect of the present invention, the attribute correlation may be a correlation between a given natural language and a natural language expressing an attribute that describes it.

[0009] In a first aspect of the present invention, a dependency correlation may be a correlation having a dependency relationship between a given natural language and a natural language that appears around the given natural language, either a mechanism, a parent-child relationship, or a cause-and-effect relationship.

[0010] In a first aspect of the present invention, the process section may be any of the process sections R, F, L, P, l, and p.

[0011] In a first aspect of the present invention, the reading unit may determine, based on one or more criteria, whether information similar to the received natural language and at least one of the item name, business process, and specified document type exists in the document information table.

[0012] In the first aspect of the present invention, the one or more criteria may be composed of at least one of a criterion for determining whether business processes are similar, a criterion for determining whether types of specified documents are similar, a criterion for determining whether item names are similar, a criterion for determining whether process classifications are similar, a criterion for determining whether dependency correlations are similar, a criterion for determining whether attribute correlations are similar, and a criterion for determining whether natural language character strings are similar.

[0013] In the first aspect of the present invention, the reading section may digitize one or more similarities based on one or more criteria.

[0014] In the first aspect of the present invention, the reading section may determine that the information is similar by comparing one or more similarities with one or more thresholds.

[0015] In the first aspect of the present invention, the registration unit may register one or more similarities in the dataset.

[0016] In a first aspect of the present invention, the registration unit may register information about natural language, item name, business process, type of specified document, process category to which the item belongs, and items that have a dependency correlation and / or attribute correlation with the item in the document information table.

[0017] The first aspect of the present invention may further comprise a data processing unit that performs statistics and organization on the data stored in the data set.

[0018] In the first aspect of the present invention, the system may further include a name matching unit that searches within the dataset based on the natural language and information about processes, items that are mutually dependent on each other, and / or items that have attribute relationships, and determines whether a natural language that is highly similar to the natural language is registered in the dataset.

[0019] The first aspect of the present invention may further include a document information generation unit that replaces all natural languages ​​contained in the document with natural languages ​​determined by the name matching unit to be highly similar, and then generates the replaced document.

[0020] In a first aspect of the present invention, a document tagging unit may be provided which, after character strings determined to be highly similar by the name matching unit are extracted from a dataset, tags all natural languages ​​contained in the document with the natural languages ​​determined to be highly similar by the name matching unit, and then generates a document in which the natural languages ​​have been replaced with the tagged natural languages.

[0021] A second aspect of the present invention is summarized as follows: an information processing device having a document information storage unit that stores a document information table, comprising: an acquisition step of acquiring a specified document; an extraction step of extracting from the specified document a natural language, the name of an item in the specified document from which the natural language has been extracted, a business process to which the specified document belongs, and / or a type of the specified document; a reading step of searching the document information table and reading information about a process category to which the item belongs and items that have a mutual dependency correlation and / or attribute correlation with the item, which are listed in the document information table and are linked to information that is identical or similar to the received natural language and / or information about the name of the item, the business process, and / or the type of the specified document; and a registration step of registering data about the natural language, the name of the item, the business process, the type of the specified document, the process category to which the item belongs, and the items that have a mutual dependency correlation and / or attribute correlation with the item in a dataset.

[0022] A third aspect of the present invention is an information processing program that causes a computer to realize an acquisition function that acquires a specified document; an extraction function that extracts at least one of natural language from the specified document, the name of an item in the specified document from which the natural language has been extracted, the business process to which the specified document belongs, and the type of the specified document; a reading function that searches a document information table and reads information about a process category to which the item belongs and items that have a mutual dependency correlation and / or attribute correlation with the item, which are listed in the document information table and are linked to information that is identical or similar to at least one of the received natural language, the name of the item, the business process, and the type of the specified document; a registration function that registers data about the natural language, the name of the item, the business process, the type of the specified document, the process category to which the item belongs, and the items that have a mutual dependency correlation and / or attribute correlation with the item in a dataset; and a document information storage function that stores the document information table.

[0023] According to the present invention, it is possible to provide an information processing device, an information processing method, and an information processing program that can improve the accuracy of classifying and organizing natural language used in business processes. [Brief explanation of the drawings]

[0024] [Figure 1] FIG. 1 is a process diagram showing the relationship between process requirements in a product development operation. [Figure 2] FIG. 2 is a diagram for explaining the process to which the requirements shown in the process diagram shown in FIG. 1 belong. [Figure 3] FIG. 2 is a diagram for explaining the process to which the requirements shown in the process diagram shown in FIG. 1 belong. [Figure 4] FIG. 2 is a diagram for explaining the process to which the requirements shown in the process diagram shown in FIG. 1 belong. [Figure 5] FIG. 2 is a diagram illustrating character strings that are highly similar to the requirements of the process diagram shown in FIG. [Figure 6] FIG. 1 is a block diagram illustrating an example of an information processing apparatus according to an embodiment. [Figure 7] FIG. 2 is a schematic diagram illustrating an example of a form. [Figure 8] FIG. 2 is a schematic diagram illustrating an example of a form. [Figure 9] FIG. 2 is a schematic diagram illustrating an example of a form. [Figure 10] FIG. 2 is a schematic diagram illustrating an example of a form. [Figure 11] FIG. 8 is a schematic diagram of a form that displays the process divisions to which items belong in the form shown in FIG. 7. [Figure 12] FIG. 12(a) is a structural diagram of items R1 to R9, and FIG. 12(b) is a table expressing the dependency correlations of items R1 to R9. [Figure 13] 13(a) and 13(b) are diagrams in which attributes of each item shown in FIG. 12(a) and FIG. 12(b) are added. [Figure 14] 12 is a schematic diagram of the form shown in FIG. 11, further displaying the process categories to which the items belong. [Figure 15] FIG. 2 is a schematic diagram illustrating the relationship between a process and a data model. [Figure 16]FIG. 16(a) is a diagram explaining the data model of business process P, FIG. 16(b) is a structural diagram of business processes P1 to P3, and FIG. 16(c) is a table expressing the dependency correlations of business processes P1 to P3. [Figure 17] FIG. 2 is a schematic diagram illustrating an example of a form. [Figure 18] FIG. 18(a) is a diagram explaining the data model of the report UI, FIG. 18(b) is a structural diagram of reports UI1 and UI2, and FIG. 18(c) is a table expressing the dependency correlation between reports UI1 and UI2. [Figure 19] Figure 19(a) is a diagram explaining a data model showing the correlation between a business process and the report UI used in the business process, Figure 19(b) is a structural diagram of business process P and the report UI, and Figure 19(c) is a table expressing the dependency correlation between business process P and the report UI. [Figure 20] 10 is a data model of each item ID-n of a form in a design basis process used in the information processing apparatus according to the embodiment. [Figure 21] FIG. 21 is a structural diagram of each item ID-n shown in FIG. 20. [Figure 22] Figure 22(a) shows the report UI and the data model of each item ID-n of the report in the design basis process, Figure 22(b) shows a portion of a structural diagram of the report UI and each item ID-n shown in Figure 20, which are in a mutually dependent relationship, and Figure 22(c) is a table expressing the dependent relationship between the report UI in Figure 22(b) and each item ID-n. [Figure 23] 1 is an example of a business process. [Figure 24] FIG. 24 is a diagram showing an example of a form used in each business process of FIG. 23. [Figure 25] An example of a data model corresponding to each business process shown in FIG. 23 and each report shown in FIG. 24 is shown below. [Figure 26] 10 is a flowchart illustrating the operation of the information processing device according to the present embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0025] Next, an embodiment of the present invention will be described with reference to the drawings. In the description of the drawings relating to the embodiment, the same or similar parts are designated by the same or similar reference numerals. However, it should be noted that the drawings are schematic, and the relationship between planar dimensions and the like may differ from the actual ones. Therefore, specific dimensions should be determined with reference to the following description. Furthermore, it goes without saying that the drawings may include parts with different dimensional relationships and ratios.

