Report creation program, report creation method, and report creation device

Abstract

To create a report of an appropriate report content.SOLUTION: A report creation program causes a computer which creates a report by using information including an extracted report viewpoint, site information and a report conclusion, based on the created report, to execute processing of: learning the report conclusion, by inputting the information to a learning unit; creating a learned format for each learned report conclusion; and creating the report by using the learned format corresponding to the report conclusion inferred from specific site information, among the learned formats for each created report conclusion.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present disclosure relates to a report creation program, a report creation method, and a report creation device. [Background technology]

[0002] A support system has been proposed that supports on-site work, and includes a design database in which design information is managed, and a work information database that stores work items individually extracted from the design information, judgment criteria for the work items, and information on the order in which the work items are performed (see, for example, Patent Document 1). The support system further includes a mobile terminal that downloads and displays on a screen a work procedure manual in which the work items are arranged based on the order information, and inputs measured values ​​for each work item. The support system further includes a site information database that uploads the measured values ​​input on the mobile terminal in association with the date of implementation. The support system further includes a report generation unit that generates a report in a predetermined format by associating at least the work items, judgment criteria, measured values, and date of implementation. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2015-018499 A Summary of the Invention [Problem to be solved by the invention]

[0004] However, with the above-mentioned support system for generating a report in a predetermined format, it is difficult to generate reports corresponding to multiple meetings, and creating the report can be time-consuming.

[0005] Thus, in one aspect, the present disclosure provides for creating a report with appropriate report content. [Means for solving the problem]

[0006] In one embodiment, a report creation program is provided for executing a process in which a computer creates a report using information including extracted reporting perspectives, on-site information, and a reporting conclusion based on a created report, inputting the information into a learning unit to learn a reporting conclusion, create a learned format for each of the learned reporting conclusions, and create the report using one of the learned learned formats for each of the reporting conclusions that corresponds to the reporting conclusion inferred from specific on-site information. Effect of the Invention

[0007] In one aspect, the present disclosure can generate reports with appropriate reporting content. [Brief description of the drawings]

[0008] [Figure 1] FIG. 1 illustrates an example of an overall configuration of a report creation device. [Diagram 2] FIG. 2 illustrates an example of a hardware configuration of a report creation device. [Diagram 3] FIG. 2 illustrates an example of a functional configuration of a preprocessing unit and a learning unit. [Figure 4] 13 is a flowchart showing the flow of pre-processing. [Diagram 5] FIG. 13 is a diagram illustrating an example of a report. [Figure 6] FIG. 13 is a diagram illustrating an example of text mining of a report. [Figure 7] 11A and 11B are diagrams illustrating an example of a report viewpoint and reference site information. [Figure 8] FIG. 11 is a diagram illustrating an example of a report viewpoint DB. [Figure 9] 13 is a flowchart showing the flow of a learning process. [Figure 10] FIG. 13 is a diagram illustrating an example of a learned format DB. [Figure 11] FIG. 13 is a diagram showing an example of a learned format (fixed portion, variable portion). [Figure 12] FIG. 2 illustrates an example of a functional configuration of an inference unit. [Figure 13] 13 is a flowchart showing the flow of a report creation process. [Figure 14] FIG. 11 is a diagram for explaining a report creation process. [Figure 15] FIG. 1 illustrates an example of a system configuration of a report creation system. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0009] Hereinafter, an embodiment will be described with reference to the accompanying drawings. In this specification and the drawings, components having substantially the same functional configurations are denoted by the same reference numerals, and redundant description may be omitted.

[0010] [Overall configuration of the report creation device] The overall configuration of a report creation device 120 according to an embodiment will be described with reference to Fig. 1. Fig. 1 is a diagram showing an example of the overall configuration of a report creation device 120 according to an embodiment.

[0011] The report creation device 120 includes a preprocessing unit 111, a learning unit 112, an inference unit 121, a report creation unit 122, an output unit 123, a learning data storage unit 114, and a learning result data storage unit 124. The learning data storage unit 114 includes a report DB (database) 115 and a site information DB 116. The learning result data storage unit 124 includes a report perspective DB 125 and a learned format DB 126. The report creation device 120 further includes the latest site information 127. The latest site information 127 indicates the current site information or the latest site information among the site information that can be acquired from the site. The report DB 115 is a database that accumulates reports that have been created in the past. The report perspective is an item that appears frequently (the frequency of appearance is equal to or greater than a preset threshold) that is extracted from the items of past reports, and is extracted by analyzing the reports accumulated in the report DB 115 using text mining, which will be described later. The site information DB116 is a store of information showing the past and current (or latest) status of the site. The site information includes work plans and work status (daily work schedules, and actual daily work results). The reported conclusion includes information showing progress, such as the degree of progress. The degree of progress is obtained from the daily work results for the daily work schedules among the site information stored in the site information DB116. However, the reported conclusion is not limited to this, and may be the monthly work results for the monthly work schedules, or the time series changes in the daily work results for the daily work schedules (for example, yesterday's delays have been made up today). The reported conclusion is included in the site information.

