CN111814018A - Record management system and device, document approval and production device and method, recording medium - Google Patents

Abstract

Translated fromChinese

一种记录管理系统，包含文档审批装置和文档制作装置。其中，文档审批装置具备：审批请求显示单元，显示包含审批栏的审批请求文档；建议输入引导单元，响应于与审批栏相关的一系列的审批输入的第1输入，显示建议输入画面；以及学习数据输出单元，接受在建议输入画面上输入的建议，并且响应于一系列的审批输入的第2输入，向学习单元输出包含文档内容和接受的建议的学习用数据集，所述文档内容包含显示的审批请求文档的关键词。由此在知识、经验丰富的审批者在决定文档时能够给予建议并作为教师数据，以备用作机械学习。

A record management system includes a document approval device and a document production device. The document approval device includes: an approval request display unit that displays an approval request document including an approval column; a suggestion input guide unit that displays a suggestion input screen in response to a first input of a series of approval inputs related to the approval column; and learning A data output unit that accepts a suggestion input on the suggestion input screen, and in response to a second input of a series of approval inputs, outputs to the learning unit a learning data set including document content and the accepted suggestion, the document content including the display The keyword for the approval request document. This allows knowledgeable, experienced approvers to give advice when deciding on documents and serve as teacher data to back up for machine learning.

Description

Translated fromChinese

记录管理系统和装置、文档审批和制作装置及方法、记录介质Record management system and device, document approval and production device and method, recording medium

技术领域technical field

本发明涉及记录管理系统、记录管理装置、文档审批装置及文档制作装置、以及记录管理方法、文档审批方法及文档制作方法、记录管理程序、文档审批程序及文档制作程序、以及记录了这些程序的记录介质。The present invention relates to a record management system, a record management device, a document approval device and a document production device, and a record management method, a document approval method and a document production method, a record management program, a document approval program and a document production program, and a document recording the procedures. recording medium.

背景技术Background technique

在对老年人、障碍者、儿童等提供护理、医疗、保育等服务的设施中，总是要求更优质且恰当的服务。特别优选能够预先察觉对象者的异常的行动并应对该异常的行动。In facilities that provide services such as nursing care, medical care, and childcare to the elderly, the disabled, and children, there is always a demand for better and more appropriate services. In particular, it is preferable to be able to detect abnormal behavior of the subject in advance and respond to the abnormal behavior.

在专利文献1所记载的护理系统中，通过各种传感器察觉被护理者的异常状况，参照存储在过去实施的数据库中的护理信息向护理者报知适当的作业内容。专利文献2所记载的柔量评价辅助系统预先对在作业指示书、护理计划制作的阶段会被认为的风险进行评价。In the nursing care system described in Patent Document 1, various sensors detect abnormal conditions of the care recipient, and refer to the nursing care information stored in the database performed in the past and notify the caregiver of appropriate work details. The compliance evaluation support system described in Patent Document 2 evaluates in advance a risk considered at the stage of preparing a work instruction sheet and a nursing plan.

在先技术文献prior art literature

专利文献Patent Literature

专利文献1：日本特开2002-149825号公报Patent Document 1: Japanese Patent Laid-Open No. 2002-149825

专利文献2：日本特开2012-175170号公报Patent Document 2: Japanese Patent Laid-Open No. 2012-175170

另一方面，在这样的设施中，一般利用被称为记录管理系统的计算机系统，采取了案例记录、护理日志、看护日志、护理计划等各种记录。这些记录包含庞大的数量的事例，能够成为用于恰当的服务的活教材。特别是对于经验不足的护理者等，只要经验丰富的经验者给出适当的建议，就能够从这些记录学到很多。On the other hand, in such a facility, various records such as a case record, a nursing log, a nursing log, and a nursing plan are generally taken using a computer system called a record management system. These records contain a huge number of examples and can be a living textbook for proper service. Especially for inexperienced caregivers, etc., a lot can be learned from these records as long as the experienced ones give appropriate advice.

发明内容SUMMARY OF THE INVENTION

本发明着眼于成为经验丰富的工作人员(主任、课长等)对由记录管理系统制作的护理日志、看护日志等业务日志、其他的报告文件等必定给予审批的(电子性的盖章)结构，由于在输入审批时必定要过目所制作的文件(日志等的记录)，因此查看在该文件中所记载的记录内容(表示对象者的状态、环境等的重要的或者特定的关键词)，能够对该对象者、环境等可能引起的情况的可能性进行建议，因此将该建议作为机械学习的教师数据使用，并对此进行了系统化。The present invention is intended to be a structure (electronic seal) in which experienced staff (director, section chief, etc.) must approve the nursing log, nursing log and other business logs created by the record management system, and other report documents. , Since it is necessary to review the created file (records such as logs) when inputting approval, check the contents of the records (important or specific keywords indicating the status of the subject person, environment, etc.) recorded in the file, Since it is possible to suggest the possibility of situations that may be caused by the subject, the environment, etc., this suggestion is used as teacher data for machine learning, and this is systematized.

在一个方式中，本发明的记录管理系统具备：审批请求显示单元，显示包含审批栏的审批请求文档；审批单元，接受与显示的审批请求文档中包含的关键词对应地输入的建议，响应于审批输入而输出包含文档内容和接受的建议的学习用数据集，所述文档内容包含审批请求文档的关键词；学习单元，基于从该审批单元输出的学习用数据集，学习与关键词对应的建议的适合的程度；文档制作单元，将文档输出到画面上；以及指导单元，基于该学习单元的学习结果，输出与从文档制作单元输出的文档中包含的关键词相适的建议，该文档制作单元在文档输出画面上显示从该指导单元输出的建议(指导)。In one mode, the record management system of the present invention includes: an approval request display unit that displays an approval request document including an approval column; an approval unit that accepts suggestions input corresponding to keywords contained in the displayed approval request document, and responds to Approving the input and outputting a learning data set containing document content and accepted suggestions, the document content containing the keywords of the approval request document; the learning unit, based on the learning data set output from the approval unit, learns the corresponding keywords The degree of suitability of the suggestion; a document making unit that outputs a document on a screen; and an instruction unit that outputs a suggestion, based on the learning result of the learning unit, which is suitable for the keywords contained in the document output from the document making unit, the document The production unit displays the advice (guidance) output from the guidance unit on the document output screen.

在另一个方式中，本发明的记录管理系统具备：审批请求显示单元，显示包含审批栏的审批请求文档；审批单元，接受与显示的审批请求文档中包含的关键词对应地输入的建议，响应于审批输入而向学习单元输出包含文档内容和接受的建议的学习用数据集，所述学习单元学习与关键词对应的建议的适合的程度，所述文档内容包含审批请求文档的关键词；文档制作单元，将文档输出到画面上；以及指导单元，基于学习单元的学习结果，输出与从文档制作单元输出的文档中包含的关键词相适的建议，该文档制作单元在文档输出画面上显示从该指导单元输出的建议(指导)。In another mode, the record management system of the present invention includes: an approval request display unit that displays an approval request document including an approval column; an approval unit that accepts suggestions input corresponding to keywords contained in the displayed approval request document, and responds to Outputting a learning data set containing document content and accepted suggestions to the learning unit upon approval input, and the learning unit learns the degree of suitability of the suggestions corresponding to the keywords, the document content including the keywords of the approval requesting document; the document; a production unit that outputs a document on a screen; and an instruction unit that outputs, based on the learning result of the learning unit, a suggestion that is suitable for keywords contained in the document output from the document production unit displayed on the document output screen Advice (guidance) output from this guidance unit.

在又一方式中，本发明的记录管理系统(装置)具备：审批请求显示单元，显示包含审批栏的审批请求文档；审批单元，接受与显示的审批请求文档中包含的关键词对应地输入的建议，响应于审批输入而向学习单元输出包含文档内容和接受的建议的学习用数据集，所述学习单元学习与关键词对应的建议的适合的程度，所述文档内容包含审批请求文档的关键词；以及文档制作单元，将在画面上显示的文档输出到指导单元，并且在画面上显示基于学习单元的学习结果而从指导单元输出的与该文档中包含的关键词相适的建议(指导)。In yet another aspect, the record management system (device) of the present invention includes: an approval request display unit that displays an approval request document including an approval column; an approval unit that accepts an input corresponding to a keyword included in the displayed approval request document Recommendation, outputting a learning data set containing document content and accepted recommendations to a learning unit in response to approval input, the learning unit learning the degree of suitability of the recommendations corresponding to the keywords, the document content containing the key to the approval requesting document word; and a document making unit that outputs the document displayed on the screen to the instruction unit, and displays on the screen, on the screen, suggestions (instruction ).

总之，文档的制作者(工作人员)在计算机上制作(输入)日报、月报、报告书等各种记录文档。在该记录文档中，用文字(单词、文章)记录有对象者(被护理者、障碍者、入院患者、儿童等)的行动、晕眩、状态、环境的问题等。然后，在制作了需要上司的审批(决定)的文档的情况下，制作审批栏，在系统上对上司进行审批的请求(审批工作流程运行)(审批请求显示单元)。上司在审批时阅读应该审批的文档，查看其中记载的内容，输入对于对象者、环境(设备、组织、工作人员个人)等可能引起的可能性、工作人员应该注意的事项等的建议，给予审批(决定)(审批单元)。响应于上司的审批的输入，被审批的文档和输入的建议被发送到学习单元(学习系统、学习服务器)。建议被定位为学习的教师数据。将审批的文档中的特定的用语(表示对象者的状态、晕眩、行动、环境的问题等的关键词)和作为教师数据的建议，作为学习用数据集，在机械学习单元中学习建议的适合性的程度。学习结果存储在学习数据库中。在审批者存在多人的情况下，优选在全部的审批或者最高位的审批者给予审批输入时，将文档和建议输送至学习单元。In short, the creator (staff) of the document creates (inputs) various recording documents such as daily reports, monthly reports, and reports on a computer. In this log file, actions, dizziness, state, environmental problems, and the like of the subject (care recipient, disabled person, hospitalized patient, child, etc.) are recorded in characters (words, texts). Then, when a document requiring approval (decision) from the boss is created, an approval column is created, and a request for approval from the boss is made on the system (approval workflow operation) (approval request display means). The boss reads the documents that should be approved when approving, checks the contents recorded in them, and inputs suggestions for possible causes such as the target person, environment (equipment, organization, individual employee), matters that the employee should pay attention to, etc., and approves it. (decision) (approval unit). In response to the supervisor's approved input, the approved documents and the input suggestions are sent to the learning unit (learning system, learning server). Teacher data suggested to be targeted for learning. Specific terms in the approved documents (keywords indicating the state of the subject, dizziness, actions, environmental problems, etc.) and suggestions as teacher data are used as learning data sets, and the suggestions are learned in the machine learning unit. degree of suitability. The learning results are stored in the learning database. In the case where there are multiple approvers, it is preferable to transmit the documents and suggestions to the learning unit when all the approvers or the highest approver gives the approval input.