[0026] Furthermore, the embodiments are merely examples of devices and methods for embodying the technical idea of ​​the present invention, and the technical idea of ​​the present invention does not limit the configuration, arrangement, layout, etc. of each component to those described below. The technical idea of ​​the present invention can be modified in various ways within the technical scope defined by the claims.

[0027] (Embodiment) The information processing device of the present invention uses a user interface (UI) such as a form used in the business process to classify natural language used in the business process by taking advantage of the fact that the data to be focused on in the data model used in the information processing device of the present invention differs for each process. The data model used in the information processing device of the present invention is a data model based on attribute correlation and dependency correlation, which will be described later.

[0028] The data model used in the information processing device of the present invention is a model that represents the attribute correlations and dependency correlations that exist between a given natural language and the natural languages ​​that appear around it. In this embodiment, the natural language refers to a language that expresses requirements in a business process, and is composed of sentences, words, numbers, mathematical expressions, symbols, or a combination of these. The natural languages ​​that appear around the given natural language express requirements that are related to the requirements expressed in the given natural language in the business process, and details of these will be described later.

[0029] An attribute correlation is a correlation at a level of abstraction or resolution between a given natural language and the natural languages ​​that appear around it, which express attributes that describe the requirements expressed in the given natural language.

[0030] A dependency correlation is a correlation between a given natural language and the natural languages ​​that appear around it, which has a dependency relationship of either mechanism, parent-child relationship, or cause-and-effect. Attribute correlation and dependency correlation will also be discussed in more detail below.

[0031] First, a data model used in the information processing device of the present invention will be described. FIG. 1 shows a process diagram illustrating the relationships between requirements for each process division related to product development work. The diagram shown in FIG. 1 is a diagram related to the product development work of an electric kick scooter, as an example. A process diagram visualizes the flow of tasks in a business process. In FIG. 1, each requirement in the diagram (e.g., requirements 31 to 39), and each requirement in each process division of the business process, is shown with a solid or dashed line indicating the relationship between the requirements. The relationship between the requirements shown with dashed lines is a dependency correlation, and the relationship between the requirements shown with solid lines is an attribute correlation. The data model used in the information processing device according to this embodiment is a data model based on these dependency correlations and attribute correlations.

[0032] In Figure 1, requirements that are mutually dependent are connected with dashed lines and arranged horizontally. Requirements that are mutually attributed are connected with solid lines and arranged vertically. Furthermore, the horizontal axis lists the divisions of the business process along a rough timeline 21: "Requirements," "Performance / Function," "Logic," "Product," and "Means of Security."

[0033] Here, the business process divisions shown in FIG. 1 (referred to as process divisions in this embodiment) are obtained by classifying each business in a business process according to the content of the business. In FIG. 1, as examples of process divisions into which each business in a business process is classified, a requirement process division, a performance / function process division, a logic process division, a product process division, and a guarantee means process division are shown along the horizontal axis (time axis 21). As will be described later, in the information processing device of the present invention, the business process divisions are classified into R (requirements), F (functions), L (logical), P (physics), l (logic), and p (parameters). The requirement process division, performance / function process division, logic process division, product process division, and guarantee means process division in FIG. 1 correspond to R, F, L, P, and l, respectively. As shown in FIG. 1, each requirement belongs to one of the process divisions R, F, L, P, l, and p.

[0034] A dependency correlation is a relationship in which one requirement indicates the mechanism, parent-child relationship, cause-and-effect, etc. of another requirement. A dependency correlation will be specifically explained below.

[0035] When developing a product, the development work generally starts from the requirements process segment, as shown in Figure 1. Specifically, as an example, in the case of the development of a kick scooter, the development starts from the user's request, Requirement 31, "I want to make local transportation easier."

[0036] Requirement 34 lists "portability" as a function necessary to satisfy the user's request in requirement 31, "I want to make local travel easier." Requirement 35 lists "people" as the target of the "portability" in requirement 34. Since the object to be transported is "people" in requirement 35, requirement 36 lists an "electric board." Requirements 31, 34, and 35 are in a parent-child relationship or a cause-and-effect relationship. In this embodiment, such a parent-child relationship or cause-and-effect relationship is called a dependency correlation.

[0037] An attribute correlation is a relationship in which one requirement embodies another requirement. Multiple requirements that are in an attribute correlation form a hierarchical structure. This hierarchical structure corresponds to the vertical relationship between the requirements in Figure 1, and as you move vertically downward within the page of Figure 1, that is, as you advance through the layers, the level of abstraction decreases or the resolution increases, and the content of the requirements becomes more specific. The degree of attribute correlation between requirements in Figure 1 corresponds to this level of abstraction or resolution. Below, we will explain attribute correlation in detail.

[0038] Requirement 32, "Self-propelled means of transport operated by a handle," is listed as one of the specific means for realizing the user request in Requirement 31, "I want to make local travel easier." The specific content of Requirement 32 is listed as "a vehicle weighing 10 kg or less, measuring 100 mm, with a three-wheeled electric motor with a handle, capable of traveling at 10 km / h in two hours."

[0039] Requirements 31, 32, and 33 are mutually attribute-correlated and form a hierarchical structure, with the content becoming more specific as you move up the hierarchy from requirement 31 to requirement 32 to requirement 33. In other words, the level of abstraction decreases and the resolution increases as you move from requirement 31 to requirement 32 to requirement 33.

[0040] Specific details of the "portable function" in requirement 34 include the "steering function" in requirement 37, the "stopping function" in requirement 38, and the "rotation transmission function" in requirement 39. Requirement 34 and requirements 37 to 39 are also mutually attribute-correlated, with requirements 37 to 39 being specific embodiments of requirement 34. Note that requirements 37, 38, and 39 are not mutually dependent.

[0041] Next, we will explain the business processes to which natural language, which is the target of classification by the information processing device according to this embodiment, belongs. As an example, a ledger, which is a document related to a company's management activities, organizes and stores the above-mentioned requirements for each business process. For example, in the case of a product specification, in the early stages of product development, the details of the specification review process enclosed by solid line 41 in FIG. 2 are described in the product specification. The specification review process enclosed by solid line 41 mainly includes information on the requirement process category. In the intermediate stages of product development, the details of the detailed design process enclosed by solid line 42 in FIG. 3 are described in the product specification. The detailed design process mainly includes specific information such as the performance / function process category, logic process category, and product process category. In the final stages of product development, the details of the evaluation process enclosed by solid line 43 in FIG. 4 are described in the product specification. The evaluation process mainly includes information such as the product process category and the security means process category.

