WO2021084807A1

WO2021084807A1 - Information-providing system

Info

Publication number: WO2021084807A1
Application number: PCT/JP2020/026693
Authority: WO
Inventors: 黒田　聡
Original assignee: 株式会社情報システムエンジニアリング
Priority date: 2019-10-31
Filing date: 2020-07-08
Publication date: 2021-05-06
Also published as: JP6703698B1; JP2021071992A

Abstract

[Problem] To provide a technique that enables facilitation of comprehension of a portion to be edited. [Solution] The information-providing system according to the present invention comprises: a relational database that stores a plurality of pieces of basic information obtained by dividing a basic content into chunk structures, and a plurality of pieces of object information obtained by dividing an object content into chunk structures, in association with each other; a database that is for calculation of basic information similarity and that is constructed by machine learning using the plurality of pieces of basic information; a basic information-obtaining means for obtaining a specific piece of basic information; a basic information-comparing means for comparing the pieces of basic information with the specific piece of basic information; a basic information similarity-calculating means for, when the specific piece of basic information is not identical to any of the pieces of basic information, calculating the basic information similarities indicating similarities between the pieces of basic information and the specific piece of basic information by referring to the database for calculation of the basic information similarity; and an object information-extraction means for extracting, as a first piece of object information, one of the pieces of object information corresponding to a selected first piece of basic information, on the basis of the basic information similarities.

Description

Information provision system

The present invention relates to an information providing system.

The information processing device of Patent Document 1 includes a reading means for reading a completed document, a recognition means for recognizing a fixed type and information structure of the document, a storage means for storing the contents written in the document, and a read document. If it is not the latest version, the information structure in the latest version of the pre-registered standard type corresponding to the recognized standard type and the comparison means for comparing the recognized information structure are changed as a result of comparison. A generation means for generating a difference document having an information structure, an output means for outputting the generated difference document, and a document entered in the difference document and a document stored in the storage means when the difference document entered by the user is read. A merging means for merging the entered contents is provided, and the storage means replaces the stored contents with the contents merged by the merging means and stores the contents.

Japanese Unexamined Patent Publication No. 2017-22431

By the way, for example, when a device is upgraded from an old version to a new version, it is necessary to edit the device manual from the old one to the new one. Since the device manual is created based on the specifications of the product, the creator of the device manual needs to create a new manual based on the new specifications.

However, the information processing apparatus disclosed in Patent Document 1 recognizes a fixed type and an information structure, and compares an old manual which is an old document with a new manual which is a new document having the same attributes. It is a thing. Therefore, there is a problem that the edited part in the manual cannot be easily grasped from the specifications having different attributes.

Therefore, the present invention was devised in view of the above-mentioned problems, and an object of the present invention is to provide an information providing system capable of easily grasping an edited part.

The information providing system according to the present invention associates a plurality of basic information in which the basic content is divided into chunk structures with a plurality of target information in which the target content having an attribute different from the basic content is divided into chunk structures. The relevance database stored in the above, the database for calculating the similarity of basic information constructed by machine learning using the plurality of basic information, the basic information acquisition means for acquiring specific basic information, the basic information, and the basic information. When the basic information comparison means for comparing the specific basic information and the basic information comparison means do not match the basic information with the specific basic information, the basic information similarity calculation database is referred to and the basics are referred to. Basic information indicating the similarity between the information and the specific basic information Basic information similarity calculation means for calculating the similarity, and the first basic information from a plurality of the basic information based on the basic information similarity. It is characterized by comprising a target information extraction means for selecting, referring to the relevance database, and extracting the target information corresponding to the first basic information as the first target information.

According to the present invention, it is possible to provide a technique capable of easily grasping a part to be edited.

FIG. 1 is a schematic diagram showing an example of the configuration of the information providing system according to the present embodiment. FIG. 2 is a schematic diagram showing an example of using the information providing system in the present embodiment. FIG. 3 is a schematic diagram showing an example of the relevance database of the information providing system in the present embodiment. FIG. 4 is a schematic diagram showing an example of a database for calculating the basic information similarity of the information providing system according to the present embodiment. FIG. 5 is a schematic diagram showing an example of a database for calculating the target information similarity of the information providing system according to the present embodiment. FIG. 6 is a schematic diagram showing an example of the configuration of the information providing device of the information providing system according to the present embodiment. FIG. 7 is a schematic diagram showing an example of the function of the information providing device of the information providing system according to the present embodiment. FIG. 8 is a flowchart showing an example of the operation of the information providing system according to the present embodiment. FIG. 9 is a schematic diagram showing a second example of the information providing system according to the present embodiment.