工作人员日常在计算机画面上制作各种文档，或者显示过去制作的、或者或他人制作的文档并进行阅览(包含文档制作单元、文档阅览)。在正在输入文档时、在阅览时、在输入结束时、在按下指导按钮时等，由工作人员制作的文档被发送到指导单元。指导单元基于工作人员制作的文档中包含的特定的用语(关键词)等，从学习单元(学习数据库)取出适合该关键词的建议或者指导，向工作人员正在输入文档的计算机发送(指导单元)。在计算机上显示所发送的建议或者指导。这样，工作人员能够对于自己正在输入的(输入了的、阅览了的)文档(对于记录在文档中的对象者、环境等)接受适当的建议。给予审批的审批者(工作人员)是经验和知识丰富的熟练者。熟练者输入的建议作为教师数据进行学习。而且，即使在不熟练的工作人员的情况下，也会与该工作人员正在输入的或者输入了的、阅览了的文档中的对象者、环境等相关的关键词相关联地通过学习结果来给予适当的指导或者建议。A worker creates various documents on a computer screen on a daily basis, or displays and reads documents created in the past or created by others (including document creation means and document viewing). The document created by the staff is sent to the guidance unit while the document is being input, at the time of viewing, at the end of the input, when the guidance button is pressed, and the like. The guidance unit extracts advice or guidance suitable for the keyword from the learning unit (learning database) based on specific terms (keywords) included in the document created by the staff, and sends it to the computer where the staff is inputting the document (guidance unit) . Display the sent advice or guidance on the computer. In this way, the worker can receive appropriate advice for the document (for the subject person, the environment, and the like recorded in the document) that he is inputting (inputting, viewing). The approver (staff) who gives the approval is a skilled person with rich experience and knowledge. Suggestions entered by the proficient are learned as teacher data. Furthermore, even in the case of an inexperienced worker, it is given by the learning result in association with keywords related to the subject person, environment, etc. in the document that the worker is inputting or inputting or viewing. Appropriate guidance or advice.

这样，根据本发明的记录管理系统，能够在给予审批的时刻输入用于学习的教师数据，并且给予审批者是熟练者，因此输入适当的教师数据。另一方面，在其他的工作人员(包含审批者)正输入文档等时，由于显示基于学习结果的适当的建议、指导，因此即使是经验不足的工作人员，该作人员也能够在文档输入中进行学习或者掌握知识、考虑注意事项。进而，如果工作人员需要，也能够得到相关的手册等详细指导信息。这样，能够从经验和知识丰富的人向经验不足的工作人员进行知识、经验的传达。In this way, according to the record management system of the present invention, it is possible to input teacher data for learning at the time when approval is given, and since the approver is a skilled person, appropriate teacher data is input. On the other hand, when another worker (including approvers) is inputting a document or the like, since appropriate advice and guidance based on the learning result are displayed, even an inexperienced worker can enter the document. To study or acquire knowledge, consider precautions. Furthermore, if the staff needs, they can also obtain detailed guidance information such as related manuals. In this way, knowledge and experience can be conveyed from a person with rich experience and knowledge to an inexperienced staff member.

本发明的文档审批装置具备：审批请求显示单元，显示包括审批栏的审批请求文档；建议输入引导单元，响应于与审批栏相关的一系列的审批输入的第1输入，显示建议输入画面；以及学习数据输出单元，接受在建议输入画面上输入的建议，并且响应于一系列的审批输入的第2输入，向学习单元输出包含文档内容和接受的建议的学习用数据集，所述文档内容包含显示的审批请求文档的关键词。该文档审批装置能够用于审批者输入作为教师数据的建议。The document approval device of the present invention includes: an approval request display unit that displays an approval request document including an approval column; a suggestion input guide unit that displays a suggestion input screen in response to a first input of a series of approval inputs related to the approval column; and The learning data output unit accepts the suggestion input on the suggestion input screen, and in response to the second input of the series of approval inputs, outputs to the learning unit a learning data set including the document content and the accepted suggestion, the document content including The keyword for the displayed approval request document. The document approval means can be used for the approver to input suggestions as teacher data.

本发明的文档制作装置具备：制作文档显示单元，在画面上显示制作文档；制作文档输出单元，向指导单元输出在画面上显示的文档；以及建议显示单元，在画面上显示基于学习单元的学习结果而从指导单元输出的与制作文档中包含的关键词相适的建议(指导)。若使用该文档制作装置(包含文档阅览装置)，工作人员能够一边输入文档一边阅览，或者在进行了输入时得到适当的建议或者指导。本发明还提供具有与上述相同的特征的记录管理方法、文档审批方法以及文档制作方法以及程序以及记录了该程序的记录介质。The document production apparatus of the present invention includes: a production document display unit that displays the produced document on a screen; a production document output unit that outputs the document displayed on the screen to the instruction unit; and a suggestion display unit that displays the learning based on the learning unit on the screen As a result, suggestions (guidance) suitable for the keywords included in the production document are output from the guidance unit. By using this document creation apparatus (including a document viewing apparatus), a worker can read while inputting a document, or can receive appropriate advice or guidance when inputting. The present invention also provides a record management method, a document approval method, and a document creation method, and a program having the same features as those described above, and a recording medium on which the program is recorded.

附图说明Description of drawings

图1是表示记录管理系统整体的框图。FIG. 1 is a block diagram showing the entire recording management system.

图2是表示记录管理客户端中的记录文档制作处理的流程图。FIG. 2 is a flowchart showing a log file creation process in the log management client.

图3是表示记录管理系统整体中的决定(审批)处理以及学习处理的流程的流程图。3 is a flowchart showing the flow of decision (approval) processing and learning processing in the entire record management system.

图4是表示学习服务器中的学习处理(处理1～3)的整体的流程的流程图。FIG. 4 is a flowchart showing an overall flow of learning processing (processing 1 to 3) in the learning server.

图5是表示学习服务器中的学习处理1的详细内容的流程图。FIG. 5 is a flowchart showing the details of the learning process 1 in the learning server.

图6是表示学习服务器中的学习处理2的详细内容的流程图。FIG. 6 is a flowchart showing the details of the learning process 2 in the learning server.

图7是表示学习服务器中的学习处理3的详细内容的流程图。FIG. 7 is a flowchart showing the details of the learning process 3 in the learning server.

图8表示记录管理客户端中的文档制作画面的例子。FIG. 8 shows an example of a document creation screen in the record management client.

图9表示记录管理客户端中的文档制作画面的例子，表示主任已审批的状态。FIG. 9 shows an example of a document creation screen in the record management client, and shows a state in which the manager has approved it.

图10表示记录管理客户端中的文档制作画面的例子，表示课长审批时显示评论输入画面的状态。FIG. 10 shows an example of a document creation screen in the record management client, and shows a state in which a comment input screen is displayed when the section manager approves.

图11表示记录管理客户端中的文档制作画面的例子，表示课长审批时输入评论的状态。FIG. 11 shows an example of a document creation screen in the record management client, and shows a state where comments are input when the section manager approves.

图12表示记录管理客户端中的文档制作画面的例子，表示课长已审批的状态。FIG. 12 shows an example of a document creation screen in the record management client, and shows a state in which the section manager has approved it.

图13表示从记录管理客户端向学习指导服务器(guidance server)(词素分析服务器)发送的数据的例子。FIG. 13 shows an example of data transmitted from the record management client to a guidance server (morphological analysis server).

图14表示学习服务器(学习数据库)中的关键词主表(keyword master)的例子。FIG. 14 shows an example of a keyword master in the learning server (learning database).

图15表示学习服务器(学习数据库)中的类别分类主表的例子。FIG. 15 shows an example of the category classification master table in the learning server (learning database).

图16表示学习服务器(学习数据库)中的学习数据组(表格)的例子。FIG. 16 shows an example of a learning data set (table) in a learning server (learning database).

图17是表示记录管理系统整体中的文档制作(文档阅览)处理以及其附带的指导显示的流程的流程图。FIG. 17 is a flowchart showing the flow of document creation (document viewing) processing in the entire record management system and the flow of the accompanying guidance display.

图18表示指导服务器中的指导制作处理的前半部分(处理4)的流程图。FIG. 18 is a flowchart showing the first half (processing 4 ) of the guidance creation processing in the guidance server.

图19表示指导服务器中的指导制作处理的后半部分(处理5)的流程图。FIG. 19 is a flowchart showing the second half (process 5 ) of the guidance creation process in the guidance server.

图20表示记录管理客户端中的文档制作(阅览)画面例。FIG. 20 shows an example of a document creation (viewing) screen in the record management client.

图21表示记录管理客户端中的文档制作(阅览)画面的例子，表示在指导区域显示指导的情况。FIG. 21 shows an example of a document creation (viewing) screen in the recording management client, and shows a case where guidance is displayed in the guidance area.

图22表示记录管理客户端中的文档制作(阅览)画面的例子，表示在窗口中显示手册的情况。FIG. 22 shows an example of a document creation (viewing) screen in the record management client, and shows a case where a manual is displayed in the window.

图23表示从记录管理客户端向词素分析服务器发送的数据的例子。FIG. 23 shows an example of data sent from the record management client to the morpheme analysis server.

图24表示从指导服务器向记录管理客户端发送的数据(信息)的例子。FIG. 24 shows an example of data (information) transmitted from the guidance server to the record management client.