[0042] As mentioned above, it is determined in advance which process category (R, F, L, P, l, and p) each requirement recorded on the form used for each business process belongs to, and the elements and process categories included in each business process. This will be explained in detail later.

[0043] Here, the natural language used in the text recorded in a document or the like may differ depending on the company, department, person in charge, etc., even if the natural language relates to the same requirement of the same process. For example, in the case of a specific requirement of a specific process division in product development work, the natural language used in the text recording the content of the requirement may differ depending on the person in charge, even if the content is the same.

[0044] As shown in Figure 5, the natural language expressing each requirement shown in Figure 1 may be replaced with natural language that is highly similar to each requirement and is written around each requirement in Figure 5. Figure 5 explains the natural language that is highly similar to each requirement in the process diagram shown in Figure 1, and is a diagram in which a part of Figure 1 is enlarged and natural language that is highly similar to each requirement is written around the requirement shown in Figure 1. For example, requirement 31, "I want to make local transportation easy," can be expressed using character strings such as "neighborhood," "walking," "easy," "walking," "commute," and "tough." Similarly, requirement 34, "portability," can be expressed using character strings such as "rideable" and "transport," and requirement 32, "self-propelled transportation means operated by handle," can be expressed using character strings such as "handle," "automatic machine," "electric driving," and "operating lever."

[0045] In business processes, when data is aggregated to be passed to tools or systems for business execution, variations in spelling, such as those mentioned above, exist. Therefore, in order to classify and organize the data, measures such as normalization are required to address the variations in spelling, but this requires human intervention to read the vast amount of data. While there are name matching technologies that utilize thesaurus data and distributed word representations, the accuracy is still insufficient. The information processing device according to this embodiment classifies and organizes the natural language expressions that express each requirement in the process diagram shown in FIG. 1, as shown in FIG. 5, and natural language expressions that are highly similar to these. In this case, the classification and organization is performed using the business process to which each requirement belongs and the data model.

[0046] The operation of the information processing device according to this embodiment will now be described. Fig. 6 is a block diagram showing the configuration of an information processing device 10 according to this embodiment. As shown in Fig. 6, the information processing device 10 according to this embodiment is composed of a CPU 11 for executing various calculations, a storage unit 12 for storing processing programs, data, etc., an I / O (input / output interface) 13, a display unit 14, and an input unit 15.

[0047] The I / O 13 is an interface, buffer, etc. for communication (transmission and reception).

[0048] The display unit 14 is a display device such as a display, and displays the results of calculations performed by the CPU 11, etc.

[0049] The input unit 15 is an input device such as a keyboard or a mouse, and is realized by a device that receives input from a user and transmits it to the CPU 11.

[0050] The block diagram in Figure 6 shows the functional units within the CPU 11. When each functional unit of the CPU 11 is realized by software, the CPU 11 realizes it by executing instructions of a program, which is software that realizes each function. In detail, the CPU 11 includes an acquisition unit 111, an extraction unit 113, a reading unit 115, and a registration unit 114. The storage unit 12 also includes a dataset storage unit 121 and a document information storage unit 122.

[0051] The acquisition unit 111 acquires a predetermined document via the input unit 15. In this embodiment, the document acquired by the acquisition unit 111 is a document that may occur in a business process, such as a ledger sheet, but is not limited to this and may be, for example, a statement, a mathematical formula, a data file, etc. Furthermore, the document acquired by the acquisition unit 111 does not necessarily have to include two or more words or a compound word, and may be a single word or a compound word.

[0052] Furthermore, the predetermined document may be saved not only in text format but also in other formats such as image, audio, etc. The extraction unit 113 may be configured to extract character strings according to the format of each document, and may, for example, recognize a sentence by image recognition, audio recognition, etc., and then extract character strings from the sentence. In this embodiment, the predetermined document is a form.

[0053] The document acquired by the acquisition unit 111 may be tagged with the type of the document and / or the business process to which the document belongs. Alternatively, when the acquisition unit 111 acquires a document, the acquisition unit 111 may acquire the type of the document and / or the business process to which the document belongs, for example, through a user's input via the input unit 15. Alternatively, the formats of the reports that may occur in the business processes often have common items for each business process, and therefore the acquisition unit 111 may automatically determine the type of the acquired document and / or the business process to which the document belongs, based on the acquired document.

[0054] Examples of specified documents are shown in Figures 7 to 10. Figure 7 is a product planning document, Figure 8 is a product specification document, Figure 9 is a product catalog, and Figure 10 is a complaint management table and service report. The forms shown in Figures 7 to 10 are, for example, forms related to business processes such as the product planning process, specification review process, detailed design process, and evaluation process, respectively.

[0055] In this embodiment, the matters described in a specified document consist of one or more items. An item in a specified document is an item in the specified document that, for example, explains at least a part of a business process. It typically consists of an item name and a field for describing the item's content. However, the item name may be omitted. Hereinafter, the name of an item is referred to as the item name, and the content of an item is referred to as the item content. The product planning document shown in FIG. 7 includes items such as a planning overview and a cost comparison table, and the respective contents are entered in the respective fields for describing the item content. The product specification sheet shown in FIG. 8 includes item names such as product name, product number, and product weight. The product catalog shown in FIG. 9 includes item names such as product name, purpose, and weight. The complaint management table and service report shown in FIG. 10 include item names such as a complaint management number and model number. As in FIG. 7, the respective fields for describing the contents of the items are entered.

[0056] The items listed in the forms shown in Figures 7 to 10 and the content entered into each item are the same as or similar to the content listed in each requirement in the process diagrams shown in Figures 1 to 4. For example, the item names and content of each item in the plan overview and cost comparison table listed in the product planning document shown in Figure 7 are rewritten to match the content of requirements 31, 32, and 33 in the requirements process in the process diagram shown in Figure 1. The items such as product name, product number, and product weight listed in the product specifications shown in Figure 8 are rewritten to match the content of the requirements for the performance and function process in the process diagram shown in Figure 1.

[0057] As described above, the item names and contents of each item written in the forms used in the business process correspond to the contents written in each requirement of the process diagram shown in Figures 1 to 4. As explained above with reference to Figures 2 to 4, the forms organize and store the requirements of the process diagram for each business process, and the requirements of the process diagram that are written as items in the form are roughly predetermined depending on each business process. The requirements of the process diagram are mutually dependently correlated or attribute-correlated, as shown by dashed or solid lines in Figure 1. Therefore, the items and contents of each item of the form are mutually dependently correlated or attribute-correlated, just like the requirements of the process diagram.

[0058] The product planning document shown in Fig. 11 shows information about which process each item in the product planning document shown in Fig. 7 belongs to among the process divisions R, F, L, P, l, and p in the process diagram shown in Fig. 1, and which other items in the document have a dependency correlation with it. In Fig. 11, R indicates that the item belongs to the process division R. The same is true for F, L, P, l, and p. In addition, R (parent) and R (child) show information about items that have a dependency correlation with the item.