[0012] The report creation device 120 receives a learning instruction from a terminal (not shown). The terminal that outputs a signal of the learning instruction or the report creation instruction may be a tablet terminal, a smartphone, or a wearable device such as an HMD (Head Mount Display) or an FMD (Face Mount Display).

[0013] In response to the learning instruction, the preprocessing unit 111 extracts "reporting perspectives" from the description items of past reports according to their frequency of appearance, and scores the relevance between each of the report perspectives. The "reporting perspectives" are report items extracted from the description items of the report that have a frequency of appearance equal to or greater than a threshold. In the example of the report in FIG. 5, if description items such as "A project progress materials," "reporting period," "year / month / day to year / month / day," "reporting date," "year / month / day," "creator," "reporter," "I. Overall summary status," "[Conversion test progress management table (compare completed)]," "II. Work status report," "III. Future plans (until the next progress meeting)," and "IV. Key issues and contact items" have a frequency equal to or greater than a preset threshold, they are extracted as report perspectives.

[0014] Furthermore, the pre-processing unit 111 determines a "report perspective reference destination" corresponding to the report perspective from past on-site information. On-site information corresponding to the report perspective is stored in the report perspective reference destination. Past reports are classified by meeting body and stored in the report DB 115 (see FIG. 5). The reports stored in the report DB 115 may be any type of electronic data (electronic file), such as text data, image data obtained by scanning paper media with a scanner, or data digitized using RPA or the like. The learned report perspective, description content, relevance between report perspectives, and report perspective reference destination for each meeting body are stored in the report perspective DB 125 (see FIG. 8).

[0015] A meeting body is a group of meetings set up for a certain purpose, and refers to a group (organization) that holds the meetings. A meeting body may be a group (organization) to which the same or similar reports can be submitted, such as a project, a department, a division, or a headquarters. Reports classified by meeting body are, for example, project reports, department reports, division reports, and headquarters reports.

[0016] The learning unit 112 learns the "description contents" and "description positions" of the report perspectives extracted for each meeting body and the report perspectives having a certain degree of relevance to the report perspectives. The learning unit 112 inputs information that associates the extracted report perspectives (including report perspectives having a certain degree of relevance or more) with the reference destination of the site information to which the report perspective refers. An example of information that associates the extracted report perspective with the reference destination of the site information to which the report perspective refers is information on the "report perspective" and the "report perspective reference destination." In other words, the information specified by the report perspective reference destination includes the site information (including the report conclusion). The learning unit 112 learns the "report conclusion" based on the past site information stored in the "report perspective reference destination" corresponding to the "report perspective," thereby generating a learned format that has learned the "report conclusion" for each meeting body. The past site information is accumulated in the site information DB 116 (see FIG. 7). The site information includes information indicating the situation of the site, such as a plan, progress, test specifications, trouble, and Q&A (questions and answers). The site information is information indicating the situation at the site, and includes information on the results (the work content performed) of the scheduled work content in any unit such as daily or monthly. In other words, the site information includes information on the progress status indicated by the results of the scheduled work content. The learned format is accumulated in the learned format DB 126.

[0017] The report creation device 120 receives a report output instruction signal from a terminal (not shown). The report output instruction signal includes a conference body to which the report is to be sent, which is specified by the terminal. The inference unit 121 infers the progress level as a report conclusion from the current site information (latest site information 127). For example, the inference unit 121 pre-learns the contents of reports by progress level that were submitted to past conference bodies. Based on the pre-learned reports by progress level, the inference unit 121 derives a report that is closest to the current progress status indicated by the latest site information 127, and infers a "report conclusion" from that report.

[0018] The report creation unit 122 determines a learned format that indicates a "report conclusion" that is the same as or closest to the inferred "report conclusion," and creates a report by applying information corresponding to the latest site information 127, such as the progress level, to the determined learned format. The report is a report created by the report creation unit 122. The output unit 123 transmits the created report to the terminal that issued the report output instruction signal, or displays it on the screen of the terminal.

[0019] [Hardware configuration] Next, a description will be given of a hardware configuration of the report creation device 120. FIG.

[0020] The report creation device 120 includes a central processing unit (CPU) 501, a read only memory (ROM) 502, and a random access memory (RAM) 503. The CPU 501, the ROM 502, and the RAM 503 form a so-called computer.