Hereinafter, an example of the information providing system according to the embodiment of the present invention will be described with reference to the drawings.

(Configuration of information providing system 100)
FIG. 1 is a block diagram showing the overall configuration of the information providing system 100 according to the present embodiment.

The information providing system 100 is used by a user such as a manual creator who newly creates a manual for the device based on the specifications of the device, for example.

As shown in FIG. 1, the information providing system 100 includes an information providing device 1. The information providing device 1 may be connected to the user terminal 5 or the server 6 via, for example, the public communication network 7.

FIG. 2 is a schematic diagram showing an example using the information providing system 100 in the present embodiment. The information providing device 1 acquires specific basic information x in the basic content X. The information providing device 1 calculates the basic information similarity with respect to the acquired specific basic information x. The information providing device 1 selects the first basic information b1 from a plurality of basic information based on the calculated basic information similarity. The information providing device 1 refers to the relevance database and extracts the target information B1 corresponding to the selected first basic information b1 as the first target information. As a result, it is possible to grasp that the target information B1 corresponding to the acquired basic information b1 similar to the specific basic information x is an edited part based on the specific basic information x. Therefore, when editing the manual or the like as the target content from the specifications or the like as the basic content, it is only necessary to edit the target information B1 and the editing work of the target content can be performed in a short time.

Further, the information providing device 1 refers to the target information similarity estimation processing database and calculates the target information similarity with respect to the first target information B1. The information providing device 1 extracts the second target information B2 different from the first target information B1 based on the calculated target information similarity. As a result, it can be understood that the second target information B2, which is similar to the first target information B1, is also an edited part based on the specific basic information x. Therefore, when editing a manual or the like as a target content from a specification or the like as a basic content, it is only necessary to edit the first target information and the second target information, and the editing work of the target content is performed in a short time. be able to.

<Relevance database>
FIG. 3 is a schematic diagram showing an example of the relevance database of the information providing system in the present embodiment. In the relevance database, a plurality of basic information and target information are stored in a one-to-one correspondence. The relevance database corresponds to a plurality of basic information in which the basic content is divided into chunk structures and a plurality of target information in which the target content having attributes different from the basic content related to the basic content is divided into chunk structures. Let me remember. The basic content and the target content include text information, and may further include chart information related to charts. The target content is the content to be edited based on the basic content.

The different attributes referred to here mean, for example, that when the basic content is a specification of the device, the target content is information other than the specification of the device, for example, a manual, an advertisement, or FAQ (Frequently) of the device. Asked Questions) etc. As described above, the target content may be information created based on the basic content. For example, when the basic content is content in a first language such as English, the target content may be content other than the first language such as Japanese. For example, when the basic content is resource data (data such as images, texts, icons, dialogs, menu contents, etc.) referenced by a program executed on the device, the target content is information other than the resource data, for example, Information such as manuals, specifications, messages, and icons related to resource data may be used.

Basic information includes text information. The basic information may further include chart information regarding charts. The basic information may include a basic information label consisting of a character string for identifying the basic information. The basic information is, for example, when the basic content is a specification of a device such as a medical device, the basic information is information divided into chunk structures in which this specification is a mass of data. The basic information is information in which basic contents such as specifications are divided into chunk structures such as sentences, chapters, paragraphs, and pages. Basic information is information used to create the target content, in addition to the specifications divided into chunk structures, for example, incident information, various treatises, information that is the source of the target content, etc. divided into chunk structures. It may be.

Target information includes text information. The target information may further include chart information related to charts. The target information may include a target information label consisting of a character string for identifying the target information. As for the target information, for example, when the basic content is a specification of a device such as a medical device, the manual as the target content created based on this specification is a mass of data in which meaningful information is collected. Information divided into chunk structures. The target information is information divided into chunk structures such as sentences, chapters, paragraphs, and pages of manuals and the like. When the basic information is created in a first language such as English, the target information may be created in a second language such as Japanese, which is different from the first language.