[0059] As shown in Figure 11, the project overview items show process categories R, F, L, and P, and each process category is further denoted by R (parent) R (child), F (parent) F (child), L (parent) L (child), or P (parent) P (child). This indicates that one or more natural languages ​​included in the content of the project overview items belong to one of the process categories R, F, L, or P, and that information indicating which items they are dependently correlated with is linked to them. In the cost comparison table, items corresponding to models are denoted by PP (parent) P (child), and items corresponding to prices are denoted by RR (parent) R (child), indicating that each item belongs to the process categories P and R, and that information indicating which items they are dependently correlated with is linked to them.

[0060] The dependency correlation expressed by R (parent) R (child) of RR (parent) R (child) shown in Fig. 11 will be explained with reference to Fig. 12(a) and Fig. 12(b). Fig. 12(a) shows a structural diagram of items R1 to R9 that belong to process partition R and have a dependency correlation with each other, and Fig. 12(b) shows a table expressing the dependency correlation.

[0061] 12(a) indicates that the items connected to the solid lines are in a parent-child relationship, i.e., a dependency correlation, and for example, item R1 has no parent and its children are items R2, R7, and R8. Item R2's parent is item R1 and its children are items R3 and R6.

[0062] The requirement item list shown in Fig. 12(b) shows at least some of the items belonging to the process classification of requirement (R), and items R1 to R9 are shown. The requirement item correlation table shown in Fig. 12(b) shows some of the dependency correlations between the items shown in the requirement item list. For example, correlation 001 indicates the dependency correlation between items R1 and R2, where item R2 is the child when item R1 is the parent, and correlation 002 indicates the dependency correlation between items R1 and R8, where item R8 is the child when item R1 is the parent.

[0063] The dependency relationships shown in FIGS. 11 and 12 are shown in the process diagram of FIG. 1 as relationships between items connected by dashed lines in the horizontal direction within the page.

[0064] Furthermore, for each item, an attribute that explains the item is called an attribute correlation. This attribute correlation is shown in Fig. 1, but not in Figs. 11 and 12. The attribute correlation will be explained with reference to Figs. 13(a) and 13(b). Figs. 13(a) and 13(b) are diagrams in which the attributes of each item have been added to the structural diagram of items R1 to R9 and the required item list shown in Figs. 12(a) and 12(b).

[0065] Referring to FIG. 13(a), for example, item R1 shows the attributes "requester: customer," "importance: level 3," and "occurrence date: month / day / 2024." Similarly, item R2 shows the attributes "requester: customer," "importance: level 1," and "occurrence date: month / day / 2024." Referring to FIG. 13(b), the attributes of each of items R1 to R9 are listed.

[0066] The attribute correlations shown in FIG. 13 are shown in the process diagram of FIG. 1 as relationships between items connected by solid lines in the vertical direction within the page.

[0067] For each item of each form in each such business process, a document information table describing the correspondence between each business process, each form, each item, the process classifications R, F, L, P, l, and p to which each item belongs that can be linked to each item, and items that have mutual dependency and / or attribute relationships with each item that can be linked to each item is stored in advance in the document information storage unit 122 described later, and the information described in the document information table is read out by the reading unit 115 described later.

[0068] The extraction unit 113 extracts natural language from a predetermined document and transmits it to the reading unit 115. The extraction unit 113 may extract any word or compound word from the predetermined document, or may extract all or a predetermined number of words and / or compound words contained in the predetermined document. The natural language extracted by the extraction unit 113 from the predetermined document is the content entered in a field for describing the content of an item described in the predetermined document, and may be part or all of the content of the item described in the predetermined document. Alternatively, when the content of the item is a sentence, such as the project summary item shown in FIG. 7, the natural language may be extracted from the content of the item.

[0069] When extracting natural language from a predetermined document, the extraction unit 113 acquires the name of the item from which the natural language was extracted, the business process to which the predetermined document belongs, and the type of the predetermined document, and sends them to the reading unit 115. If the predetermined document is a form, the type of the predetermined document is, for example, a product plan, product specification, product catalog, complaint management sheet, service report, etc., and if the predetermined document is not a form, it is, for example, a calculation sheet, formula, data file, etc.

[0070] When acquiring the business process to which a predetermined document belongs and the type of the predetermined document, if the predetermined document is tagged with the document type and / or the business process to which the document belongs, the extraction unit 113 may acquire the business process and the type of the predetermined document from the information tagged with the predetermined document. If the acquisition unit 111 acquires the document type and / or the business process to which the document belongs via the input unit 15, or if the acquisition unit 111 automatically determines the acquired document type and / or the business process to which the document belongs, the extraction unit 113 may receive the business process and the type of the predetermined document from the acquisition unit 111.

[0071] When the extraction unit 113 extracts multiple natural languages ​​from a predetermined document, for example, the extraction unit 113 may store the extracted multiple natural languages ​​in the storage unit 12 or the like, and read out the stored multiple natural languages ​​one by one in order and transmit them to the reading unit 115 and the registration unit 114. After the reading unit 115 and the registration unit 114 perform processing on the natural languages ​​received from the extraction unit 113, the reading unit 115 and the registration unit 114 may repeat the processing for all of the stored multiple natural languages.

[0072] After receiving the natural language extracted from the specified document by the extraction unit 113, the name of the item from which the natural language was extracted, the business process to which the specified document belongs, and the type of the specified document, the reading unit 115 searches the document information table stored in the document information storage unit 122 to determine whether information identical and / or similar to at least any of the received natural language, name of the item, business process, and type of the specified document exists in the document information table.

[0073] Here, when the reading unit 115 determines whether similar information exists in the document information table, the criteria for determining similarity are defined by one of the following criteria or a combination thereof. The reading unit 115 may quantify each similarity based on each criterion when determining whether similar information exists in the document information table based on each of the following criteria. For example, the reading unit 115 may determine that the similarity is high when each of the similarities based on each criterion exceeds one or more predetermined thresholds. Alternatively, the reading unit 115 may determine that the similarity is high when the sum of the similarities based on each criterion is the total similarity and the total similarity exceeds a second predetermined threshold, and / or the similarity is low when the total similarity is lower than a third predetermined threshold. Note that the determination threshold may be given as an absolute value. Alternatively, the determination threshold may be derived as a relative value obtained as a result of learning. Furthermore, weights or coefficients such as probability or importance may be applied. The predetermined threshold and the second predetermined threshold may be stored in the storage unit 12, or may be acquired via the input unit 15 by input by the user, for example.

[0074] The criterion for determining whether information is similar is whether the business process to which the predetermined document from which the extraction unit 113 extracted natural language belongs is the same as or similar to the business process to which the information listed in the document information table belongs. Here, "close" refers to whether the vectors are close when the natural language constituting the names of the business processes are vectorized, whether the vectors are close when the natural language and the natural language in the attribute correlation and dependency correlation are vectorized, and / or whether the order of the business processes themselves is close. Similarly, the term "close" used in the similarity determination criteria described below refers to whether the vectors are close when the natural language constituting the names of the objects to be determined for similarity are vectorized, whether the vectors are close when the natural language and the natural language and the attribute correlation and dependency correlation are vectorized, and / or whether the order of the objects to be determined for similarity is close. Furthermore, the business process to which the predetermined document from which the extraction unit 113 extracted natural language belongs and the business process to which the information listed in the document information table belongs are the same as each other in terms of parent-child dependency correlation or attribute correlation. The parent and child of the dependency correlation from each perspective are, for example, as will be described later, if the business process to which a specific document from which the extraction unit 113 extracted natural language belongs in Figure 15 is business process P2, the parent of the dependency correlation is business process P1 and the child is business process P3.