附图标记的说明Explanation of reference numerals

10：记录管理系统，10: Records Management System,

11：记录管理客户端(系统)，11: Record management client (system),

12：记录管理服务器，12: Records Management Server,

13：记录管理数据库，13: Records management database,

14：词素分析服务器，14: Morpheme Analysis Server,

15：学习服务器，15: Learning Server,

16：学习数据库，16: Learning Database,

17：指导服务器，17: Guidance Server,

18：手册数据库，18: Manual Database,

19：学习指导服务器(系统)。19: Learning guidance server (system).

具体实施方式Detailed ways

图1表示记录管理系统整体的结构。FIG. 1 shows the overall configuration of the recording management system.

一般情况下，在记录管理系统10中设置有多台(也可以是一台)记录管理客户端(计算机)(系统)11。这些记录管理客户端11能够与记录管理服务器12相互通信。记录管理服务器12具备记录管理数据库13。Generally, a plurality of (or one) record management clients (computers) (systems) 11 are provided in therecord management system 10 . Theserecord management clients 11 can communicate with each other with therecord management server 12 . Therecord management server 12 includes arecord management database 13 .

记录管理客户端11作为审批请求显示单元、审批单元(建议输入引导单元、学习数据输出单元)、文档制作单元(文档制作画面显示单元、制作文档输出单元、建议显示单元)等发挥功能。此外，记录管理客户端11也被称为文档审批装置、文档制作装置。记录管理客户端11可以单独地或者与记录管理服务器12以及记录管理数据库13一起被认为是记录管理系统或者记录管理装置。记录管理服务器12有时构成审批单元(学习数据输出单元)的一部分，有时也构成文档制作单元(制作文档输出单元)的一部分。Therecord management client 11 functions as approval request display unit, approval unit (advice input guidance unit, learning data output unit), document creation unit (document creation screen display unit, created document output unit, suggestion display unit), and the like. In addition, therecord management client 11 is also called a document approval apparatus or a document creation apparatus. Therecords management client 11 may be considered a records management system or a records management apparatus alone or together with therecords management server 12 and therecords management database 13 . Therecord management server 12 may constitute a part of an approval unit (learning data output unit), and may also constitute a part of a document creation unit (creation document output unit).

在记录管理系统10的整体中包含词素分析服务器14、学习服务器(学习单元)15以及指导服务器(指导单元)17，在学习服务器15附带有学习数据库16，在指导服务器17附属有手册数据库18。学习数据库16能够从学习服务器15以及指导服务器17访问。指导服务器17能够访问手册数据库18。有时将这些服务器14、15、17以及数据库16、18的集合称为学习指导服务器(系统)19。The entirerecord management system 10 includes amorphological analysis server 14, a learning server (learning unit) 15, and an instruction server (instruction unit) 17. The learningserver 15 is attached with alearning database 16, and theinstruction server 17 is attached with amanual database 18. Thelearning database 16 is accessible from the learningserver 15 and theinstruction server 17 . Theinstruction server 17 has access to themanual database 18 . The collection of theseservers 14 , 15 , 17 anddatabases 16 , 18 is sometimes referred to as a study guide server (system) 19 .

在一个实施方式中，学习指导服务器19能够利用被称为云的因特网上的服务器群(云计算)来实现。记录管理服务器12以及记录管理数据库13也能够利用云计算来实现。在这种情况下，各服务器执行的后述的处理程序被上传到云。词素分析服务器14以及学习服务器15也能够分别利用现有的词素分析服务器以及学习服务器来构筑该记录管理系统。In one embodiment, the learningguidance server 19 can be implemented using a server farm on the Internet called a cloud (cloud computing). Therecord management server 12 and therecord management database 13 can also be realized by cloud computing. In this case, a processing program to be described later executed by each server is uploaded to the cloud. Themorphological analysis server 14 and the learningserver 15 can also construct the record management system by using the existing morphological analysis server and learning server, respectively.

以下详述通过执行在构成记录管理系统10的各服务器中安装的程序而实现的功能或者动作的详细内容。The details of the functions or operations realized by executing the programs installed in each server constituting therecording management system 10 will be described in detail below.

首先，参照图2，对记录文档制作处理进行说明。First, with reference to FIG. 2 , the log file creation process will be described.

记录文档制作是在护理设施、医院、保育园等中使用记录管理客户端11由工作人员(护理者等负责人)日常进行的处理(操作)。所制作的记录文档的例子在图8中示出。也参照图8，工作人员在记录管理客户端11中启动记录编辑画面，从该画面的右栏的表单菜单中选择给护理管理者的报告书(护理管理报告书)。工作人员从画面上的下拉列表使用鼠标选择或者键盘等输入对象者(例如被护理者)的姓名(使用者：静冈太郎)、自己的姓名(报告者：田中一郎)、日期等，还输入“特别事项”的文章(文本数据)(图2、S11)。特别事项是报告事项的内容(文档内容)。另外，在该记录编辑画面的下部还显示有追加、保存、复制、删除等的框(按钮)。Record file creation is a process (operation) that is routinely performed by a staff (person in charge such as a caregiver) using therecord management client 11 in a nursing facility, hospital, nursery school, or the like. An example of the created log file is shown in FIG. 8 . Referring also to FIG. 8 , the staff member activates the record editing screen in therecord management client 11, and selects a report for the care manager (nursing management report) from the form menu in the right column of the screen. From the drop-down list on the screen, the staff member selects the name of the target person (for example, the care recipient) (user: Taro Shizuoka), his own name (reporter: Ichiro Tanaka), the date, etc., using the mouse selection or the keyboard, etc., and also inputs Article (text data) of "special matters" (Fig. 2, S11). The special matters are the contents (document contents) of the report matters. In addition, boxes (buttons) for adding, saving, copying, deleting, etc. are also displayed in the lower part of the record editing screen.

在工作人员制作的文档中有需要由上司(管理责任者等)审批的文档和不需要审批的文档。在需要审批的文档中，在记录编辑画面上显示审批栏(决定栏)A。在需要审批的文档中有各种业务日志、各种报告书等。在不需要审批的文档中不显示审批栏。通过审批栏控制(程序)如下所述地控制审批栏。Among the documents created by the staff, there are documents that require approval by a supervisor (person in charge of management, etc.) and documents that do not require approval. In a document requiring approval, an approval column (decision column) A is displayed on the record editing screen. There are various business logs, various reports, etc. in the documents that need to be approved. Approval columns are not displayed in documents that do not require approval. The approval bar is controlled as described below by the approval bar control (program).

在审批栏A中显示的审批者的数量根据文档的种类而不同。在此，为了简单，设置有能够指定2名审批者的框。工作人员输入主任佐藤和课长铃木作为审批者(图2、S12)(审批请求文档的制作、显示)。这样，通过输入并指定审批者的姓(名)，记录编辑画面上的文档成为审批请求文档。在有多个审批者的情况下确定审批的顺序(审批顺序)。在图8的例子中，为主任(佐藤)、课长(铃木)的顺序。审批工作流程以该审批顺序运行。The number of approvers displayed in the approval column A varies depending on the type of document. Here, for the sake of simplicity, there is provided a box where two approvers can be specified. The staff input Director Sato and Section Chief Suzuki as approvers (Fig. 2, S12) (production and display of approval request document). In this way, by inputting and specifying the surname (first name) of the approver, the document on the record editing screen becomes the approval request document. The order of approvals (approval order) is determined when there are multiple approvers. In the example of FIG. 8, it is the order of the director (Sato) and the section manager (Suzuki). Approval workflows run in this approval order.

审批工作流程本身是公知的技术。例如，审批请求文档被发送到记录管理服务器12，审批者和审批顺序的数据被保存。审批请求文档被存储在记录管理数据库13。然后，如果审批者从记录管理客户端11登录，则从记录管理服务器12向该记录管理客户端11发送表示对特定的文档被指定为审批者的意思的通知。审批者在记录管理客户端11中确认在审批状况查看器中自己应审批的文档的有无。审批者起动记录编辑器，将需要审批的文档显示在记录管理客户端11。工作人员、审批者分别具有专用的记录管理客户端，在这些记录管理客户端通过网络连接的情况下，也可以将表示被指定为审批者的意思的通知从制作了审批请求文档的工作人员的客户端11直接发送至审批者的客户端11。The approval workflow itself is a well-known technique. For example, an approval request document is sent to therecords management server 12, and data of approvers and approval order is saved. The approval request document is stored in therecords management database 13 . Then, when an approver logs in from therecords management client 11 , a notification indicating that a specific document is designated as an approver is transmitted from therecords management server 12 to therecords management client 11 . In therecord management client 11, the approver confirms the presence or absence of the document to be approved by himself/herself in the approval status viewer. The approver activates the record editor, and displays the document requiring approval on therecord management client 11 . The staff member and the approver each have a dedicated record management client, and when these record management clients are connected through a network, a notification indicating that they are designated as the approver may be sent from the staff member who created the approval request document. Theclient 11 sends directly to the approver'sclient 11.

接着，参照图3，对记录管理客户端11以及服务器12中的审批处理以及包含基于词素分析服务器14以及学习服务器15的词素分析处理在内的机械学习处理进行说明。3 , the approval process in therecord management client 11 and theserver 12 and the machine learning process including the morpheme analysis process by themorpheme analysis server 14 and the learningserver 15 will be described.

当审批者佐藤主任在记录管理客户端11中点击了审批栏A的框(印章栏)a1并且确定审批时，如图9所示，在框a1中显示佐藤的审批印。由此，由佐藤主任进行的审批(决定)结束(图3、S21、S22)。如从后述部分可知，佐藤主任查看特别事项(文档内容)的文章，进行“痴呆症的风险”这样的标记(建议的输入)。When the approver Director Sato clicks the frame (seal column) a1 of the approval column A in therecord management client 11 and confirms the approval, as shown in FIG. 9 , Sato's approval seal is displayed in the frame a1. Thereby, the approval (decision) by Director Sato is completed (FIG. 3, S21, S22). As can be seen from the later part, Director Sato checked the article of special matters (document content), and made a mark such as "risk of dementia" (suggested input).