[0021] The report creation device 120 also has an auxiliary storage device 504, a display device 505, an operation device 506, a communication device 507, and a drive device 508. In the report creation device 120, each piece of hardware is connected to each other via a bus 509.

[0022] The CPU 501 is a computing device that executes various programs installed in the auxiliary storage device 504 (for example, a pre-processing program, a learning program, a report creation processing program, etc.).

[0023] The ROM 502 is a non-volatile memory. The ROM 502 functions as a main storage device that stores various programs, data, etc., required for the CPU 501 to execute various programs installed in the auxiliary storage device 504. Specifically, the ROM 502 functions as a main storage device that stores boot programs such as a BIOS (Basic Input / Output System) and an EFI (Extensible Firmware Interface).

[0024] The RAM 503 is a volatile memory such as a dynamic random access memory (DRAM) or a static random access memory (SRAM). The RAM 503 functions as a main storage device that provides a working area in which various programs installed in the auxiliary storage device 504 are expanded when the CPU 501 executes them.

[0025] The auxiliary storage device 504 is an auxiliary storage device that stores various programs and data. For example, the learning data storage unit 114 and the learning result data storage unit 124 are realized in the auxiliary storage device 504. The display device 505 is a display device that displays the internal state of the report creation device 120. The operation device 506 is an input device that allows a user of the report creation device 120 to input various instructions to the report creation device 120. The communication device 507 is a communication device that allows the report creation device 120 to transmit and receive information to and from other devices.

[0026] The drive device 508 is a device for setting a computer-readable recording medium. The recording medium referred to here includes media that record information optically, electrically, or magnetically, such as a CD-ROM, a flexible disk, or a magneto-optical disk. Alternatively, the recording medium may include a semiconductor memory that records information electrically, such as a ROM or a flash memory.

[0027] In addition, the various programs to be installed in the auxiliary storage device 504 may be installed, for example, by setting a distributed recording medium in the drive device 508 and reading out the various programs recorded on the recording medium by the drive device 508.

[0028] For example, the preprocessing unit 111, the learning unit 112, the inference unit 121, and the report creation unit 122 are realized by the CPU 301 executing the recipe and other programs loaded onto the RAM 303. Also, for example, the learning data storage unit 114 and the learning result data storage unit 124 are realized in the auxiliary storage device 504. Also, for example, the output unit 123 is realized in the display device 505.

[0029] [Functional configuration of the preprocessing unit and learning unit] The functional configurations of the preprocessing unit 111 and the learning unit 112 will be described with reference to Fig. 3. Fig. 3 is a diagram showing an example of the functional configurations of the preprocessing unit 111 and the learning unit 112. The preprocessing unit 111 has a report perspective determination unit 201, a relevance scoring unit 202, and a reference destination determination unit 203.

[0030] The report viewpoint determination unit 201 analyzes the description items in each report stored in the report DB 115 by text mining, extracts the description items for each report and keywords appearing in the description items, and determines the report viewpoint according to the frequency of appearance of the keywords. For example, the report viewpoint determination unit 201 determines a keyword whose frequency of appearance is higher than a threshold as a "report viewpoint" and stores it in the report viewpoint DB 125.

[0031] The relevance scoring unit 202 scores the relevance between the determined reporting viewpoints. For example, the relevance scoring unit 202 stores in the reporting viewpoint DB 125 the frequency of simultaneous appearance between the determined reporting viewpoints as a reporting viewpoint relevance degree.

[0032] The reference destination determination unit 203 analyzes each piece of site information stored in the site information DB 116 by text mining, and extracts characteristic keywords. The reference destination determination unit 203 records the storage destination of site information having a keyword that matches the report perspective determined by the report perspective determination unit 201 among the characteristic keywords as a "report perspective reference destination" in the report perspective DB 125 (see FIG. 8). This links the report perspective with the keyword of the site information of the report perspective reference destination. A storage destination in which site information such as a URL is stored may be specified as the report perspective reference destination (reference of site information). When the site information DB 116 is implemented as a relational database (RDB), a report perspective node may store SQL (Structured Query Language) for extracting report items from the site information, and the site information may be referenced. The report perspective node may have a script for referencing the site information.

[0033] 3, the learning unit 112 includes a report viewpoint learning unit 204 and a format creating unit 205. The report viewpoint learning unit 204 learns the report conclusion of the report for each meeting body from each case of the report.

[0034] The reporting perspective learning unit 204 compares the keywords appearing in the report with the reporting perspectives (stored in the reporting perspective DB 125) determined by the reporting perspective determination unit 201, and learns the description positions of the report for each reporting perspective.

[0035] The report perspective learning unit 204 learns the contents (fixed parts, variable parts) written in the description position for each report perspective in the report. For example, in a report, the headings are mainly the contents of the fixed parts, while the sections stating the results of consideration of progress, etc., tables, assignments, and contact information are the contents including the variable parts. Note that the report that is initially provided to the report perspective learning unit 204 as learning data is preferably a standardized report.