[0075] The criterion for determining that the types of predetermined documents are similar is whether the type of the predetermined document from which the extraction unit 113 extracted natural language is the same as or close to the type of document from which the information recorded in the document information table is described. Furthermore, since the type of document is determined by the business process, the criterion for determining the similarity of the types of predetermined documents may be whether the business process to which the predetermined document from which the extraction unit 113 extracted natural language belongs and the business process to which the information recorded in the document information table belongs are the same as each other in the parent-child relationship of the dependency correlation or the attribute correlation.

[0076] The criteria for determining whether item names are similar is whether the name of the item from which the extraction unit 113 extracted natural language information is the same as or similar to the name of the item from which information is recorded in the document information table. Alternatively, the name of the item from which the extraction unit 113 extracted natural language information and the name of the item from which information is recorded in the document information table share at least some of the names of parent and child items in a dependency correlation or high-level and low-level items in an attribute correlation. For example, in the process diagram shown in FIG. 1, high-level and low-level items in an attribute correlation correspond to requirements located vertically above and below a given requirement, which are directly connected by a solid line and have an attribute correlation with the given requirement. For example, if the name of the item from which the extraction unit 113 extracted natural language information is requirement 32, "self-propelled vehicle operated by handle operation," then the high-level item in the attribute correlation is requirement 33, and the low-level item in the attribute correlation is requirement 31. Hereinafter, requirements with high and low abstraction levels for attribute correlations will be expressed as before and after attribute correlations.

[0077] The criterion for determining whether process divisions are similar is whether the process division to which the natural language extracted by the extraction unit 113 belongs is the same as or close to the process division to which the information described in the document information table belongs, or whether the process division and the information belong to a parent-child relationship in a dependency correlation or an attribute correlation.

[0078] The criterion for determining that the dependency correlation is similar is whether the natural language extracted by the extraction unit 113 and the parent and child requirements of the dependency correlation are similar to the information described in the document information table.

[0079] The criterion for determining that the attribute correlation is similar is whether the natural language extracted by the extraction unit 113 and the conditions before and after the attribute correlation are close to the information described in the document information table.

[0080] The criterion for determining that the natural language extracted by the extraction unit 113 and the information listed in the document information table are similar in terms of vectors is whether they are similar in terms of distributed representations of words in existing technology or similar in terms of document vectors.

[0081] The criterion for determining that natural language strings are similar is that, among the strings constituting the natural language or item names extracted by the extraction unit 113 from a specified document in the document information table, there is a natural language or item name with at least a specified number of characters that is the same, or there is a natural language or item name with at least a specified string that is the same.

[0082] The above criteria may be weighted or multiplied by a coefficient, such as a probability or importance. For example, if the dependency correlation or attribute correlation is similar, the natural language extracted by the extraction unit 113 and the information described in the document information table are considered to have a high degree of similarity. Also, for example, if the business processes are similar, the similarity is considered to be low based on the similarity of the business processes alone. By using weights or coefficients based on the degree of similarity when only each of the above criteria is satisfied, the degree of similarity when one or more criteria are satisfied can be expressed as a numerical value. This numerical value can be used by the registration unit 114 when determining whether to register the information extracted by the extraction unit 113 in a dataset.

[0083] When the reading unit 115 determines that there is information in the document information table that is identical and / or similar to at least one of the natural language, the name of the item, the business process, and the type of the specified document, it reads out information about the process classifications R, F, L, P, l, and p to which the item belongs, and the items that have a dependency correlation and / or attribute correlation with each item, which are listed in the document information table and are linked to the identical and / or similar information in the document information table, and transmits this information to the registration unit 114 together with the natural language, the name of the item from which the natural language was extracted, the business process to which the specified document belongs, and the type of the specified document.

[0084] Here, the reading unit 115 may determine that only information determined to have a high degree of similarity is similar information, or may determine that the degree of similarity is high if the total degree of similarity is higher than a second predetermined threshold. And / or, for example, if the reading unit 115 determines that information similar to at least one of natural language, item name, business process, and predetermined document type exists in the document information table, the reading unit 115 may notify the user of the determination result by displaying the similar information on the display unit 14 along with the degree of similarity for each of the one or more determination criteria described above, and may obtain a user's input via the input unit 15 as to whether to transmit the similar information to the registration unit 114 and register it in the dataset, and transmit the similar information to the registration unit 114 based on the obtained determination. When notifying the user of the determination result described above, the reading unit 115 may display an alert on the display unit 14, etc.

[0085] If the reading unit 115 determines that the document information table contains information identical to at least one of the natural language, the name of the item, the business process, and the type of the specified document, the reading unit 115 may send to the registration unit 114 the information read from the document information table, the natural language, the name of the item from which the natural language was extracted, the business process to which the specified document belongs, and the type of the specified document, as well as information indicating which of the natural language, the name of the item, the business process, and the type of the specified document in the document information table the information was identical to.

[0086] If the reading unit 115 determines that there is information in the document information table that is similar to at least one of the natural language, the name of the item, the business process, and the type of specified document, the reading unit 115 may transmit to the registration unit 114 the information read from the document information table, the natural language, the name of the item from which the natural language was extracted, the business process to which the specified document belongs, and the type of specified document, as well as which information the information was similar to, and / or the judgment results based on each of the one or more judgment criteria described above, and / or the reading unit 115 may transmit to the registration unit 114 the similarity of each of the one or more judgment criteria described above, and / or the total similarity to the registration unit 114.

[0087] When the reading unit 115 determines that there is no information in the document information table that is identical or similar to at least one of the natural language, the name of the item, the business process, or the type of the specified document, the reading unit 115 may transmit the determination result that there is no identical or similar information to the registration unit 114, together with the natural language, the name of the item from which the natural language was extracted, the business process to which the specified document belongs, and the type of the specified document.

[0088] The registration unit 114 registers in the dataset the natural language received from the reading unit 115, the name of the item from which the natural language was extracted, the business process to which the specified document belongs, the type of the specified document, the process classification R, F, L, P, l, and p to which the item belongs, which is listed in the document information table, and items that have a dependency correlation and / or attribute correlation with the item.

[0089] In addition to these, the registration unit 114 may register information such as the date and time the natural language was registered, information on the sentence, document, business process, etc. in which the natural language was confirmed, and the number of times the character string was confirmed in the dataset.

[0090] Furthermore, in addition to these, the registration unit 114 may transmit to the registration unit 114 the information read from the document information table received from the reading unit 115, the natural language, the name of the item from which the natural language was extracted, the business process to which the specified document belongs, and the type of specified document, as well as information indicating which information in the document information table the natural language, the name of the item, the business process, and the type of specified document was identical to, and / or which information it was similar to, and / or the judgment results based on each of the one or more judgment criteria described above, and / or the reading unit 115 may transmit to the registration unit 114 the similarity of each of the one or more judgment criteria described above, and / or register the total similarity in the dataset.