由于审批的顺序是佐藤主任、铃木课长的顺序，因此当佐藤主任的审批结束后，在铃木课长的记录管理客户端11被通知有应审批的文档。在记录管理客户端11的显示画面上显示铃木课长应审批的文档(图3、S21)(审批请求文档的显示)。然后，铃木课长点击审批栏A的自己的姓名下的框(印章栏)a2。于是，如图10所示，显示评论(建议)输入画面B(建议输入引导(单元))。Since the order of approval is the order of Director Sato and Section Chief Suzuki, when the approval of Director Sato is completed, therecord management client 11 of Section Chief Suzuki is notified that there are documents to be approved. On the display screen of therecord management client 11, a document to be approved by Section Chief Suzuki is displayed (FIG. 3, S21) (display of approval request document). Then, Section Manager Suzuki clicks the box (seal column) a2 under his name in the approval column A. Then, as shown in FIG. 10, a comment (suggestion) input screen B (suggestion input guide (unit)) is displayed.

在该评论输入画面中，显示有审批状态的下拉列表/框b1、审批者名的显示框b2、审批日框b3、标签的下拉列表/框b4、追加按钮b5、确定按钮b6、取消按钮b7等。审批者铃木课长在审批状态框b1中选择审批。当然，认为审批者应阅读给护理管理者的报告书的特别事项(报告内容、文档内容)而应进行建议。在该例子中，着眼于“含痰”、“噎”这样的用语(如后所述被称为关键词)，从标签的下拉列表选择“咽下障碍的风险”(参照图11)。这样，当审批者阅读报告内容并选择成为适当的评论(建议)的一个或者多个标签(建议)时，该被选择的标签作为机械学习的教师数据在客户端11被接受。在标签列表中没有成为适当的评论(建议)的用语的情况下，能够按下追加按钮b5来追加评论。所追加的评论也被接受作为教师数据。当审批者按压确定按钮b6时，审批(决定)给护理管理者的报告书(图3、S22)。在建议中与审批请求文档的内容相对应地不仅包括对象者(被护理者等)，还包括与各种环境(设备、组织、工作人员个人)等相关的内容。On this comment input screen, a drop-down list/box b1 for approval status, a display box b2 for an approver name, an approval date box b3, a drop-down list/box b4 for tags, an add button b5, an OK button b6, and a cancel button b7 are displayed. Wait. Approver Manager Suzuki selects Approval in Approval Status Box b1. Of course, it is considered that the approver should read the special matters (report contents, document contents) of the report to the nursing manager, and should make suggestions. In this example, focusing on terms such as "sputum" and "choking" (referred to as keywords as described later), "risk of swallowing disturbance" is selected from the pull-down list of the label (see Fig. 11 ). In this way, when the approver reads the report content and selects one or more tags (suggestions) to be appropriate comments (suggestions), the selected tags are accepted at theclient 11 as teacher data for machine learning. When there is no term suitable for a comment (suggestion) in the tag list, the comment can be added by pressing the add button b5. The appended comments are also accepted as teacher data. When the approver presses the OK button b6, the report to the care manager is approved (decided) (FIG. 3, S22). The advice includes not only the target person (care recipient, etc.) but also content related to various environments (equipment, organization, individual staff member) and the like corresponding to the content of the approval request document.

由此，评论输入画面B消失，如图12所示，在审批栏A的审批者铃木的姓名下的印章栏a2中显示有铃木的印影。在审批栏A中作为审批者显示的全部人员进行了审批的输入(图3、S23中为是)，因此在记录编辑画面中显示的文档(给护理管理者的报告书)成为审批(决定)完毕文档。As a result, the comment input screen B disappears, and as shown in FIG. 12 , Suzuki's imprint is displayed in the stamp column a2 under the name of the approver Suzuki in the approval column A. Since all the persons displayed as approvers in the approval column A have entered the approval (YES in FIG. 3 , S23 ), the document (report to the nursing manager) displayed on the record editing screen is approved (decided). Complete the document.

记录管理客户端11响应最后的(全部的)审批者的审批输入，将已审批文档和输入的标签发送至记录管理服务器12(图3、S24)(审批单元、学习数据输出单元)。接收到该请求，记录管理服务器12将决定完毕的文档和输入的标签(这些构成学习用数据集)发送到词素分析服务器14(学习指导服务器19)(审批单元、学习数据输出单元)(图3、S25)。Therecords management client 11 responds to the approval input of the last (all) approvers, and transmits the approved documents and the input tags to the records management server 12 (FIG. 3, S24) (approval unit, learning data output unit). Upon receipt of this request, therecord management server 12 transmits the determined document and the input tag (these constitute the learning data set) to the morphological analysis server 14 (learning guidance server 19) (approval unit, learning data output unit) (FIG. 3 , S25).

从记录管理服务器12向词素分析服务器14(学习指导服务器19)发送的发送数据的例子在图13中示出。该发送数据包含发送目的地、文档种类(护理管理报告书＝给护理管理者的报告书)、迄今为止回复的记录文档的全部件数(作为一例为3000件，因此图13的文档(记录)为第3001件)、报告书的内容(“特别事项”、文档内容)、以及审批者(决定者)的姓名、职务、该审批者输入的标签(评论)和该审批者的加权值。在之前的例子中，铃木课长输入建议(评论)(标签)“咽下障碍的风险”，佐藤主任输入“痴呆症的风险”。加权值是审批者的权重，铃木课长输入的“咽下障碍的风险”的权重为0.7，佐藤主任输入的“痴呆症的风险”的权重为0.3，在后述的机械学习中使用。加权表示审批者输入的建议的确定性，考虑其经验、知识等而预先决定，并按照每个审批者存储在记录管理客户端11或者服务器12。An example of transmission data transmitted from therecord management server 12 to the morphological analysis server 14 (learning guidance server 19 ) is shown in FIG. 13 . The transmission data includes the transmission destination, the type of document (care management report=report to the nursing manager), and the total number of recorded documents that have been responded to so far (for example, 3000, so the document (record) in FIG. 13 is as follows: 3001), the contents of the report ("special matters", document contents), and the name and title of the approver (decision person), the label (comment) input by the approver, and the weighted value of the approver. In the previous example, Section Chief Suzuki entered advice (comment) (label) "risk of swallowing disorders" and Director Sato entered "risk of dementia". The weighting value is the weight of the approver. The weight of "risk of swallowing disorder" input by Section Chief Suzuki is 0.7, and the weight of "risk of dementia" input by Director Sato is 0.3, which are used in the machine learning described later. The weight indicates the certainty of the suggestion input by the approver, is determined in advance in consideration of the experience, knowledge, etc. of the approver, and is stored in therecord management client 11 or theserver 12 for each approver.

作为已审批文档，列举了给护理管理者的报告书，但除此之外还有受理单(intakesheet)(问诊单)、面单(face sheet)(详细的个人信息)、护理计划票、日报、月报等业务日志等。As the approved documents, the report to the nursing manager is listed, but there are also an intake sheet (inquiry sheet), a face sheet (detailed personal information), a nursing plan ticket, Daily, monthly and other business logs.

词素分析服务器14是进行将从记录管理服务器12发送的已审批文档中的文档内容(图13的文档内容、即图8等的特别事项)的文章分解为词素并判别其词性等的处理的服务器。以图13的文档内容为例提取动词和名词(如果需要，则提取它们的合成词)。例如，活力(vital)测量、心跳过速、血压安定、已参加(参加中)、含痰、噎、吃光、纸尿裤、排泄失败等用语。将这些用语称为关键词(以下，作为关键词的例子，为了简化，从图13的文档提取含痰、噎、心跳过速、排泄失败)。词素分析服务器14也可以具备收集了在与放置有该记录管理系统的设施相关的领域(护理、医疗、幼儿教育等)中经常使用的用语的辞典，参照该辞典仅提取相应领域的关键词。此外，动词和名词以及它们的结合语、合成词有时具有相同意思(例如，含痰和痰堵)，因此也可以具有将相同的意思的术语统一为名词或者动词中的任一个的处理功能。Themorphological analysis server 14 is a server that performs a process of decomposing the text of the document content (the document content of FIG. 13 , that is, the special matters of FIG. 8 and the like) in the approved document sent from therecord management server 12 into morphemes, and discriminating the part of speech, etc. . Taking the document content of FIG. 13 as an example, verbs and nouns (and their compound words, if necessary) are extracted. For example, terms such as vital (vital) measurement, tachycardia, stable blood pressure, participating (participating), phlegm, choking, eating up, diaper, and excretion failure. These terms are referred to as keywords (hereinafter, as examples of keywords, phlegm, choking, tachycardia, and voiding failure are extracted from the document of FIG. 13 for simplicity). Themorphological analysis server 14 may include a dictionary that collects terms frequently used in fields (nursing care, medical care, early childhood education, etc.) related to the facility where the record management system is installed, and refer to the dictionary to extract only keywords in the corresponding field. Also, since verbs and nouns, as well as their conjunctions and compound words sometimes have the same meaning (for example, phlegm and phlegm), it may also have a processing function of unifying terms with the same meaning into either noun or verb.

包含上述关键词的文档内容和对其审批时输入的标签(评论、建议)(教师数据)构成学习用数据集，但也可以将文档内容的关键词组和与其对应地输入的标签(评论、建议)的集合称为学习用数据集。这些学习用数据集被发送到学习服务器15(图3、S26)。The document content containing the above-mentioned keywords and the tags (comments, suggestions) (teacher data) entered when approving the keywords constitute the learning data set, but the keyword groups of the document content and the tags (comments, suggestions) input corresponding to them may be combined. ) is called a learning dataset. These learning data sets are sent to the learning server 15 ( FIG. 3 , S26 ).

学习服务器15进行有教师数据的机械学习。在此，作为一个例子，利用贝叶斯定理，预测其分类发生概率最高的(与关键词对应的标签(建议)的适合性(适合的程度)最高的或者超过规定的阈值)。The learningserver 15 performs machine learning with teacher data. Here, as an example, Bayes' theorem is used to predict the highest probability of occurrence of classification (the label (suggestion) corresponding to the keyword has the highest suitability (degree of suitability) or exceeds a predetermined threshold).