FIG. 4 is a schematic diagram showing an example of a database for calculating the basic information similarity of the information providing system according to the present embodiment. FIG. 5 is a schematic diagram showing an example of a database for calculating the target information similarity of the information providing system according to the present embodiment.

<Database for calculating basic information similarity>
The database for calculating the similarity of basic information is constructed by machine learning using the basic information. As a method of machine learning, for example, basic information is vectorized and learned using a learning program as teacher data. The basic information is stored in the basic information similarity calculation database as a parameter in a vectorized state in correspondence with the basic information label in the basic information. The basic information may be stored in the basic information similarity calculation database as a parameter in a vectorized state in correspondence with the basic information.

<Database for target information similarity estimation processing>
The database for target information similarity estimation processing is constructed by machine learning using the target information. As a method of machine learning, for example, target information is vectorized and learned using a learning program as teacher data. The target information is stored in the target information similarity estimation processing database as a parameter in a vectorized state in correspondence with the target information label in the target information. The target information may be stored in the target information similarity estimation processing database as a parameter in a vectorized state in correspondence with the target information.

<Information providing device 1>
FIG. 6 is a schematic diagram showing an example of the configuration of the information providing device 1 of the information providing system according to the present embodiment. As the information providing device 1, in addition to a personal computer (PC), an electronic device such as a smartphone or a tablet terminal may be used. The information providing device 1 includes a housing 10, a CPU 101, a ROM 102, a RAM 103, a storage unit 104, and I / F 105 to 107. Each configuration 101 to 107 is connected by an internal bus 110.

The CPU (Central Processing Unit) 101 controls the entire information providing device 1. The ROM (Read Only Memory) 102 stores the operation code of the CPU 101. The RAM (Random Access Memory) 103 is a work area used during the operation of the CPU 101. The storage unit 104 stores various information such as basic information, target information, basic information similarity calculation database, and target information similarity calculation database. As the storage unit 104, for example, in addition to an HDD (Hard Disk Drive), an SSD (solid state drive) or the like is used.

The I / F 105 is an interface for transmitting and receiving various information to and from the user terminal 5 and the like via the public communication network 7. The I / F 106 is an interface for transmitting and receiving various information to and from the input portion 108. For example, a keyboard is used as the input portion 108, and a user who uses the information providing system 100 inputs or selects various information or a control command of the information providing device 1 via the input portion 108. The I / F 107 is an interface for transmitting and receiving various information to and from the output portion 109. The output unit 109 outputs various information stored in the storage unit 104, the processing status of the information providing device 1, and the like. A display is used as the output portion 109, and a touch panel type may be used, for example. In this case, the output portion 109 may be configured to include the input portion 108.

FIG. 7 is a schematic diagram showing an example of the function of the information providing device 1 of the information providing system according to the present embodiment. The information providing device 1 includes a basic information acquisition unit 31, a basic information comparison unit 32, a basic information similarity calculation unit 33, a target information extraction unit 34, a target information similarity calculation unit 35, and an input unit 15. It includes an output unit 16, a storage unit 17, and a control unit 18. Each function shown in FIG. 7 is realized by the CPU 101 executing a program stored in the storage unit 104 or the like with the RAM 103 as a work area. Moreover, each function may be controlled by artificial intelligence, for example. Here, the "artificial intelligence" may be based on any well-known artificial intelligence technology.

<Basic information acquisition department 31>
The basic information acquisition unit 31 acquires various types of information such as basic information and specific basic information. The specific basic information is the basic information for which the basic information similarity should be calculated from now on.

<Basic information comparison unit 32>
The basic information comparison unit 32 compares the basic information stored in the relevance database with the specific basic information acquired by the basic information acquisition unit 31. The basic information comparison unit 32 determines whether the basic information and the specific basic information match or do not match.

In the example of FIG. 4, it is assumed that the specific basic information in the basic content X acquired by the basic information acquisition unit 31 includes "basic information x", "basic information a1", and "basic information c1". Then, the basic information comparison unit 32 compares the "basic information x", "basic information a1", and "basic information c1" included in the specific basic information with the basic information stored in the relevance database. It is assumed that "basic information a1" and "basic information c1" are stored in the relevance database, and "basic information x" is not stored. At this time, the basic information comparison unit 32 determines that the "basic information a1" and "basic information c1" included in the specific basic information match the basic information stored in the relevance database database, and the determination ends. Further, the basic information comparison unit 32 determines that the "basic information x" does not match the basic information stored in the relevance database.