[0091] When the registration unit 114 receives from the reading unit 115 a determination result that identical or similar information does not exist, the natural language, the name of the item from which the natural language was extracted, the business process to which the specified document belongs, and the type of the specified document, the registration unit 114 registers the natural language, the name of the item from which the natural language was extracted, the business process to which the specified document belongs, and the type of the specified document in a dataset.

[0092] Furthermore, the registration unit 114 may register in the document information table the natural language received from the reading unit 115, the name of the item from which the natural language was extracted, the business process to which the specified document belongs, the type of the specified document, the process classifications R, F, L, P, l, and p to which the item belongs, and items that have a dependency correlation and / or attribute correlation with the item, which are listed in the document information table. The reading unit 115 searches the document information table and reads out data listed in the document information table, so by the registration unit 114 registering data in the document information table as described above, the reading unit 115 can search for and read out data from a document information table in which a larger amount of data has been accumulated.

[0093] The dataset storage unit 121 stores a dataset. The dataset includes natural language, the name of the item from which the natural language was extracted, the business process to which a specific document belongs, the type of the specific document, the process classifications R, F, L, P, l, and p to which the item belongs, which are listed in the document information table, and items that have a dependency correlation and / or attribute correlation with the item. The dataset may be created in advance and stored in the dataset storage unit 121, or may be generated by the information processing device according to this embodiment.

[0094] The document information storage unit 122 stores a document information table that describes the correspondence between each item of each form in each business process, each business process, each form, each item, the process classifications R, F, L, P, l, and p to which each item belongs that can be linked to each item, and items that have a dependency correlation and / or attribute correlation with each item that can be linked to each item. The document information table is created in advance and stored in the document information storage unit 122.

[0095] The document information table is created in advance based on experience, for example, by a person in charge of creating reports, but as described in the second embodiment, once data is accumulated in a dataset, it becomes possible to change the information recorded in the document information table and the information recorded in the dataset to more appropriate information based on the data accumulated in the dataset.

[0096] By repeating a series of operations in the information processing device according to this embodiment, including acquisition of a predetermined document by the acquisition unit 111, extraction of natural language by the extraction unit 113, search and readout by the readout unit 115, and registration in a dataset by the registration unit 114, natural language is accumulated in a dataset along with item names, business processes, predetermined document types, process categories to which the items belong, and items in dependency correlations and / or attribute correlations with the items. Accumulating data facilitates the extraction and inference of words and sentences with higher similarity probabilities, enabling statistics and learning from the data. This enables the data to be utilized for various purposes, such as the collection and organization of training data used in machine learning, name matching, and data management. When classifying natural language, the information processing device according to this embodiment uses not only item names, business processes, and predetermined document types, but also process categories to which the items belong and items in dependency correlations and / or attribute correlations with the items, thereby improving the accuracy of statistics and learning using data.

[0097] The product planning document shown in Fig. 14 shows natural language corresponding to each requirement shown in the process diagram, which is included in the sentences written in the planning overview section of the product planning document shown in Fig. 11, and for each of these natural languages, the process categories R, F, L, P, l, and p to which it belongs, and information on items that have a dependency correlation and / or attribute correlation with each item that can be linked to each item. Specifically, for example, it shows that the compound word "market of 20 trillion yen scale" belongs to process category R, ​​and is linked to items that have a dependency correlation with "market of 20 trillion yen scale."

[0098] When extracting natural language corresponding to each requirement shown in the process diagram from the text written in the planning overview section of the product planning document shown in Figure 11, more accurate extraction is possible when natural language processing is performed on the text, and extraction is performed using natural language processing on the text, together with information on the process category to which the content written in the planning overview section belongs, and on items that have dependency correlations and attribute correlations, compared to extraction using only natural language processing on the text.

[0099] The operation of the information processing device according to this embodiment, which has been specifically described with reference to FIGS. 1 to 14, will now be described in an abstract manner with reference to FIGS.

[0100] Fig. 15 shows a schematic diagram illustrating the relationship between business processes and data models, which will be described later. Fig. 15 shows data models 131 to 135 corresponding to multiple business processes P1, P2, P3, P4, and P5, respectively. As an example, the business processes P1 to P5 shown in Fig. 15 are arranged in the order of the progress of the business processes, with business processes P1 to P3 being a planning process, business process P4 being a design review process, and business process P5 being a testing process. The relationships between the business processes P1 to P5 are as follows: n Business Process P, one step before n-1 is business process P n Parent of Business Process P n Business Process P after one step of n+1 is business process P n 15, if business process P2 is the process in question, the parent process of business process P2 is business process P1, and the child process is business process P3.

[0101] The process divisions that require attention in each business process differ depending on the order of the stages in the business process and the dependency relationships between the business process and other business processes. Here, the process divisions that require attention refer to the process divisions to which the requirements shown in Figure 1 belong, or the process divisions to which the items shown in Figure 11 belong, that are most frequently included in a given business process, and are any of the process divisions R, F, L, P, l, and p. For example, in Figure 15, the process division that requires attention in business process P1, which is first in the order of the stages in the business process, is R. In the case of a planning process, the process division that requires attention in business processes P1 to P3 is R.

[0102] With reference to Figures 2 to 4, we have already given examples in which the product specification, which is a document used in the specification review process, mainly contains information on the requirement process category, the product specification, which is a document used in the detailed design process, mainly contains information on the performance / function process category, logic process category, and product process category, and the product specification, which is a document used in the evaluation process, mainly contains information on the security means process category. The above-mentioned noteworthy items correspond to the requirement process category in the specification review process, the performance / function process category, logic process category, and product process category in the detailed design process, and the security means process category in the evaluation process in the examples shown in Figures 2 to 4.

[0103] Fig. 16(a) shows a data model of business process P used in the information processing device according to this embodiment, Fig. 16(b) shows an example structural diagram of business processes P1 to P3, which are interdependently related to one another, and Fig. 16(c) shows a table expressing the interdependence of business processes P1 to P3 in Fig. 16(b). PID-1, PID-2, ... in Fig. 16(a) represent interdependence correlations between business processes P1 to P3 shown in Fig. 16(b) and correspond to business process correlation_IDs shown in Fig. 16(c). Business process P in Fig. 16(a) corresponds to the process ID in the process item list in Fig. 16(c), and P(parent) and P(child) in Fig. 16(a) correspond to "parent" and "child" in the process correlation table in Fig. 16(c).

[0104] An example of a form used in a business process is shown in Fig. 17. Fig. 17 shows a business process P in which a form UI is used, each item (user interface) UI-n (n is an integer equal to or greater than 1) written in the form UI, and process classifications R, L, and P to which the natural language entered in item UI-n belongs.

[0105] FIG. 18(a) shows a data model of a form UI used in a business process in an information processing device according to this embodiment. FIG. 18(b) shows an example of a structure diagram of forms UI1 and UI2, which are interdependently related to each other. FIG. 18(c) shows a table expressing the interdependence between forms UI1 and UI2 in the business process shown in FIG. 18(b). As in the case of the business process shown in FIG. 16, UI-ID-1, UI-ID-2, ... in FIG. 18(a) represent the interdependence between the forms UI1 and UI2 shown in FIG. 18(b) and correspond to the process UI correlation_ID shown in FIG. 18(c). The form UI in FIG. 18(a) corresponds to the process UI in the process item list in FIG. 18(c), and the UI(parent) and UI(child) in FIG. 18(a) correspond to the "parent" and "child" in the process correlation table in FIG. 18(c).