贝叶斯定理由以下公式表示。Bayes' theorem is expressed by the following formula.

P(A)＞0的情况，In the case of P(A)>0,

P(B|A)＝P(A|B)×P(B)/P(A)····式(1)P(B|A)=P(A|B)×P(B)/P(A)  … Equation (1)

在此，here,

P(A)：发生现象A的概率P(A): the probability of occurrence of phenomenon A

P(B)：发生现象B的概率P(B): the probability of occurrence of phenomenon B

P(A|B)：在发生了现象B后发生现象A的概率P(A|B): The probability of occurrence of phenomenon A after occurrence of phenomenon B

P(B|A)：在发生了现象A后发生现象B的概率。P(B|A): The probability of occurrence of phenomenon B after occurrence of phenomenon A.

如果将该贝叶斯定理代入包含关键词A和建议B的学习数据集，则如下所述。If this Bayes' theorem is substituted into a learning dataset containing keyword A and suggestion B, it is as follows.

P(A)：整体中的A这样的关键词(单词)的比例P(A): The ratio of keywords (words) such as A in the whole

P(B)：整体中的B这样的标签(建议)的比例P(B): Proportion of labels like B (suggested) in the ensemble

P(A|B)：在适用B这样的标签(建议)的情况下包含A这样的关键词的概率P(A|B): The probability of including a keyword such as A when a label such as B is applied (suggestion)

P(B|A)：在包含A这样的关键词的情况下适用B这样的标签(建议)的(适当的)概率P(B|A): The (appropriate) probability of applying a label (suggestion) such as B when a keyword such as A is included

在此，整体是指审批完毕(决定完毕)文档(记录)的整体(的数量)。Here, the whole refers to the whole (number) of the approval-completed (decided-completed) documents (records).

作为一例，在包含“噎”这样的单词的情况下，作为与“咽下障碍的风险相关的记录(文档)”的概率P(B|A)能够通过以下的计算求出。As an example, when the word "choking" is included, the probability P(B|A) that is "a record (document) related to the risk of swallowing disturbance" can be obtained by the following calculation.

在与“咽下障碍的风险相关的记录”中包含“噎”这样的单词的概率P(A|B)×全部的记录中的“与咽下障碍的风险相关的记录”的比例P(B)÷在全部的记录中包含“噎”这样的单词的比例P(A)。Probability P(A|B) that the word "choking" is included in "records related to the risk of swallowing disorders" × ratio of "records related to the risk of swallowing disorders" among all the records P(B )÷Proportion P(A) of the word "choking" in all records.

图14～图16表示存储在学习数据库16中的关键词主表(图14)、类别分类主表(图15)以及学习数据组(表格)(图16)。表示这些主表、数据组中记述的各比例的数值是使用过去的3000件记录(文档)计算出的值。所谓记录是指决定(审批)的文档的意思，记录的数量是文档的数量。14 to 16 show the keyword master table ( FIG. 14 ), the category classification master table ( FIG. 15 ), and the learning data group (table) ( FIG. 16 ) stored in thelearning database 16 . The numerical values representing the respective ratios described in these master tables and data sets are values calculated using the past 3000 records (documents). The so-called record means the document to be decided (approved), and the number of records is the number of documents.

参照图14，关键词主表与关键词对应地存储包含该关键词的决定完毕(审批结束)记录(文档)的比例P(A)。Referring to FIG. 14 , the keyword master table stores the ratio P(A) of the decision completed (approval completed) records (documents) including the keyword in association with the keyword.

参照图15，类别分类主表存储按照每个类别分类标记的决定完毕(审批结束)记录(文档)的比例P(B)。在此，将标签(建议、评论)称为类别分类，分别附加类别分类代码。在护理设施的情况下，根据被护理者的典型的或者代表的(成为管理的对象)行动或者状态被划分为区分，在各区分中附加有区分代码。而且，在各区分中存在多个类别分类。进而，在类别分类主表中记述有网站的URL，该URL关于各类别分类保存有手册，该手册记载了应该提供给工作人员的有利的信息(行动方针、注意事项等)。Referring to FIG. 15 , the category classification master table stores the ratio P(B) of the records (documents) that have been decided (approved) marked for each category classification. Here, labels (suggestions, comments) are called category classifications, and category classification codes are added to each. In the case of a nursing facility, categories are divided according to typical or representative actions or states of care recipients (subject to management), and a category code is added to each category. Furthermore, there are a plurality of classifications in each division. Furthermore, the URL of the website is described in the category classification master table, and this URL stores a manual for each category classification. The manual describes useful information (action guidelines, precautions, etc.) to be provided to the staff.

图16表示机械学习中的或者通过学习而得到的数据组。区分代码、类别分类代码、标记的决定完毕记录的比例P(B)与图15所示的相同。与类别分类对应地，存储有各关键词(A)、含有关键词的决定完毕的记录的比例P(A)、在被标记有类别分类(B)的决定完毕的记录中包含关键词(A)的比例P(A|B)以及包含关键词(A)的情况下被标记为类别分类(B)的比例P(B|A)。FIG. 16 shows a data set in machine learning or obtained by learning. The ratio P(B) of the determined records of the classification code, the category classification code, and the flag is the same as that shown in FIG. 15 . Corresponding to the category classification, each keyword (A), the ratio P(A) of the determined records containing the keyword, and the keyword (A) in the determined records marked with the category classification (B) are stored. ) and the proportion P(B|A) that is marked as a category classification (B) when the keyword (A) is included.

图3所示的S27的学习服务器15中的处理的详细内容如图4～图7所示。图4中的S31的处理1的详细内容如图5所示，S32中的处理2的详细内容如图6所示，S33中的处理3的详细内容如图7所示。Details of the processing in the learningserver 15 in S27 shown in FIG. 3 are shown in FIGS. 4 to 7 . The details of the process 1 in S31 in FIG. 4 are shown in FIG. 5 , the details of the process 2 in S32 are shown in FIG. 6 , and the details of the process 3 in S33 are shown in FIG. 7 .

图4的处理1(S31)是更新图14的关键词主表的处理。参照图5，从作为词素分析服务器14的分析结果而得到的关键词列表中一个一个地取得关键词，重复S42～S45的处理(S41)。在关键词列表中，如上所述，包含含痰、噎等的关键词。查看取得的关键词是否在关键词主表(图14)中(S42、S43)，如果有相应的关键词，则取出包含该关键词的记录的比例P(A)，并更新该值。例如，关键词“噎”的P(A)是0.0020，这是到全记录数3000件的值。即，在3000件的记录中关键词“噎”出现的件数3000件乘以0.0020而成为6件。这次，由于有包含关键词“噎”的记录(文档)，因此，新的P(A)成为，Process 1 ( S31 ) of FIG. 4 is a process of updating the keyword master table of FIG. 14 . Referring to FIG. 5 , keywords are acquired one by one from the keyword list obtained as an analysis result of themorphological analysis server 14 , and the processes of S42 to S45 are repeated ( S41 ). In the keyword list, as described above, keywords containing phlegm, choking, and the like are included. Check whether the acquired keyword is in the keyword master table (FIG. 14) (S42, S43). If there is a corresponding keyword, take out the ratio P(A) of records containing the keyword, and update the value. For example, P(A) of the keyword "choking" is 0.0020, which is a value up to 3000 pieces of the total number of records. That is, in the records of 3000 cases, the number of cases in which the keyword "choke" appears in 3000 cases is multiplied by 0.0020 to obtain 6 cases. This time, since there is a record (document) containing the keyword "choke", the new P(A) becomes,

P(A)＝(6+1)件÷3001件＝0.0023····式(2)。P(A)=(6+1)pieces÷3001pieces=0.0023... Formula (2).

在关键词主表中，将关键词“噎”的P(A)从0.0020更新为0.0023(S45)。In the keyword master table, the P(A) of the keyword "choking" is updated from 0.0020 to 0.0023 (S45).

在相应的关键词不是关键词主表的情况下(未登记的情况)，将该关键词追加到关键词主表，计算包含的记录的比例P(A)，并写入关键词主表(S44)。If the corresponding keyword is not in the keyword master table (if it is not registered), add the keyword to the keyword master table, calculate the proportion P(A) of the records included, and write it into the keyword master table ( S44).

作为词素分析的结果，若提取了例如“心跳过速”这样的关键词，则该关键词在图14的关键词主表中没有登记，因此，其比例P(A)计算为As a result of the morphological analysis, if a keyword such as "tachycardia" is extracted, the keyword is not registered in the keyword master table of Fig. 14, and therefore, its ratio P(A) is calculated as

P(A)＝1件÷3001件＝0.0003····式(3)，P(A) = 1 piece ÷ 3001 pieces = 0.0003... Formula (3),

与“心跳过速”这样的关键词一起，将其比例0.0003登记到关键词主表。Together with the keyword "tachycardia", the ratio of 0.0003 is registered in the keyword master table.

图4的处理2(S32)是更新图15的类别分类主表的处理。参照图6，取得从记录管理服务器12发送的数据中包含的全部的类别分类(评论、建议)，分别重复以下的S52～S55的处理。接收到的类别分类如下所述。Process 2 ( S32 ) in FIG. 4 is a process of updating the category classification master table in FIG. 15 . Referring to FIG. 6 , all categories (comments, suggestions) included in the data transmitted from therecord management server 12 are acquired, and the following processes of S52 to S55 are respectively repeated. The received category classification is described below.

咽下障碍的风险 铃木课长 加权值0.7Risk of Dysphagia Mr. Suzuki Weighted 0.7

痴呆症的风险 佐藤主任 加权值0.3Risk of Dementia Director Sato Weighted 0.3

上述的类别分类中的“咽下障碍的风险”存在于类别分类主表中。即，“咽下障碍的风险，0.1222”(S52、S53)。首先，将咽下障碍的风险的件数作为3000件×0.1222＝366.6件而求出。接下来，如下求出新的比例P(B)。此时，使用加权值0.7。The "risk of swallowing disorders" in the category classification above is present in the category classification main table. That is, "risk of dysphagia, 0.1222" (S52, S53). First, the number of cases of the risk of swallowing disturbance was obtained as 3000 cases×0.1222=366.6 cases. Next, a new ratio P(B) is obtained as follows. At this time, a weight value of 0.7 is used.