[0036] In this way, the report viewpoints of the report for each meeting body, the respective description positions, description contents, and conclusion reports are learned. The format creation unit 205 creates a learned format for each report conclusion and for each meeting body, in which each description content is described in each description position.

[0037] The format creation unit 205 refers to the relevance obtained by scoring the relevance between each report perspective, and adds to the learned format any report perspective that has a relevance level equal to or higher than a certain level and that has not yet been reflected in the learned format.

[0038] When a new report is created, the learning unit 112 can learn the change tendency from the previously learned report and update the learned format for each report conclusion for each meeting body.

[0039] [Preprocessing] Next, a flow of pre-processing according to an embodiment will be described with reference to Figs. 4 to 8. Fig. 4 is a flowchart showing the flow of pre-processing according to an embodiment. Fig. 5 is a diagram showing an example of a report. Fig. 6 is a diagram showing an example of text mining of a report. Fig. 7 is a diagram showing an example of a report perspective and on-site information of a reference destination. Fig. 8 is a diagram showing an example of a report perspective DB 125.

[0040] When the preprocessing of FIG. 4 starts, in step S41, the reporting perspective determination unit 201 analyzes the description items in the reports stored in the report DB 115 for each meeting body by text mining, and extracts keywords that appear in the description items for each report. For example, the report shown in FIG. 5 shows a progress document that is an example of a report for a meeting body of project A. The reporting perspective determination unit 201 uses text mining to extract useful information from the report. For example, the reporting perspective determination unit 201 divides the sentences in the report into keywords (words such as nouns, verbs, adjectives, etc.), and analyzes the frequency of appearance of those words and the relevance of a word to other words (such as the frequency of simultaneous appearance in a document).

[0041] Figure 6 shows an example of the results of using text mining to extract what words appear in sentences from department reports along with their frequency of occurrence, and displaying this as a department report perspective. In the department report perspective, words that appear every time in a department report are given an occurrence frequency of "1.0." Occurrence frequencies range from 0 to 1, with higher values ​​indicating higher occurrence frequency. For example, each word after the word "progress" (improvement trend, as planned, proceeding, delay, issue, problem) is a word that appears simultaneously with the word "progress," and appears with a frequency of 0.1 to 0.5.

[0042] As described above, the report perspective determination unit 201 uses text mining to learn a large number of reports, and learns what words are present in the reports of each meeting body, and whether each word appears in each report every time or with a predetermined frequency. The report perspective determination unit 201 also learns whether the presence or absence of appearance of each word following the word "progress" and the degree of appearance change depending on the progress status.

[0043] In step S42 of FIG. 4, the reporting perspective determination unit 201 determines whether the occurrence frequency of a word is higher than a desired threshold. The reporting perspective determination unit 201 determines a word (keyword) whose occurrence frequency is higher than the desired threshold as a "reporting perspective" and stores it in the reporting perspective DB 125. In step S42, if the reporting perspective determination unit 201 determines that the occurrence frequency is equal to or lower than the desired threshold, it proceeds to step S45. In step S42, if the reporting perspective determination unit 201 determines that the occurrence frequency is higher than the desired threshold, it proceeds to step S43, where it determines the high-frequency keyword (whose occurrence frequency is higher than the desired threshold) as the reporting perspective.

[0044] Next, in step S44, the relevance scoring section 202 scores the relevance between the determined report viewpoints.

[0045] Next, in step S45, the reference destination determination unit 203 uses text mining to analyze the site information stored in the site information DB 116. Text mining of site information is similar to text mining of reports, and therefore a description thereof will be omitted.

[0046] Next, in step S46, the reference destination determination unit 203 extracts characteristic keywords from the site information.

[0047] Next, in step S47, the reference destination determination unit 203 determines whether there is a keyword that matches the report perspective determined by the report perspective determination unit 201 among the characteristic keywords. If there is a keyword that matches the report perspective, in step S48, the reference destination determination unit 203 determines the site information having the keyword that matches the report perspective as the "report perspective reference destination". Next, in step S49, the reference destination determination unit 203 associates the determined "report perspective reference destination" with the matching report perspective and records it in the report perspective DB 125, and ends the pre-processing. As a result, the "report perspective reference destination" corresponding to the report perspective determined from the past site information accumulated in the site information DB 116 is stored in the report perspective DB 125. In step S47, if there is no site information having a keyword that matches the determined report perspective, the pre-processing ends as it is.