<Basic information similarity calculation unit 33>
When the basic information and the specific basic information do not match by the basic information comparison unit 32, the basic information similarity calculation unit 33 refers to the basic information similarity calculation database and stores it in the basic information similarity calculation database. The basic information similarity degree indicating the similarity between the basic information and the specific basic information acquired by the basic information acquisition unit 31 is calculated. The basic information similarity calculation unit 33 calculates the basic information similarity using the feature amount of the basic information. As a feature amount of basic information, for example, basic information may be vectorized and expressed. The basic information similarity calculation unit 33 vectorizes specific basic information and then performs vector calculation with the vectorized basic information in the basic information similarity calculation database to obtain the specific basic information and the basic information. Calculate the basic information similarity.

Note that the basic information similarity calculation unit 33 does not calculate the basic information similarity when the basic information and the specific basic information match by the basic information comparison unit 32.

The basic information similarity indicates the degree of similarity between the specific basic information and the basic information, for example, a decimal number of 100 steps from 0 to 1 such as "0.98", a percentage, 10 steps, or 5 steps. It is shown in three or more stages such as.

In the example of FIG. 4, the "basic information x" included in the specific basic information by the basic information comparison unit 32 does not match the basic information stored in the relevance database. In such a case, the basic information similarity calculation unit 33 refers to the basic information similarity calculation database, and the "basic information x" included in the specific basic information and the "basic" stored in the basic information similarity calculation database. The basic information similarity is calculated for each of "information a1", "basic information b1", "basic information c1", and "basic information b2". For the basic information similarity between the "basic information x" and the "basic information a1", the inner product of the "feature amount q2 of the basic information x" and the "feature amount p1 of the basic information a1" is calculated, for example, "0. It is calculated as "20". Similarly, the basic information similarity between the "basic information x" and the "basic information a1" is "0.98". The basic information similarity between the "basic information x" and the "part information a1" is "0.33". The basic information similarity between the "basic information x" and the "basic information a1" is "0.85". In this case, "basic information x" indicates that it is more similar to "basic information b1" than, for example, "basic information a1".

<Target information extraction unit 34>
The target information extraction unit 34 selects the first basic information from a plurality of basic information based on the calculated basic information similarity, refers to the relevance database, and selects the target information corresponding to the selected first basic information. Extract as the first target information. When one first basic information is selected from a plurality of basic information, the target information extraction unit 34 extracts one target information corresponding to the selected first basic information as the first target information. Further, when the target information extraction unit 34 selects a plurality of first basic information, the target information corresponding to each of the selected first basic information may be extracted as the first target information.

The target information extraction unit 34 may select as the first basic information from the respective basic information labels included in the plurality of basic information based on the calculated basic information similarity. The target information extraction unit 34 may extract the target information corresponding to the basic information label stored in the relevance database as the first target information from the selected basic information label (first basic information). For example, the target information extraction unit 34 selects the basic information label 21, and extracts the target information B1 corresponding to the basic information label 21 stored in the relevance database as the first target information from the selected basic information label 21. You may. Since the basic information label is composed of a character string, the capacity of the basic information similarity calculation database can be reduced as compared with storing the basic information having the text information.

In the example of FIG. 4, as a result of calculating the basic information similarity, the target information extraction unit 34 has the highest basic of "basic information a1", "basic information b1", "basic information c1", and "basic information b2". The "basic information b1" for which the information similarity is calculated is selected as the first basic information. When selecting as the first basic information, a threshold value may be set for the basic information similarity, and the basic information obtained by calculating the basic information similarity equal to or lower than the threshold value may be selected. This threshold value can be appropriately set on the user side.

Then, the target information extraction unit 34 refers to the relevance database and extracts the "target information B1" corresponding to the "basic information b1" selected as the first basic information as the first target information.

Further, the target information extraction unit 34 further extracts one or more second target information different from the first target information from the relevance database based on the target information similarity to be described later.

The target information extraction unit 34 may select one or a plurality of target information labels from the target information labels included in the plurality of target information based on the calculated target information similarity. The target information extraction unit 34 may extract the target information corresponding to the target information label stored in the relevance database from the selected target information label as the second target information. For example, the target information extraction unit 34 selects the target information label 122, and extracts the target information B2 corresponding to the target information label 122 stored in the relevance database as the second target information from the selected basic information label 122. You may. Since the target information label is composed of a character string, the capacity of the target information similarity calculation database can be reduced as compared with storing the target information having the text information.