P(B)＝(366.6件+1件×0.7)÷3001件＝0.1224····式(4)P(B) = (366.6 pieces + 1 piece × 0.7) ÷ 3001 pieces = 0.1224... Equation (4)

用新计算出的值0.1224更新类别分类主表中的咽下障碍的风险0.1222(S55)。The risk of dysphagia 0.1222 in the main category classification table is updated with the newly calculated value 0.1224 (S55).

关于痴呆症的风险，由于在类别分类主表中没有登记，因此对其按如下方式计算P(B)。使用加权值0.3。Regarding the risk of dementia, since it was not registered in the main category classification table, P(B) was calculated as follows. Use a weighting value of 0.3.

痴呆症的风险的比例Dementia risk ratio

P(B)＝1件×0.3÷3001件＝0.0001····式(5)P(B) = 1 piece × 0.3 ÷ 3001 pieces = 0.0001... Equation (5)

在类别分类主表中新追加痴呆症的风险，并登记其比例P(B)＝0.0001(S54)。The risk of dementia is newly added to the main table of category classification, and its proportion P(B)=0.0001 is registered (S54).

图4的处理3(S33)是更新图16的学习数据组的处理。参照图7，对从记录管理服务器12接收到的被标记的全部的类别分类(评论、建议)和通过词素分析服务器14的分析而得到的所有的关键词的全部组合，重复从S62到S65的处理。Process 3 ( S33 ) of FIG. 4 is a process of updating the learning data set of FIG. 16 . Referring to FIG. 7, for all combinations of all categories (comments, suggestions) marked received from therecord management server 12 and all keywords obtained by the analysis of themorphological analysis server 14, the steps from S62 to S65 are repeated. deal with.

从学习数据组(图16)提取上述的类别分类和关键词的全部的组合的记录(S62)。在该例子中，全部的记录如下所述。Records of all combinations of the above-mentioned category classifications and keywords are extracted from the learning data set ( FIG. 16 ) ( S62 ). In this example, all records are as follows.

咽下障碍的风险+噎 P(A|B)＝0.0151Risk of dysphagia + choking P(A|B)=0.0151

咽下障碍的风险+心跳过速 没有记录Risk of dysphagia + tachycardia Not documented

咽下障碍的风险+排泄失败 没有记录Risk of dysphagia + voiding failure Not documented

........ .... ............ ......

痴呆症的风险+噎 没有记录Risk of dementia + choking Not documented

....... .... .... …  

由于存在第1个组合，因此对于该组合，如下所述地计算新的比例P(B|A)(S65)。Since the first combination exists, for this combination, a new ratio P(B|A) is calculated as follows (S65).

即，对于咽下的风险+“噎”，P(A|B)＝0.0151，总记录数为3000件，加权为0.7(铃木课长)。That is, for the risk of swallowing + "choking", P(A|B)=0.0151, the total number of records is 3000, and the weight is 0.7 (Section Manager Suzuki).

过去的咽下障碍的风险被标记的记录件数＝3000件×0.1222＝366.6件Number of records marked for risk of swallowing disorders in the past = 3000 x 0.1222 = 366.6

新的咽下障碍的风险被标记的记录件数＝(366.6件+(1件×0.7))＝367.3件Number of records marked for new risk of dysphagia = (366.6 + (1 x 0.7)) = 367.3

新的咽下障碍的风险被标记的记录中包含“噎”的记录件数＝366.6件×0.0151＝5.5357件New risk-marked records containing "choking" = 366.6 x 0.0151 = 5.5357

因此P(A|B)＝(5.5357件+1件)÷367.3件)＝0.0178Therefore P(A|B)=(5.5357+1)÷367.3)=0.0178

根据贝叶斯定理，P(B|A)＝P(A|B)×P(B)÷P(A)，因此使用P(A)＝0.0023(参照式(2))、P(B)＝0.1224(参照式(4))，P(A|B)＝0，0178×0.1224÷0.0023＝0.9473····式(7)According to Bayes' theorem, P(B|A)=P(A|B)×P(B)÷P(A), so use P(A)=0.0023 (refer to equation (2)), P(B) =0.1224 (refer to formula (4)), P(A|B)=0, 0178×0.1224÷0.0023=0.9473... formula (7)

在学习数据组(图16)中，对于咽下障碍的风险和噎的组合，使用通过式(6)以及式(7)表示的值，更新这些数据(S65)。In the learning data group ( FIG. 16 ), these data are updated using the values represented by the equations (6) and (7) for the combination of the risk of swallowing disorder and choking ( S65 ).

在图16的学习数据组中不存在类别分类和关键词的组合的记录的情况下，计算P(A|B)和P(B|A)，加入学习数据组(S64)。When there is no record of the combination of category classification and keyword in the learning data set of FIG. 16, P(A|B) and P(B|A) are calculated and added to the learning data set (S64).

例如，对于咽下障碍的风险+“心跳过速”，P(A|B)＝0，总记录件数3000件，加权0.7(铃木课长)。For example, for the risk of dysphagia + "tachycardia", P(A|B)=0, the total number of records is 3000, and the weight is 0.7 (Section Manager Suzuki).

过去的咽下障碍的风险被标记的记录件数＝3000件×0.1222＝366.6件Number of records marked for risk of swallowing disorders in the past = 3000 x 0.1222 = 366.6

新的咽下障碍的风险被标记的记录件数＝(366.6件+(1件×0.7))＝367.3件Number of records marked for new risk of dysphagia = (366.6 + (1 x 0.7)) = 367.3

在新的咽下障碍的风险被标记的记录中包含“心跳过速”的记录件数＝366.6件×0＝0件Number of records containing "tachycardia" in records flagged at risk for new dysphagia = 366.6 x 0 = 0

因此P(A|B)＝(0件+1件)÷367.3件)＝0.0027····式(8)Therefore, P(A|B)=(0+1)÷367.3)=0.0027... Equation (8)

此外，对于P(B|A)，假设P(A)、P(B)被计算为P(A)＝0.0003，P(B)＝0.0001，Furthermore, for P(B|A), assuming that P(A), P(B) are calculated as P(A)=0.0003, P(B)=0.0001,

则P(B|A)＝P(A|B)×P(B)÷P(A)＝0.0027×0.0003÷0.0001＝0.0009····式(9)Then P(B|A)=P(A|B)×P(B)÷P(A)=0.0027×0.0003÷0.0001=0.0009... Equation (9)

在学习数据组(图16)中，新设置有咽下障碍的风险+心跳过速的记录，分别登记由式(8)、(9)计算出的值。In the learning data set ( FIG. 16 ), a record of risk of swallowing disturbance + tachycardia is newly set, and the values calculated by equations (8) and (9) are registered, respectively.

对应用了贝叶斯定理的机械学习进行了详述，但也可以通过其他机械学习，例如模糊推理来进行有教师数据的机械学习。Machine learning that applies Bayes' theorem is detailed, but machine learning with teacher data is also possible through other machine learning, such as fuzzy inference.

接下来，说明在工作人员基于上述的机械学习的结果在记录管理客户端系统11中要输入文档时、正在输入时、输入结束时等，对输入的(正在输入的)内容进行与护理、看护、保育等相关的建议(指导)的结构。Next, when a worker is about to input a document in the recordmanagement client system 11 based on the result of the above-mentioned machine learning, when inputting a document, when inputting is completed, etc., it will be described how to perform nursing care and nursing on the inputted (inputting) content. , childcare, etc. related advice (guidance) structure.

图20表示在记录管理客户端系统11中显示的画面的例子(文档制作、文档制作单元)。工作人员在表单菜单中，例如选择早餐时，进餐记录输入(编辑)画面(窗口)C出现，工作人员通过数字、文章、确认等输入早餐的进餐量、进餐的情况、口腔护理。在画面的下方，除了删除、复制、保存、追加等通常的按钮以外，还设置有指导显示按钮D。进行输入的工作人员(记录者)被显示为特许一郎，设施的使用者(被护理者)显示为静冈五郎。FIG. 20 shows an example of a screen displayed in the record management client system 11 (document creation, document creation means). For example, when a staff member selects breakfast in the form menu, the meal record input (edit) screen (window) C appears, and the staff member enters the amount of breakfast meals, meal conditions, and oral care through numbers, texts, and confirmation. In the lower part of the screen, in addition to the usual buttons such as delete, copy, save, and add, a guide display button D is provided. The staff member (recording person) who made the input is displayed as Jōichiro, and the user (care receiver) of the facility is displayed as Shizuoka Goro.

当记录者(工作人员)按下指导显示按钮D或者保存按钮D时，如后所述，输入的记录内容从记录管理客户端11(也可以经由记录管理服务器12)被发送到词素分析服务器14(制作文档输出、制作文档输出单元)。词素分析服务器14从记录内容中提取与护理(看护、保育)等关联性高的关键词(名词、动词)。例如，在图20的输入(记录)画面(窗口C)的例子中，提取“有嗜睡”、“有晕眩”、“噎”等，这些关键词被发送到指导服务器17。When the recorder (personnel) presses the instruction display button D or the save button D, the input record content is transmitted from the record management client 11 (or via the record management server 12 ) to themorphological analysis server 14 as will be described later. (Make document output, make document output unit). Themorphological analysis server 14 extracts keywords (nouns, verbs) that are highly related to nursing care (nursing care, childcare) and the like from the recorded content. For example, in the example of the input (recording) screen (window C) of FIG. 20 , keywords such as “drowsiness”, “dizziness”, and “choking” are extracted, and these keywords are sent to theguidance server 17 .

指导服务器17对于从词素分析服务器14发送的关键词，参照学习数据库16内的学习数据组(图16)，提取与这些关键词关联性强的(高的)建议并制作指导，发送至记录管理客户端11。在客户端11中，如图21所示，从指导服务器发送的指导被显示在指导区域E(建议显示、建议显示单元)。Theguidance server 17 refers to the learning data group ( FIG. 16 ) in thelearning database 16 for the keywords sent from themorphological analysis server 14 , extracts suggestions that are highly (high) related to these keywords, creates guidance, and sends them to therecord management Client 11. In theclient 11, as shown in FIG. 21, the guidance transmitted from the guidance server is displayed in the guidance area E (advice display, advice display means).