[0048] FIG. 7 is a diagram showing an example of a report perspective and site information of a reference. The reference determination unit 203 extracts a reference of the site information to be embedded in the report perspective of the learned format by using text mining. For example, in the example of FIG. 7, the site information accumulated in the site information DB 116 has keywords of item names such as project name, asset classification, planned actual (work plan / work performance), daily work schedule value (e.g., 10), and work performance value (e.g., 8). The reference determination unit 203 extracts, by text mining, site information having keywords matching each of the report perspectives of "project" and "asset classification" in FIG. 7 from the department's report perspectives extracted by text mining in FIG. 6. The reference determination unit 203 records a "report perspective reference" indicating a storage destination of the site information having keywords matching each of the report perspectives of "project" and "asset classification". When creating a report, the system determines an appropriate trained format and embeds in the report perspective numerical data indicating the latest progress status of on-site information in the "report perspective reference destination" linked to the "report perspective" of the trained format. This allows the system to create the latest report.

[0049] In pre-processing, the report itself for each meeting body is used as learning data and machine-learned before the learning process described below. Then, in steps S41 to S44 in Fig. 4, the report format for each meeting body is learned. Also, in steps S41 to S44, the report content for each meeting body is learned. This completes the preparation of data for learning by the learning unit 112 described below.

[0050] The data learned in the pre-processing is stored in the report perspective DB 125. Fig. 8 shows an example of the report perspective DB 125. The report perspective DB 125 stores the report perspective of the report for each meeting body. In the example of Fig. 8, each piece of information on the report perspective name, report perspective content (description content), relevance between the report perspectives, and report perspective reference destination is stored for each meeting body (group name).

[0051] [Learning process] Next, a flow of a learning process according to an embodiment will be described with reference to Fig. 9 to Fig. 11. Fig. 9 is a flowchart showing the flow of the learning process according to an embodiment. Fig. 10 is a diagram showing an example of a learned format DB. Fig. 11 is a diagram showing an example of a learned format (fixed portion, variable portion).

[0052] When the learning process in FIG. 9 is started, in step S91, the report viewpoint learning unit 204 acquires past reports.

[0053] Next, in step S92, the reporting perspective learning unit 204 compares the keywords appearing in the report with the acquired reporting perspectives for each description item of the report, and learns the description positions for each reporting perspective.

[0054] Next, in step S93, the report viewpoint learning unit 204 identifies the description content, fixed part, and variable part of the report viewpoint. By distinguishing between fixed parts and variable parts at the beginning when creating a report format, in the next and subsequent learning, it is possible to provide standardized information to the fixed parts, making further learning of the fixed parts unnecessary, and improving the learning efficiency and accuracy.

[0055] Next, in step S94, the report perspective learning unit 204 extracts and learns a "report conclusion" based on a "report perspective reference" corresponding to the "report perspective" from the past site information stored in the site information DB 116.

[0056] Next, in step S95, the format creation unit 205 creates a learned format for each "meeting body" and for each "report conclusion" based on the description position of the "report perspective".

[0057] Next, in step S96, the report viewpoint learning unit 204 adds to the learned format a report viewpoint having a certain or higher degree of relevance to the report viewpoint described in the learned format while referring to the relevance between the report viewpoints in the report viewpoint DB. After executing step S96, the learning process ends.

[0058] The created trained formats are trained formats classified by report conclusion for each meeting body, and are stored in the trained format DB 126. In the example of the trained format DB in Fig. 10, the format name, report perspective, report conclusion, description item, description content, and description position are stored for each meeting body. In the example of Fig. 10, the meeting body is Project A, and trained formats (Project A Reports 1 to 9) are created for each report conclusion.

[0059] FIG. 11 is a diagram showing an example of fixed and variable parts of a learned format. In the example of FIG. 11, the shaded parts are variable parts, and the non-shaded parts are fixed parts. The contents of the report perspective have either fixed or variable parts. In the case of a fixed part, a fixed sentence is displayed at the specified writing position. In the case of a variable part, the latest site information 127 is displayed at the specified writing position based on the "report perspective reference destination."

[0060] 10 and 11, the reporting unit is the project unit, and a format for each report conclusion (degree of progress) for each project is created by learning the progress report for the project. However, this is not limited to this, and a format for each report conclusion for each department or business division may be created by learning the progress report for each division or business division.

[0061] [Functional configuration of inference section, report creation section, and output section] The functional configuration of the inference unit 121 will be described with reference to Fig. 12. Fig. 12 is a diagram showing an example of the functional configuration of the inference unit 121. The inference unit 121 has a report conclusion inference unit 301 and a format determination unit 302.

[0062] The report conclusion inference unit 301 infers a "report conclusion" corresponding to the latest site information 127 based on the report perspective DB 125 and on the site information (particularly the "progress rate") of the "report perspective reference" linked to the "report perspective" of the report of the "meeting body" specified in the report output instruction.