<Target information similarity calculation unit 35>
The target information similarity calculation unit 35 refers to the target information similarity estimation processing database, and determines the target information similarity indicating the similarity between the target information and the first target information extracted by the target information extraction unit 34. calculate. The target information similarity calculation unit 35 calculates the target information similarity using the feature amount of the target information. As the feature amount of the target information, for example, the target information may be vectorized and expressed. The target information similarity calculation unit 35 vectorizes the specific target information, and then performs a vector operation with the vectorized target information in the target information similarity estimation processing database to obtain the specific target information and the target information. Calculate the similarity of the target information of.

The target information similarity indicates the degree of similarity between the first target information and the target information, for example, a decimal number of 100 steps from 0 to 1 such as "0.95", a percentage, 10 steps, or 5 steps. It is shown in three or more stages such as.

In the example of FIG. 5, the target information similarity calculation unit 35 refers to the target information similarity calculation database, and is similar to the target information “target information B1” extracted as the first target information by the target information extraction unit 34. The degree of similarity of the target information is calculated with "target information A1", "target information B1", "target information C1", and "target information B2" stored in the degree calculation database. For the target information similarity between the "target information B1" and the "target information A1", the inner product of the "feature amount Q1 of the target information B1" and the "feature amount P1 of the target information A1" is calculated, for example, "0. It is calculated as "30". Similarly, the target information similarity between the "target information B1" and the "target information B1" is "1.00". The target information similarity between the "target information B1" and the "target information C1" is "0.20". The target information similarity between the "target information B1" and the "target information B2" is "0.95". In this case, "target information B1" indicates that it is more similar to "target information B2" than, for example, "target information A1".

As described above, the target information extraction unit 34 further extracts one or more second target information different from the first target information based on the degree of similarity of the target information.

In the example of FIG. 5, as a result of calculating the target information similarity, the target information extraction unit 34 determines a predetermined target among "target information A1", "target information B1", "target information C1", and "target information B2". The "target information B2" for which the information similarity is calculated is extracted as the second target information. When selecting the second target information, a threshold value may be set for the target information similarity, and the target information for which the target information similarity equal to or less than the threshold value is calculated may be selected. This threshold value can be appropriately set on the user side. Since the target information for which the target information similarity degree "1.00" is calculated matches the first target information, it may be excluded from being selected as the second target information.

<Input unit 15>
The input unit 15 inputs various information to the information providing device 1. The input unit 15 inputs various information such as learning data, basic information, and basic contents via the I / F 105, and also inputs various information from the input portion 108 via the I / F 106, for example.

<Output unit 16>
The output unit 16 outputs various information such as target information to the output unit 109 or the like. The output unit 16 transmits various information such as target information to the user terminal 5 or the like via, for example, the public communication network 7.

<Memory unit 17>
The storage unit 17 stores various information such as basic information and target information in the storage unit 104, and retrieves various information stored in the storage unit 104 as needed. Further, the storage unit 17 stores various databases such as a database for calculating the basic information similarity and a database for calculating the target information similarity in the storage unit 104, and retrieves various databases stored in the storage unit 104 as needed. ..

<Control unit 18>
The control unit 18 executes machine learning for constructing a database for calculating basic information similarity using a plurality of basic information. Further, the control unit 18 executes machine learning for constructing a database for calculating the similarity of target information using a plurality of target information. The control unit 18 executes machine learning by linear regression, logistic regression, support vector machine, decision tree, regression tree, random forest, gradient boosting tree, neural network, bays, time series, clustering, ensemble learning, and the like.

<User terminal 5>
The user terminal 5 indicates a terminal owned by the user. As the user terminal 5, in addition to electronic devices such as mobile phones (mobile terminals), smartphones, tablet terminals, wearable terminals, personal computers, and IoT (Internet of Things) devices, those embodied in all kinds of electronic devices are used. You may. As the user terminal 5, a holo lens (registered trademark), which is one type of HMD (head-mounted display), may be used. The user terminal 5 may be connected to the information providing device 1 via, for example, the public communication network 7, or may be directly connected to, for example, the information providing device 1. The user can acquire the first target information from the information providing device 1 by using the user terminal 5, and can display various acquired information on the display unit of the user terminal 5. Further, the user may use the user terminal 5 to perform various controls of the information providing device 1.