在图21所示的例子中，在指导区域E中，显示有由指导服务器17选择出的睡眠障碍的风险和咽下障碍的风险，对于各风险，显示其详细内容、护理手册、预防手册等的文字(详细指导项目)。当工作人员(记录者)选择(点击)某个详细指导项目时，其详细的内容显示在客户端11的画面上。在图22中，作为选择了“误咽防止手册”的结果，将误咽防止手册显示在窗口F中。在该手册中，工作人员能够进一步选择详细内容(防止误咽的要点、进餐的姿势、进餐的内容、口腔护理)，来阅读所显示得解说，对于误咽进行学习，加深理解，能够对(护理等的)业务有用。In the example shown in FIG. 21 , in the guidance area E, the risk of sleep disorder and the risk of swallowing disorder selected by theguidance server 17 are displayed, and for each risk, its details, nursing manual, prevention manual, etc. are displayed. text (detailed guidance items). When a staff member (recorder) selects (clicks) a certain detailed instruction item, its detailed content is displayed on the screen of theclient 11 . In FIG. 22 , as a result of the selection of the "Accidental Swallowing Prevention Manual", the window F is displayed with the Aspiration Prevention Manual. In this manual, the staff can further select detailed contents (points to prevent swallowing, posture of eating, contents of eating, oral care), read the explanations displayed, learn about swallowing, deepen understanding, and be able to ( Nursing, etc.) business is useful.

以下，参照图17～图19，详述根据工作人员在客户端11上的操作(输入)而产生图20～图22所示那样的显示的客户端11、按照词素分析服务器14和指导服务器17的程序的处理。17 to 19 , theclient 11 , themorphological analysis server 14 , and theguidance server 17 , which generate a display as shown in FIGS. 20 to 22 according to the operation (input) of the staff on theclient 11 , will be described in detail. program processing.

工作人员(包括主任、课长等管理者)在客户端11的画面上选择表单的种类，并输入与此相应的记录。或者，工作人员从记录管理服务器12(记录管理数据库13)中提取规定的文档(表单)(包含其他工作人员制作的文档)并显示在客户端11画面上进行阅览(图17、S71)。A staff member (including managers such as a director and a section manager) selects the type of form on the screen of theclient 11, and inputs the corresponding record. Alternatively, the worker extracts a predetermined document (form) (including documents created by other workers) from the record management server 12 (record management database 13), displays it on the screen of theclient 11, and browses it (FIG. 17, S71).

客户端11将所显示的记录文档发送到词素分析服务器14(S72)。客户端11发送文档的时刻可以是如上所述按下指导显示按钮D时、保存按钮被按下时、其他任意时刻(例如，从开始输入起每隔一定时间、一定时间没有输入时等)。在图23中示出了从记录管理客户端11发送到词素分析服务器14的信息(数据)的一例。发送目的地可以不是学习指导服务器(系统)，而是词素分析服务器。文档的种类是早餐记录、文档内容被输入窗口C并被显示的文章等。Theclient 11 transmits the displayed log document to the morphological analysis server 14 (S72). The time when theclient 11 transmits the document may be when the guidance display button D is pressed as described above, when the save button is pressed, or at any other time (for example, every certain time from the start of input, when there is no input for a certain period of time, etc.). An example of information (data) transmitted from therecord management client 11 to themorphological analysis server 14 is shown in FIG. 23 . The transmission destination may not be a learning guidance server (system) but a morphological analysis server. The types of documents are breakfast records, articles whose contents are input into the window C and displayed, and the like.

词素分析服务器14进行接收到的文档内容的词素分析，参照所需的专业术语辞典(护理、看护、医疗、保育等辞典)，提取该领域(设施)特有的关键词(名词、动词)。在上述的例子中，提取出有嗜睡、晕眩(有晕眩)、噎等。将所提取得关键词从词素分析服务器14发送到指导服务器17(S73)。Themorphological analysis server 14 performs morphological analysis of the received document content, refers to a necessary terminology dictionary (nursing, nursing, medical, nursing, etc.), and extracts keywords (nouns, verbs) specific to the field (facility). In the above-mentioned example, drowsiness, dizziness (with dizziness), choking, and the like were extracted. The extracted keywords are sent from themorphological analysis server 14 to the guidance server 17 (S73).

指导服务器17在接收到关键词后，进行处理4(S74)、处理5(S75)。处理4、处理5的详细内容分别在图18、图19中示出。After receiving the keyword, theguidance server 17 performs process 4 (S74) and process 5 (S75). The details of the process 4 and the process 5 are shown in FIG. 18 and FIG. 19 , respectively.

处理4使用从词素分析服务器14发送的全部的关键词，搜索学习数据库16的学习数据组，提取与一致的关键词对应的类别分类(建议)中的、概率高的类别分类(图17、S74)。即，参照图18，首先在学习数据组(图16)中，提取具有与接收的关键词一致的关键词的记录(图18、S81)。在图16的例子中，提取出具有噎、有嗜睡、晕眩这样的关键词的记录。接下来，提取出的关键词的记录中具有规定的阈值(例如设为0.5)以上的P(B|A)的记录被提取出(图18、S82)。在图16的学习数据组中，关键词“噎”的P(B|A)的值为0.9226，关键词“有嗜睡”的P(B|A)的值为0.6452，关键词“有晕眩”的P(B|A)为0.3047，因此具有阈值以上的P(B|A)的关键词为“噎”和“有嗜睡”。阈值要通过试行错误而确定，以得到适当的指导。Process 4 uses all the keywords sent from themorphological analysis server 14 to search the learning data set of thelearning database 16, and extracts a category with a high probability among category categories (suggestions) corresponding to the matching keyword ( FIG. 17 , S74 ). ). That is, referring to FIG. 18, first, in the learning data group (FIG. 16), a record having a keyword matching the received keyword is extracted (FIG. 18, S81). In the example of FIG. 16 , records having keywords such as choking, drowsiness, and dizziness are extracted. Next, among the records of the extracted keywords, records having a P(B|A) greater than or equal to a predetermined threshold (for example, set to 0.5) are extracted ( FIG. 18 , S82 ). In the learning data set of Fig. 16, the value of P(B|A) for the keyword "choking" is 0.9226, the value of P(B|A) for the keyword "drowsiness" is 0.6452, and the value of the keyword "dizziness" is 0.6452. The P(B|A) of " is 0.3047, so the keywords with P(B|A) above the threshold are "choking" and "somnolence". Thresholds are to be determined by trial and error for proper guidance.

在具有这样提取的关键词的记录中，还存在类别分类重复的记录，因此除去重复的类别分类之外，制作提取出的关键词的记录中包含的类别分类的列表(图18、S83)。在上述的例子中，虽然提取了“噎”和“有嗜睡”作为关键词，但这些关键词的记录中包含的类别分类由于是咽下障碍的风险和睡眠障碍的风险，因此没有重复，制作这两个类别分类的列表。Among the records having the keywords extracted in this way, there are records with overlapping categories. Therefore, a list of categories included in the records of the extracted keywords is created in addition to the overlapping categories ( FIG. 18 , S83 ). In the above example, although "choking" and "drowsiness" were extracted as keywords, the categories included in the records of these keywords were not duplicated because of the risk of swallowing disorder and the risk of sleep disorder. A list of these two categories.

在处理5中，指导服务器17对于这样制作的类别分类的列表中的全部的类别分类，参照类别分类主表(图15)将有用的手册关联起来发送到记录管理客户端11(图17、S75)(指导、指导单元)。即，在图19中，参照类别分类主表(图15)，取得与在图18、S83中选择出的类别分类对应的各种手册(在全部的手册或者特定的条件下选择的手册)，存储在响应信息(响应数据)中，并编辑成HTML形式(图19、S91、S92)，并发送给记录管理客户端11(图19、S93)。从指导服务器17发送到记录管理客户端11的信息的例子如图24所示。与类别分类对应的手册的URL等以显示用HTML形式进行编辑而被包含。In process 5, theguidance server 17 refers to the category master table ( FIG. 15 ) for all the categories in the category list created in this way, and associates the useful manuals with the record management client 11 ( FIG. 17 , S75 ). ) (guidance, guidance unit). That is, in FIG. 19, referring to the category classification master table (FIG. 15), various manuals (all manuals or manuals selected under specific conditions) corresponding to the category selected in FIG. 18 and S83 are acquired, It is stored in the response information (response data), edited into HTML (FIG. 19, S91, S92), and sent to the record management client 11 (FIG. 19, S93). An example of the information sent from theguidance server 17 to therecord management client 11 is shown in FIG. 24 . The URL of the manual corresponding to the category classification is included by editing it in HTML format for display.

返回图17，接收到来自指导服务器17的回复的记录管理客户端11将类别分类和有用的手册的菜单显示在显示画面中的指导区域E中(S76)，当被显示的类别分类或者其中的详细指导项目被点击，则使用URL从网站取入相应的手册，制作指导窗口F并显示在其内部(S77)。Returning to FIG. 17 , therecord management client 11, which has received the reply from theguidance server 17, displays the category classification and the menu of useful manuals in the guidance area E in the display screen (S76), when the displayed category classification or among the menus are displayed (S76). When the detailed instruction item is clicked, the corresponding manual is fetched from the website using the URL, and an instruction window F is created and displayed in it (S77).

Claims (23)

1. A record management system is provided with:

an approval request display unit which displays an approval request document containing an approval bar;

an approval unit that accepts a suggestion input in correspondence with a keyword included in a displayed approval request document, and outputs a learning dataset including document contents and the accepted suggestion in response to the approval input, the document contents including the keyword of the approval request document;

a learning unit that learns the degree of suitability of the advice corresponding to the keyword based on the learning dataset output from the approval unit;

a document creation unit that outputs a document onto a screen; and

a guidance unit that outputs a suggestion appropriate for the keyword included in the document output from the document creation unit based on a learning result of the learning unit,

the document producing unit displays the advice output from the guidance unit on a document output screen.