[0063] The format determination unit 302 determines a learned format having a reporting conclusion corresponding to the reporting conclusion of the report inferred by the inference unit 121. The format determination unit 302 may select a plurality of learned formats having a reporting conclusion that matches and / or is closest to the reporting conclusion of the report inferred by the inference unit 121. In this case, the selected plurality of learned formats are displayed selectably on the screen of the terminal that transmitted the report output instruction. When the user of the terminal selects one of the plurality of learned formats displayed on the screen, the format determination unit 302 selects one learned format according to the user operation.

[0064] When the learned format is determined, the report creation unit 122 determines what reporting perspectives exist for each meeting body on the determined learned format, and then arranges the latest on-site information 127 for each reporting perspective to complete the report.

[0065] The output unit 123 displays the completed report on the screen of the terminal that sent the report output instruction. The output unit 123 may also send the completed report to the terminal that sent the report output instruction. This makes it possible to create a report with appropriate report content that corresponds to the inferred report conclusion in the determined format.

[0066] [Report creation process] Next, a flow of a report creation process according to an embodiment will be described with reference to Fig. 13 and Fig. 14. Fig. 13 is a flowchart showing the flow of the report creation process. Fig. 14 is a diagram for explaining the report creation process.

[0067] When the report creation device 120 receives a report output instruction from a terminal (not shown), the report creation process of FIG. 13 is started, and in step S101, the report conclusion inference unit 301 identifies the meeting entity specified in the report output instruction signal.

[0068] Next, in step S102, the report conclusion inference unit 301 acquires the “report perspective” of the report of the specified “meeting body” based on the report perspective DB 125.

[0069] Next, in step S103, the report conclusion inference unit 301 acquires on-site information of the "report perspective reference destination" linked to the acquired report perspective. For example, Fig. 14(a) shows an example of on-site information of the "report perspective reference destination" where the specified "meeting body" is linked to the report perspective of the A project report of the department report.

[0070] 13, the report conclusion inference unit 301 infers a "report conclusion" corresponding to the latest site information 127 based on the site information of the acquired "report viewpoint reference destination." For example, the report conclusion inference unit 301 acquires information on the progress status from the latest site information 127, calculates a progress rate (degree of progress), and selects a learned format that indicates a report conclusion that is closest to the calculated progress rate.

[0071] For example, the progress rate is calculated by comparing the schedule and actual results for June 25th, which is shown in the bold frame of the latest site information in FIG. 14(a). The progress rate is calculated as 180 / 200=90% from the reporting perspective of online (A). Here, June 25th is the target date for inferring the report conclusion, and is the current date or the latest date for inferring the report conclusion. FIG. 14(b) shows an example of a trained format. In the example of FIG. 14(b), among the trained formats of the A project report in which the "meeting body" is the report perspective "progress data" and the progress rate is 90%, the report conclusion shown in the bold frame matches and / or is closest to the report conclusion that matches the calculated progress rate. In other words, "A project report 5", "A project report 6", and "A project report 7", which are trained formats having the report conclusions "slightly behind", "improving", and "improving compared to the previous time", are selected.

[0072] In this case, the selected multiple trained formats and the estimated report conclusion may be displayed on the screen of the terminal that sent the report output instruction. When the user selects one trained format from the displayed multiple trained formats, the format determination unit 302 selects one trained format according to the user operation.

[0073] The format determination unit 302 may select one learned format of a reporting conclusion that matches the reporting conclusion of the report inferred by the inference unit 121. The format determination unit 302 may select multiple learned formats having a reporting conclusion that is expected to be (closest to) the reporting conclusion of the report inferred by the inference unit 121.

[0074] Next, in step S105 of FIG. 13, the report creation unit 122 arranges the latest site information 127 in accordance with the description positions for each report viewpoint present in the selected learned format, and completes the report.

[0075] In addition, fixed text is embedded in the fixed part of the report perspective existing in the determined trained format. Therefore, the report creation unit 122 can omit placing the site information in the fixed part of the report perspective by placing information such as the progress of the latest site information 127 in the variable part of the report perspective existing in the determined trained format and creating a report. This makes it possible to create reports efficiently.

[0076] For example, in the example of FIG. 14(c), "Project A" is placed in the variable part of report perspective 401 (the shaded part of the learned format in FIG. 11). The "report period" indicated by report perspective 402 is a fixed part. "June 12, 2021" and "June 25, 2021" are placed in the variable part of report perspective 403 (the shaded part of the learned format in FIG. 11).

[0077] The heading of the report perspective 404 is a fixed part. The report perspective 405 is a variable part, and a sentence such as "As of June 25th, the progress rate is 90.0%, which is slightly behind schedule (about 2 days)" is arranged, and the latest progress status is shown.