<Server 6>
The server 6 stores the above-mentioned various information. Various information sent via, for example, the public communication network 7 is stored in the server 6. For example, the server 6 stores the same information as the storage unit 104, and may send and receive various information to and from the information providing device 1 via the public communication network 7. That is, the information providing device 1 may use the server 6 instead of the storage unit 104.

<Public communication network 7>
The public communication network 7 is an Internet network or the like to which the information providing device 1 and the like are connected via a communication circuit. The public communication network 7 may be composed of a so-called optical fiber communication network. Further, the public communication network 7 is not limited to the wired communication network, and may be realized by a known communication network such as a wireless communication network.

(First example of operation of information providing system 100)
Next, an example of the operation of the information providing system 100 in this embodiment will be described. FIG. 8 is a flowchart showing an example of the operation of the information providing system 100 in the present embodiment.

<Basic information acquisition step S31>
The basic information acquisition unit 31 acquires one or more basic information in which basic contents such as specifications are divided into chunk structures as specific basic information (basic information acquisition step S31). The basic information acquisition unit 31 may acquire specific basic contents including specific basic information.

<Basic information comparison step S32>
Next, the basic information comparison unit 32 compares the basic information stored in the relevance database with the specific basic information acquired by the basic information acquisition unit 31 (basic information comparison step S32). The basic information comparison unit 32 determines whether the basic information and the specific basic information match or do not match.

<Basic information similarity calculation step S33>
Next, when the basic information similarity calculation unit 33 does not match the basic information and the specific basic information as a result of comparison by the basic information comparison unit 32, the basic information similarity calculation unit 33 refers to the basic information similarity calculation database and the basic information similarity degree. The basic information similarity degree indicating the similarity between the basic information stored in the calculation database and the specific basic information acquired by the basic information acquisition unit 31 is calculated (basic information similarity calculation step S33).

<First target information extraction step S34>
The target information extraction unit 34 selects the first basic information from a plurality of basic information based on the calculated basic information similarity, refers to the relevance database, and selects the target information corresponding to the selected first basic information. It is extracted as the first target information (first target information extraction step S34).

<Target information similarity calculation step S35>
Next, the target information similarity calculation unit 35 refers to the target information similarity estimation processing database, and the target information stored in the target information similarity estimation processing database and the target information extracted by the target information extraction unit 34. 1 The target information similarity indicating the similarity with the target information is calculated (target information similarity calculation step S35).

<Second target information extraction step S36>
Next, the target information extraction unit 34 further extracts one or a plurality of second target information different from the first target information based on the target information similarity (second target information extraction step S36).

This completes an example of the operation of the information providing system 100.

According to the present embodiment, a plurality of basic information in which the basic content is divided into chunk structures and a plurality of target information in which target content having an attribute different from the basic content is divided into chunk structures are stored in correspondence with each other. Relevance database, basic information similarity calculation database constructed by machine learning using a plurality of basic information, basic information acquisition unit 31 for acquiring specific basic information, basic information, and specific basic information. When the basic information and the specific basic information do not match by the basic information comparison unit 32 and the basic information comparison unit 32, the basic information similarity calculation database is referred to, and the basic information and the specific basic information are displayed. The basic information similarity calculation unit 33 that calculates the basic information similarity indicating the similarity of, and the first basic information are selected from a plurality of basic information based on the basic information similarity, and the relevance database is referred to. It includes a target information extraction unit 34 that extracts target information corresponding to the first basic information as the first target information.

According to the present embodiment, the basic information similarity calculation unit 33 calculates the basic information similarity for specific basic information that does not match the basic information stored in the relevance database by the basic information comparison unit 32. That is, it is not necessary to calculate the basic information similarity for the specific basic information that matches the basic information stored in the relevance database by the basic information comparison unit 32. Therefore, the calculation of the basic information similarity can be performed more efficiently.

In particular, according to the present embodiment, the first basic information is selected from a plurality of basic information based on the similarity of the basic information, the relevance database is referred to, and the target information corresponding to the first basic information is the first target. Extract as information. Thereby, by selecting the first basic information similar to the specific basic information based on the quantitatively evaluated basic information similarity, the accuracy of selecting the first basic information can be improved.