[0106] For an item UI-n described in a form, there exist parent UI-n and child UI-n that are in a dependent relationship with the item UI-n, just like the business process P and the form UI. The parent UI-n and child UI-n that are in a dependent relationship with the item UI-n may be described in the same form as the item UI-n, or they may be described in different forms. Each form usually corresponds to a business process, and as shown in Figures 2 to 4, the requirements included in each business process are roughly determined for each business process. Therefore, the parent UI-n and child UI-n that are in a dependent relationship with the item UI-n may be described in a form used in the same business process as the item UI-n, or they may be described in a form used in a business process different from the item UI-n.

[0107] FIG. 19(a) shows a data model used in the information processing device according to this embodiment, illustrating the correlation between a business process and a form UI used in the business process. FIG. 19(b) shows an example of a structure diagram of a business process P and a form UI, which are interdependently correlated with each other. FIG. 19(c) shows a table expressing the interdependent correlation between the business process P and the form UI in FIG. 19(b). Assume that the form used in business process P3 is form U1. P-UI-ID-1, P-UI-ID-2, ... in FIG. 19(a) represent the interdependent correlation between business process P3 and form U1 shown in FIG. 19(b) and correspond to the process UI correlation_ID shown in FIG. 19(c). The form UI in FIG. 19(a) corresponds to the PU correlation in FIG. 19(c), and the UI(parent) and UI(child) in FIG. 19(a) correspond to the "parent" (P3) and "child" (U1) in the process correlation table in FIG. 19(c).

[0108] Fig. 20 shows a data model indicating the correlation between items of each item ID-n of a form in a design basis process as an example of a business process used in the information processing device according to this embodiment, and Fig. 21 shows a structural diagram of each item ID-n shown in Fig. 20, which are in a mutually dependent correlation. The ID-n shown in Fig. 21 is the dependent correlation between each element shown in Fig. 20.

[0109] Figure 22(a) shows a data model showing the correlation between the report UI used in the business process and each item ID-n of the report in the design basis process. Figure 22(b) shows a portion of a structural diagram of the report UI and each item ID-n shown in Figure 20, which are in a mutually dependent correlation. Figure 22(c) shows a table expressing the dependent correlation between the report UI in Figure 22(b) and each item ID-n. The UI shown in Figure 22(a) is a dependent correlation between each element shown in Figure 22(b). As shown in Figure 22, it can be seen that there is a dependent correlation between the report UI and report item R1.

[0110] As shown in Figures 17 to 22, there can be dependency correlations between elements of the same type, such as between business processes of multiple business processes P, between forms of multiple form UIs, and between items of each item ID-n of a form. However, dependency correlations are not limited to between elements of the same type; they can also exist between elements of different types, such as between a business process P and a form UI, or between a form UI and each item ID-n of a form. On the other hand, attribute correlations refer to correlations with attributes that explain specific items. Although attribute correlations are not shown in Figures 17 to 22, attribute correlations differ from dependency correlations in that elements in an attribute correlation are of the same type, and attribute correlations do not exist between elements of different types.

[0111] The data model for each business process will be described with reference to Figures 23 to 25. As shown in Figure 23, business processes P1, P2, and P3 are exemplified by a planning business process, a design review business process, and a design guarantee and verification business process.

[0112] Figure 24 shows examples of forms used in each business process shown in Figure 22. In the planning business process, design review business process, and design assurance and verification business process, a proposal document, a design review document, and design verification materials are used as forms U1, U2, and U3, respectively. Each of forms U1, U2, and U3 has items U1-1, U1-2..., U2-1U2-2..., U3-1, U3-1.... As shown in Figure 23, each of forms U1, U2, and U3 has items that are common to other forms and items that are not common to other forms.

[0113] Figure 25 shows an example of a data model corresponding to each business process shown in Figure 23 and each report shown in Figure 24. It shows that each item U1-1, U1-2..., U2-1U2-2..., U3-1, U3-1... shown in Figure 24 has items belonging to processes R, F, L, P, l, and p. As with Figure 24, each of the data models shown in Figure 25 has items that are common to other data models and items that are not common to other data models. The data model of the entire business process based on the three data models shown in Figure 25 is shown in Figure 22.

[0114] Next, the operation of the information processing device according to this embodiment will be described with reference to the flowchart of FIG.

[0115] In step S801, the acquisition unit 111 acquires a predetermined document.

[0116] In step S802, the extraction unit 113 extracts at least one of the natural language from the specified document, the name of the item in the specified document from which the natural language was extracted, the business process to which the specified document belongs, and the type of the specified document.

[0117] In step S803, the reading unit 115 searches the document information table to determine whether information identical and / or similar to the received natural language and at least one of the item name, business process, and specified document type exists in the document information table, and based on the result of the determination, reads information listed in the document information table about the process category to which the item belongs and items that have a mutual dependency correlation and / or attribute correlation with the item.

[0118] In step S804, the registration unit 114 registers information about the natural language, the name of the item, the business process, the type of specified document, the process category to which the item belongs, and items that have a dependency correlation and / or attribute correlation with the item in the dataset.

[0119] (Second embodiment) When data is accumulated in the dataset, it becomes possible to change the information recorded in the document information table to be more appropriate based on the data accumulated in the dataset. As described above, the document information table is created in advance based on experience, for example, by a person in charge of creating forms, but the correspondences recorded in the document information table between each business process, each form, each item, the process classification to which each item belongs that can be linked to each item, and the items that can be linked to each item and have a dependency correlation and / or attribute correlation with each other are not necessarily optimal at the time the document information table is created.

[0120] By performing statistics and organizing the data stored in the dataset, it is possible to change the information listed in the document information table to make it more appropriate. For example, it is possible to find commonalities and characteristics between the natural language associated with a specific item and the natural language associated with another specific item, and appropriately modify the correlation between a specific item and another specific item based on those commonalities and characteristics.

[0121] The information processing device according to this embodiment may further include a data processing unit that performs statistics and organization on the data stored in the dataset. The data processing unit may sort each of the data, for example, each business process, each report, each item, the process classification to which each item belongs that can be linked to each item, and items that have a dependency correlation and / or attribute correlation with each item that can be linked to each item, and extract fluctuations in other data relative to the sorted data, or may extract commonalities and characteristics between other data. After extracting the data fluctuations, commonalities and characteristics between other data, the data processing unit may prompt the user to check, organize, and correct the document information table, for example, by displaying the extracted results on the display unit 14.