[0078] The heading of report perspective 406 and the title of 407 are fixed parts. In the variable parts in the table of report perspective 408 (the shaded parts of the learned format in FIG. 11), various local information such as asset classification, overall, previous, and current is arranged. Note that in FIG. 14(c), the next schedule in the table of FIG. 11 and the report perspectives below the table of FIG. 11 are omitted.

[0079] 13, next, in step S106, the output unit 123 displays the report on the screen of the terminal that sent the report output instruction. The output unit 123 may send the report to the terminal that sent the report output instruction. After that, the report creation process ends.

[0080] The output unit 123 may store the created report in the report DB 115 as an example of a past report (see FIG. 1). The created report may be corrected by a user or the like and then stored in the report DB 115. The created report and the corrected report become input data for the learning process as past report examples. This is expected to improve the accuracy of the report content of the next report to be created.

[0081] By using the report creation program for executing the above-described pre-processing, learning process, and report creation process, the report creation device 120 creates a report with appropriate report content suited to the situation. This reduces the effort required to create similar reports (materials) for each person involved, even when there are many people involved.

[0082] [Variations] The above-described pre-processing, learning processing, and report creation processing may be executed by a report creation system 100 shown in Fig. 15, instead of the report creation device 120 in Fig. 1. Fig. 15 is a diagram showing an example of a system configuration of the report creation system 100.

[0083] 2. Report creation system 100 includes a learning device 110 and a report creation device 120'. Note that the learning device and report creation device 120' in FIG. 15 are assumed to have the same hardware configuration as report creation device 120 in FIG.

[0084] The learning device 110 has a preprocessing unit 111, a learning unit 112, and a learning data storage unit 114. The preprocessing unit 111, the learning unit 112, and the learning data storage unit 114 have the same functions as the preprocessing unit 111, the learning unit 112, and the learning data storage unit 114 shown in FIG. 1, and therefore their explanations will be omitted.

[0085] The report creation device 120′ has an inference unit 121, a report creation unit 122, an output unit 123, and a learning result data storage unit 124. The inference unit 121, the report creation unit 122, the output unit 123, and the learning result data storage unit 124 have the same functions as the inference unit 121, the report creation unit 122, the output unit 123, and the learning result data storage unit 124 shown in FIG. 1, and therefore the description thereof will be omitted.

[0086] The learning results learned by the learning device 110 are stored in a learning result data storage unit 124. A report creation device 120′ creates a report with appropriate report content according to the report destination and situation, based on the various data stored in the learning result data storage unit 124.

[0087] Although the report creation program, report creation method, and report creation device according to the present disclosure have been described above using the above-mentioned embodiments, the report creation program, report creation method, and report creation device according to the present disclosure are not limited to the above-mentioned embodiments, and various modifications and improvements are possible within the scope of the present disclosure. Furthermore, when there are a plurality of the above-mentioned embodiments and modifications, they can be combined within a range that does not contradict each other.