In particular, according to the present embodiment, the relevance database is referred to, and the target information corresponding to the first basic information is extracted as the first target information. When new information is included in the specific basic information or when there is a change, the part extracted as the first target information corresponds to the edited part in the target content such as the manual. Therefore, when editing the target content such as a manual, it is only necessary to edit the target information extracted as the first target information, and the editing work of the target content can be performed in a short time.

For example, if a device is upgraded from an old version to a new version and a part of the past specifications is changed to become a new specification, the past of the product created based on the past specifications The manual also needs to be created in a new manual. At this time, in the past, it was difficult to grasp the edited part in the past manual from the new specification. As described above, according to the present embodiment, the past specifications that are candidates for editing are selected from the new specifications as the basic contents, and the past manuals corresponding to the past specifications are the new specifications. It can be grasped that it is the target content to be edited by the book. At this time, a new specification as specific basic content acquired by the basic information acquisition unit 31, a past specification which is a set of a plurality of basic information stored in the relevance database, and a set of a plurality of target information A past manual is divided into chunk structures. Therefore, it is possible to efficiently extract only the parts changed by the new specifications from the past manuals. Therefore, based on the new specifications, the user can easily grasp the edited part in the past manual. Therefore, for example, when creating a new manual, the past manual can be used as it is for the parts that are not changed in the specifications, and only the parts that are changed in the new specifications can be newly created. So to speak, it is only necessary to edit the difference in the changed part in the specifications. Therefore, the manual editing work can be easily performed.

Further, according to the present embodiment, the target information similarity estimation processing database constructed by machine learning using a plurality of target information and the target information similarity estimation processing database are referred to, and the target information and the first. The target information similarity calculation unit 35 for calculating the similarity between the target information and the target information, and the target information extraction unit 34 is different from the first target information based on the target information similarity. The second target information is further extracted.

According to this embodiment, the second target information different from the first target information is further extracted based on the similarity of the target information. As a result, the accuracy of selecting the second target information can be improved by selecting the second target information similar to the first target information based on the quantitatively evaluated degree of similarity of the target information. Therefore, when new information is included in the specific basic information or when there is a change, the second target information similar to the first target information is also extracted, so that any part of the target information in which the target content is divided is extracted. The user can immediately grasp whether or not the above applies. Therefore, when editing the target content, it is only necessary to edit the first target information and the target information extracted as the second target information, and the editing work of the target content can be performed in a short time.

That is, if a device has multiple versions and some of the past specifications are changed to new specifications, the product created based on the multiple past specifications. Each past manual also needs to be created in a new manual. According to the present embodiment, a past specification that is a candidate to be changed is selected from the new specifications, and a past manual corresponding to this past specification and another past manual similar to the past manual are selected. However, it can be understood that changes are necessary due to the new specifications. At this time, the new specifications, the past specifications, and the past manuals are each divided into chunk structures. Therefore, it is possible to efficiently extract only the parts changed by the new specifications from the past manuals. At this time, a plurality of similar past manuals can be extracted as targets. Therefore, the user can easily and simultaneously grasp the relevant parts of a plurality of past manuals that should be based on the new specifications. Therefore, for example, when creating a new manual, the past manual can be used as it is for the parts that are not changed in the specifications, and only the parts that are changed in the specifications can be newly created. So to speak, it is only necessary to edit the difference in the changed part in the specifications. Therefore, the manual editing work can be easily performed.

According to this embodiment, the basic information acquisition step S31 is performed after the target information selection step S14. As a result, the user can compare the first target information selected by the target information selection unit 14 and the first target information and the second target information extracted by the target information extraction unit 34. Therefore, in the first target information such as a manual, the relevant part to be edited can be immediately grasped.

<Second example of information providing device 1>
FIG. 9 is a schematic view showing a second example of the information providing system 100 according to the present embodiment. In the second example of the information providing system 100, a plurality of basic information in which the basic content is divided into chunk structures and a plurality of target information in which the target content classified by the attribute is divided into chunk units are stored. It has multiple relevance databases.