[0122] (Application example) The information processing device according to this application example may further include a name identification unit, as compared with the information processing device according to the embodiment. In the name identification unit, the acquisition unit 111 acquires a predetermined document, the extraction unit 113 extracts a natural language from the predetermined document, and the reading unit 115 reads, from the document information table, information about each business process, each form, and each item that can be linked to the item, the process to which the item belongs, and the items that have a dependency relationship and / or attribute relationship with the item. The name identification unit then searches the dataset to determine whether a natural language highly similar to the extracted natural language is registered in the dataset. When searching the dataset, the name identification unit performs the search based on the natural language extracted by the extraction unit 113 as well as the information read by the reading unit 115. Among the natural languages ​​registered in the dataset, a natural language that has the same process, among processes R, F, L, P, l, and p, to which the item from which the natural language was acquired belongs and the same items that have a dependency relationship and / or attribute relationship with the item can be determined to have a high similarity to the extracted natural language.

[0123] The information processing device according to this application example may include a data name similarity determination unit in addition to name matching using processes and data models. When aggregating data in business processes to perform name matching, the data name similarity determination unit performs name matching based on the similarity of data names.

[0124] The information processing device according to this application example may further include a data similarity determination unit that, when aggregating data in a business process to perform name matching, performs the name matching based on similarities in data format, etc.

[0125] The information processing device according to this application example may further include a document information generation unit, which may acquire a document by the acquisition unit 111, extract all words or compound words contained in the document by the extraction unit 113, replace all words or compound words contained in the document with natural language words determined by the name matching unit to have high similarity, generate document information describing the replaced document, and store the generated document information in the storage unit 12.

[0126] The information processing device according to this application example may further include a document tagging unit, which may perform the following steps: acquire a document by the acquisition unit 111, extract all words or compound words contained in the document by the extraction unit 113, extract character strings determined to have high similarity by the name matching unit from a dataset, tag all words or compound words contained in the document with the character strings determined to have high similarity by the name matching unit, and then generate document information describing the document in which the tagged character strings have been replaced; and store the document information in the storage unit 12.

[0127] The present invention naturally includes various embodiments not described herein. Therefore, the technical scope of the present invention is defined only by the invention-specifying matters according to the scope of the claims that are appropriate from the above description. [Explanation of symbols]

[0128] 10. Information processing equipment 11 CPU 12 Storage section 13 I / O 14 Display section 15 Input section 111 Acquisition Department 113 Extraction part 114 Registration Department 115 Reading part 121 Dataset Storage 122 Document information storage unit 131~135 Data Model 21 Timeline 31~39 Requirements 41~43 solid line

Claims

1. an acquisition unit that acquires a predetermined document; an extraction unit that extracts at least one of a natural language from the predetermined document, a name of an item in the predetermined document from which the natural language has been extracted, a business process to which the predetermined document belongs, and a type of the predetermined document; a reading unit that searches a document information table to determine whether information identical and / or similar to at least one of the received natural language, the name of the item, the business process, and the type of the predetermined document exists in the document information table, and based on the result of the determination, reads out information about the process category to which the item belongs and items that have a mutual dependency correlation and / or attribute correlation with the item, which are described in the document information table; a registration unit that registers data on the natural language, the name of the item, the business process, the type of the predetermined document, the process category to which the item belongs, and items that are in a dependency correlation and / or attribute correlation with the item in a dataset; a document information storage unit that stores the document information table; An information processing device comprising:

2. 2. The information processing apparatus according to claim 1, wherein the attribute correlation is a correlation between a given natural language and a natural language expressing an attribute that describes the given natural language.

3. 2. The information processing device according to claim 1, wherein the dependency correlation is a correlation having at least one of a mechanism, a parent-child relationship, or a cause-and-effect dependency correlation between a predetermined natural language and natural languages ​​that appear around the predetermined natural language.

4. 2. The information processing apparatus according to claim 1, wherein the process classification is any one of R, F, L, P, l, and p processes.

5. The information processing device described in claim 1, characterized in that the reading unit determines whether information similar to the received natural language and at least one of the name of the item, the business process, and the type of the specified document exists in the document information table based on one or more judgment criteria.

6. The information processing device according to claim 5, characterized in that the one or more judgment criteria are composed of at least one of a judgment criterion for determining whether business processes are similar, a judgment criterion for determining whether types of specified documents are similar, a judgment criterion for determining whether item names are similar, a judgment criterion for determining whether process classifications are similar, a judgment criterion for determining whether dependency correlations are similar, a judgment criterion for determining whether attribute correlations are similar, and a judgment criterion for determining whether natural language character strings are similar.

7. The information processing apparatus according to claim 5 , wherein the reading unit quantifies one or more similarities based on the one or more criteria.

8. 8. The information processing apparatus according to claim 7, wherein the reading unit determines that the information is similar by comparing the one or more similarities with one or more thresholds.

9. The information processing apparatus according to claim 7 , wherein the registration unit registers the one or more similarities in the data set.

10. The information processing device according to claim 1, characterized in that the registration unit registers information about the natural language, the name of the item, the business process, the type of the specified document, the process category to which the item belongs, and items that have a mutual dependency correlation and / or attribute correlation with the item in the document information table.

11. 2. The information processing apparatus according to claim 1, further comprising a data processing unit that performs statistics and organization on the data stored in the data set.

12. 2. The information processing device according to claim 1, further comprising a name matching unit that searches the dataset based on the natural language, the business process, and information about items that have mutual dependencies and / or attribute relationships with each of the items, and determines whether a natural language that is highly similar to the natural language is registered in the dataset.

13. 13. The information processing device according to claim 12, further comprising a document information generation unit that replaces all natural languages ​​included in the document with natural languages ​​determined by the name matching unit to have high similarity, and then generates the replaced document.

14. 13. The information processing device according to claim 12, further comprising a document tagging unit that extracts character strings determined to have a high similarity by the name matching unit from the dataset, tags all natural languages ​​included in the document with the natural languages ​​determined to have a high similarity by the name matching unit, and then generates a document in which the natural languages ​​have been replaced with the tagged natural languages.

15. In an information processing apparatus having a document information storage unit that stores a document information table, an acquisition step of acquiring a predetermined document; an extraction step of extracting at least one of a natural language from the predetermined document, a name of an item in the predetermined document from which the natural language has been extracted, a business process to which the predetermined document belongs, and a type of the predetermined document; a reading step of searching the document information table and reading information about the process category to which the item belongs and items that have a mutual dependency correlation and / or attribute correlation with the item, which are listed in the document information table and are linked to information that is the same as or similar to at least one of the received natural language, the name of the item, the business process, and the type of the predetermined document; a registration step of registering data on the natural language, the name of the item, the business process, the type of the predetermined document, the process category to which the item belongs, and items that are in a dependency correlation and / or attribute correlation with the item in a dataset; An information processing method comprising:

16. On the computer, an acquisition function for acquiring a predetermined document; an extraction function for extracting at least one of a natural language from the predetermined document, a name of an item in the predetermined document from which the natural language has been extracted, a business process to which the predetermined document belongs, and a type of the predetermined document; a reading function that searches a document information table and reads information about a process category to which the item belongs and items that are in a mutual dependency correlation and / or attribute correlation with the item, which are listed in the document information table and are linked to information that is the same as or similar to at least one of the received natural language, the name of the item, the business process, and the type of the predetermined document; a registration function for registering data on the natural language, the name of the item, the business process, the type of the specified document, the process category to which the item belongs, and items that are in a mutual dependency correlation and / or attribute correlation with the item in a dataset; a document information storage function for storing the document information table; An information processing program characterized by realizing the above.

Images