[0088] In addition to the above description, the following items are disclosed. (Appendix 1) a computer that creates a report using the extracted information including the report viewpoint, the field information, and the report conclusion based on the created report; The information is input to a learning unit to learn a report conclusion, and a learned format is created for each of the learned report conclusions; creating the report using the learned format corresponding to the report conclusion inferred from specific on-site information among the learned formats for the report conclusions created; A report-writing program to run the process. (Appendix 2) Based on the reports created for each meeting body, the extracted multiple reporting perspectives and the multiple pieces of on-site information to which the multiple reporting perspectives refer are input to the learning unit, thereby creating a learned format for each meeting body for each learned report conclusion, creating the report using the learned format corresponding to the report conclusion inferred from the specific on-site information, among the learned formats of the conference body designated when instructing to create the report; Attachment 1, a report generation program. (Appendix 3) Extracting the relevance between the plurality of report perspectives according to the frequency of occurrence of the description items in the report; extracting a plurality of the report perspectives according to the relevance; The information associating the extracted multiple reporting perspectives with the multiple pieces of on-site information to which the multiple reporting perspectives refer, respectively, is input to the learning unit, thereby creating the learned format for each meeting body for each learned report conclusion. Attachment 2, a report generation program. (Appendix 4) The specific site information is the latest site information, Select the learned format that matches or is closest to the report conclusion inferred from the latest on-site information from among the learned formats of the specified conference body, generating said report using said selected learned format; Attachment 2, a report generation program. (Appendix 5) If multiple learned formats are selected, display the selected multiple learned formats; creating the report using one of the learned formats selected in response to a user operation from among the displayed plurality of learned formats; Attachment 4, a report generation program. (Appendix 6) The learned format includes information on the description position and description content of each of the plurality of report viewpoints, and the information on the description content includes information on a fixed part or a variable part; For a reporting perspective having the variable portion in the description content among the plurality of reporting perspectives, the latest on-site information associated with the reporting perspective is placed in the variable portion. Attachment 4, a report generation program. (Appendix 7) A computer that creates a report using information including the extracted report viewpoint, the field information, and the report conclusion based on the created report, The information is input to a learning unit to learn a report conclusion, and a learned format is created for each of the learned report conclusions; creating the report using the learned format corresponding to the report conclusion inferred from specific on-site information among the learned formats for the report conclusions created; A reporting method that performs an action. (Appendix 8) A report creation device that creates a report using information including an extracted report viewpoint, on-site information, and a report conclusion based on a created report, a learning unit that learns a report conclusion by inputting the information and creates a learned format for each of the learned report conclusions; a report creation unit that creates the report by using the learned format corresponding to the report conclusion inferred from specific site information among the learned formats for each report conclusion that have been created; A report creation device having the above configuration. (Appendix 9) A learning device that learns to create a report using information including an extracted report viewpoint, on-site information, and a report conclusion based on a created report, a pre-processing unit that extracts the information based on the report, the information including the report perspective, the site information, and the report conclusion; a learning unit that learns a report conclusion by inputting the information and creates a learned format for each of the learned report conclusions; A learning device having the above configuration. (Appendix 10) A report creation device that creates a report using information including an extracted report viewpoint, on-site information, and a report conclusion based on a created report, an inference unit that infers a report conclusion by inputting the information including the report viewpoint, the site information, and the report conclusion; A report creation unit that creates the report using a learned format corresponding to a report conclusion inferred from the specific site information, among learned formats for each report conclusion learned by the learning device described in Supplementary Note 9; A report creation device having the above configuration. [Explanation of symbols]

[0089] 111 Pretreatment section 112 Learning Department 114 Learning data storage unit 115 Report DB 116 Field information DB 120 Report writing device 121 Reasoning part 122 Report Writing Department 123 Output section 124 Learning result data storage unit 125 Reporting Perspective DB 126 Learned Format DB 127 Latest site information

Claims

1. a computer that creates a report using the extracted information including the report viewpoint, the field information, and the report conclusion based on the created report; The information is input to a learning unit to learn a report conclusion, and a learned format is created for each of the learned report conclusions; creating the report using the learned format corresponding to the report conclusion inferred from specific on-site information among the learned formats for the report conclusions created; A report-writing program to run the process.

2. Based on the reports created for each conference body, the extracted multiple reporting perspectives and the multiple on-site information references to which the multiple reporting perspectives refer are input to the learning unit, thereby creating a learned format for each conference body for each learned report conclusion, creating the report using the learned format corresponding to the report conclusion inferred from the specific on-site information, among the learned formats of the conference body designated when instructing to create the report; The report creation program according to claim 1 .

3. Extracting the relevance between the plurality of report perspectives according to the frequency of occurrence of the description items in the report; extracting a plurality of the report perspectives according to the relevance; The information associating the extracted multiple reporting perspectives with the multiple on-site information references to which the multiple reporting perspectives refer is input to the learning unit, thereby creating the learned format for each meeting body for each learned report conclusion. The report creation program according to claim 2.

4. The specific site information is the latest site information, Select, from among the learned formats of the specified conference body, the learned format that matches or is closest to the report conclusion inferred from the latest on-site information; generating said report using said selected learned format; The report creation program according to claim 2.

5. If multiple learned formats are selected, display the selected multiple learned formats; creating the report using one of the learned formats selected in response to a user operation from among the displayed plurality of learned formats; The report creation program according to claim 4.

6. The learned format includes information on the description position and description content of each of the plurality of report viewpoints, and the information on the description content includes information on a fixed part or a variable part; For a reporting perspective having the variable portion in the description content among the plurality of reporting perspectives, the latest on-site information associated with the reporting perspective is placed in the variable portion. The report creation program according to claim 4.

7. A computer that creates a report using information including the extracted report viewpoint, the field information, and the report conclusion based on the created report, The information is input to a learning unit to learn a report conclusion, and a learned format is created for each of the learned report conclusions; creating the report using the learned format corresponding to the report conclusion inferred from specific on-site information among the learned formats for the report conclusions created; A reporting method that performs an action.

8. A report creation device that creates a report using information including an extracted report viewpoint, on-site information, and a report conclusion based on a created report, a learning unit that learns a report conclusion by inputting the information and creates a learned format for each of the learned report conclusions; a report creation unit that creates the report by using the learned format corresponding to the report conclusion inferred from specific site information among the learned formats for each report conclusion that have been created; A report creation device having the above configuration.

Images