2. A record management system is provided with:

an approval request display unit which displays an approval request document containing an approval bar;

an approval unit that accepts a suggestion input in correspondence with a keyword included in a displayed approval request document and outputs a learning data set including document content and the accepted suggestion to a learning unit in response to the approval input, the learning unit learning a degree of suitability of the suggestion corresponding to the keyword, the document content including the keyword of the approval request document;

a document creation unit that outputs a document onto a screen; and

a guidance unit that outputs a suggestion appropriate for the keyword included in the document output from the document production unit based on the learning result of the learning unit,

the document producing unit displays the advice output from the guidance unit on a document output screen.

3. A record management system is provided with:

an approval request display unit which displays an approval request document containing an approval bar;

an approval unit that accepts a suggestion input in correspondence with a keyword included in a displayed approval request document and outputs a learning data set including document content and the accepted suggestion to a learning unit in response to the approval input, the learning unit learning a degree of suitability of the suggestion corresponding to the keyword, the document content including the keyword of the approval request document; and

and a document creation unit that outputs the document displayed on the screen to the guidance unit, and displays on the screen a suggestion that is output from the guidance unit based on a learning result of the learning unit and that is appropriate for the keyword contained in the document.

4. The record management system according to any one of claims 1 to 3,

the approval unit includes:

a suggestion input guide unit which displays a suggestion input screen in response to a1 st input of a series of approval inputs related to the approval bar; and

and a learning data output unit that accepts the advice input on the advice input screen and outputs a learning data set including document contents including the keyword of the displayed approval request document and the accepted advice to the learning unit in response to a2 nd input of the series of approval inputs.

5. The record management system according to any one of claims 1 to 3,

the document creation unit includes:

a production document display unit that displays a production document on a screen;

a document creation output unit that outputs the document displayed on the screen to the guidance unit; and

and a suggestion display unit that displays on the screen suggestions that are output from the guidance unit based on the learning result of the learning unit and that are appropriate for the keywords included in the production document.

6. A recording management device is provided with:

an approval request display unit which displays an approval request document containing an approval bar;

an approval unit that accepts a suggestion input in correspondence with a keyword included in a displayed approval request document and outputs a learning data set including document content and the accepted suggestion to a learning unit in response to the approval input, the learning unit learning a degree of suitability of the suggestion corresponding to the keyword, the document content including the keyword of the approval request document; and

and a document creation unit that outputs the document displayed on the screen to the guidance unit, and displays on the screen a suggestion that is output from the guidance unit based on a learning result of the learning unit and that is appropriate for the keyword contained in the document.

7. A document approval apparatus includes:

an approval request display unit which displays an approval request document including an approval bar;

a suggestion input guide unit which displays a suggestion input screen in response to a1 st input of a series of approval inputs related to the approval bar; and

and a learning data output unit that accepts the advice input on the advice input screen and outputs a learning data set including document contents including the keyword of the displayed approval request document and the accepted advice to the learning unit in response to a2 nd input of the series of approval inputs.

8. The document approval apparatus of claim 7,

the learning data output unit outputs an identification code of an approver who inputs a series of approval inputs together with the learning data set.

9. The document approval apparatus of claim 7,

the learning data output unit outputs weighted data of an approver together with a learning data set.

10. A document creation device is provided with:

a production document display unit that displays a production document on a screen;

a document creation output unit that outputs the document displayed on the screen to the guidance unit; and

and a suggestion display unit that displays on the screen suggestions that are output from the guidance unit based on the learning result of the learning unit and that are appropriate for the keywords included in the production document.

11. The document making device according to claim 10,

the advice display unit displays the detailed information output from the guidance unit together with the advice.

12. A method for managing a record of a recording medium,

the approval request display unit displays an approval request document including an approval bar,

the approval unit accepts a suggestion input corresponding to a keyword included in the displayed approval request document, outputs a learning data set including document contents and the accepted suggestion in response to the approval input, the document contents including the keyword of the approval request document,

the learning unit learns the degree of suitability of the advice corresponding to the keyword based on the learning dataset output from the approval unit,

the document producing unit outputs the document onto the screen,

the guiding unit outputs a suggestion suitable for the keyword included in the document output from the document producing unit based on the learning result of the learning unit, and then

The document producing unit displays the advice output from the guidance unit on a document output screen.

13. A method for managing a record of a recording medium,

the approval request display unit displays an approval request document including an approval bar,

the approval unit accepts a suggestion input corresponding to a keyword included in the displayed approval request document, outputs a learning data set including document contents and the accepted suggestion to the learning unit in response to the approval input, the learning unit learns a degree of suitability of the suggestion corresponding to the keyword, the document contents including the keyword of the approval request document,

the document producing unit outputs the document onto the screen,

the guiding unit outputs a suggestion appropriate for the keyword included in the document output from the document producing unit based on the learning result of the learning unit, and then

The document producing unit displays the advice output from the guidance unit on a document output screen.

14. A method for managing a record of a recording medium,

the approval request display unit displays an approval request document including an approval bar,

the approval unit accepts a suggestion input corresponding to a keyword contained in the displayed approval request document, outputs a learning data set containing document contents and the accepted suggestion to a learning unit in response to the approval input, the learning unit learns a degree of suitability of the suggestion corresponding to the keyword, the document contents contain the keyword of the approval request document, and then

The document producing unit outputs the document displayed on the screen to the guidance unit, and displays on the document display screen a suggestion that is appropriate for the keyword contained in the document and that is output from the guidance unit based on the learning result of the learning unit.

15. A method for managing a record of a recording medium,

the approval requesting unit displays an approval requesting document including an approval bar,

the approval unit accepts a suggestion input corresponding to a keyword contained in the displayed approval request document, outputs a learning data set containing document contents and the accepted suggestion to a learning unit in response to the approval input, the learning unit learns a degree of suitability of the suggestion corresponding to the keyword, the document contents contain the keyword of the approval request document, and then

The document producing unit outputs the document displayed on the screen to the guidance unit, and displays on the document display screen a suggestion that is appropriate for the keyword contained in the document and that is output from the guidance unit based on the learning result of the learning unit.

16. A document examination and approval method is provided,

the approval requesting unit displays an approval requesting document including an approval bar,

the suggestion input guide unit displays a suggestion input screen in response to a1 st input of a series of approval inputs related to the approval bar, and then

The learning data output unit accepts a suggestion input on the suggestion input screen and outputs a data set for learning containing document contents containing keywords of the displayed approval request document and the accepted suggestion to the learning unit in response to a2 nd input of a series of approval inputs.

17. A method for making a document includes the steps of,

the production document display unit displays the production document on the screen,

the produced document output unit outputs the document displayed on the screen to the guide unit, and then

The suggestion display unit displays, on a document creation screen, suggestions that are appropriate for the keywords included in the created document and that are output from the guidance unit based on the learning result of the learning unit.

18. A recording medium having recorded thereon a recording management program that controls a computer to execute:

the approval request display unit displays an approval request document including an approval bar,

the approval unit accepts a suggestion input corresponding to a keyword included in the displayed approval request document, outputs a learning data set including document contents and the accepted suggestion in response to the approval input, the document contents including the keyword of the approval request document,

the learning unit learns the degree of suitability of the advice corresponding to the keyword based on the learning dataset output from the approval unit,

the document producing unit outputs the document onto the screen,

the guiding unit outputs a suggestion suitable for the keyword included in the document output from the document producing unit based on the learning result of the learning unit, and then

The document producing unit displays the advice output from the guidance unit on a document output screen.

19. A recording medium having recorded thereon a recording management program that controls a computer to execute:

the approval request display unit displays an approval request document including an approval bar,

the approval unit accepts a suggestion input corresponding to a keyword included in the displayed approval request document, outputs a learning data set including document contents and the accepted suggestion to the learning unit in response to the approval input, the learning unit learns a degree of suitability of the suggestion corresponding to the keyword, the document contents including the keyword of the approval request document,

the document producing unit outputs the document onto the screen,

the guiding unit outputs a suggestion appropriate for the keyword included in the document output from the document producing unit based on the learning result of the learning unit, and then

The document producing unit displays the advice output from the guidance unit on a document output screen.

20. A recording medium having recorded thereon a recording management program that controls a computer to execute:

the approval request display unit displays an approval request document including an approval bar,

the approval unit accepts a suggestion input corresponding to a keyword contained in the displayed approval request document, outputs a learning data set containing document contents and the accepted suggestion to a learning unit in response to the approval input, the learning unit learns a degree of suitability of the suggestion corresponding to the keyword, the document contents contain the keyword of the approval request document, and then

The document producing unit outputs the document displayed on the screen to the guidance unit, and displays on the document display screen a suggestion that is appropriate for the keyword contained in the document and that is output from the guidance unit based on the learning result of the learning unit.

21. A recording medium having recorded thereon a recording management program that controls a computer to execute:

the approval requesting unit displays an approval requesting document including an approval bar,

the approval unit accepts a suggestion input corresponding to a keyword contained in the displayed approval request document, outputs a learning data set containing document contents and the accepted suggestion to a learning unit in response to the approval input, the learning unit learns a degree of suitability of the suggestion corresponding to the keyword, the document contents contain the keyword of the approval request document, and then

The document producing unit outputs the document displayed on the screen to the guidance unit, and displays on the document display screen a suggestion that is appropriate for the keyword contained in the document and that is output from the guidance unit based on the learning result of the learning unit.

22. A recording medium having recorded thereon a document approval program that controls a computer to execute:

the approval requesting unit displays an approval requesting document including an approval bar,

the suggestion input guide unit displays a suggestion input screen in response to a1 st input of a series of approval inputs related to the approval bar, and then

The learning data output unit accepts a suggestion input on the suggestion input screen and outputs a data set for learning containing document contents containing keywords of the displayed approval request document and the accepted suggestion to the learning unit in response to a2 nd input of a series of approval inputs.

23. A recording medium having recorded thereon a document making program that controls a computer to execute:

the production document display unit displays the production document on the screen,

the produced document output unit outputs the document displayed on the screen to the guide unit, and then

The suggestion display unit displays on a screen suggestions that are appropriate for the keywords included in the production document and that are output from the guidance unit based on the learning result of the learning unit.

Images