In the example of FIG. 9, the information providing system 100 includes a first relevance database, a second relevance database, a third relevance database, and a fourth relevance database. For example, when the basic content is a specification of the device, a plurality of target information regarding the target content as a manual of the device is stored in the first relevance database. The second relevance database stores a plurality of target information regarding the target content as an advertisement of the device. A plurality of target information regarding the target content as FAQ of the device is stored in the third relevance database. The fourth relevance database stores a plurality of target information regarding the target content as resource data (data such as images, texts, icons, dialogs, menu contents, etc.) referred to by the program executed by the device.

The operation of the information providing system 100 according to the second example will be described.

<Basic information acquisition step S31>
First, the basic information acquisition unit 31 includes specific basic information x1 to x5, specific basic information e1, specific basic information f2, specific basic information g3, and specific basic information h2, which are included in the specific basic content X. (Basic information acquisition step S31).

<Basic information comparison step S32>
The basic information comparison unit 32 compares the basic information stored in the relevance database with the specific basic information acquired by the basic information acquisition unit 31. The basic information comparison unit 32 determines whether the basic information and the specific basic information match or do not match.

In the example of FIG. 9, the basic information comparison unit 32 includes "basic information x1" to "basic information x5", "basic information e1", "basic information f2", "basic information g3", which are included in the specific basic information. The "basic information h2" is compared with the basic information stored in the relevance database. "Basic information e1" is stored in the first relevance database, "basic information f2" is stored in the second relevance database, and "basic information g3" is stored in the third relevance database. , "Basic information h2" is stored in the fourth relevance database, and "basic information x1" to "basic information x5" are not stored in any of these relevance databases. At this time, the basic information comparison unit 32 stores the "basic information e1", "basic information f2", "basic information g3", and "basic information h2" included in the specific basic information in the relevance database. Is determined to match. Further, the basic information comparison unit 32 determines that the "basic information x1" to "basic information x5" do not match the basic information stored in the relevance database.

<Basic information similarity calculation step S33>
When the basic information and the specific basic information do not match by the basic information comparison unit 32, the basic information similarity calculation unit 33 refers to the basic information similarity calculation database and stores it in the basic information similarity calculation database. The basic information similarity degree indicating the similarity between the basic information and the specific basic information acquired by the basic information acquisition unit 31 is calculated.

In the example of FIG. 9, the "basic information x1" to "basic information x5" included in the specific basic information by the basic information comparison unit 32 do not match the basic information stored in the respective relevance databases. In such a case, the basic information similarity calculation unit 33 refers to the basic information similarity calculation database, and refers to the basic information similarity calculation database for "basic information x1" to "basic information x5" included in the specific basic information. Calculate the degree of similarity of basic information with each basic information stored in.

<First target information extraction step S34>
The target information extraction unit 34 selects the first basic information from a plurality of basic information based on the calculated basic information similarity, refers to each relevance database, and corresponds to the selected first basic information. Information is extracted as the first target information. When one first basic information is selected from a plurality of basic information, the target information extraction unit 34 extracts one target information corresponding to the selected first basic information as the first target information. Further, when the target information extraction unit 34 selects a plurality of first basic information, the target information corresponding to each of the selected first basic information may be extracted as the first target information.

In the example of FIG. 9, as a result of calculating the basic information similarity, the target information extraction unit 34 uses the "basic information e1" for which the highest basic information similarity is calculated as the first basic information for the "basic information x1". select. When selecting as the first basic information, a threshold value may be set for the basic information similarity, and the basic information obtained by calculating the basic information similarity equal to or lower than the threshold value may be selected. This threshold value can be appropriately set on the user side. Similarly, the target information extraction unit 34 has "basic information f1" for "basic information x2", "basic information f3" for "basic information x3", and "basic information g2" for "basic information x4". , And for "basic information x5", "basic information h2" is selected as the first basic information.

Then, the target information extraction unit 34 refers to the first relevance database and extracts the "target information E1" corresponding to the "basic information e1" selected as the first basic information as the first target information. Similarly, the target information extraction unit 34 refers to the second relevance database and selects the "target information F1" corresponding to the "basic information f1" selected as the first basic information as the first basic information "basic". The "target information F3" corresponding to the "information f3" is extracted as the first target information, respectively. The target information extraction unit 34 refers to the third relevance database and extracts the "target information G2" corresponding to the "basic information g2" selected as the first basic information as the first target information. The target information extraction unit 34 refers to the fourth relevance database and extracts the "target information H2" corresponding to the "basic information h2" selected as the first basic information as the first target information.

The target information extraction unit 34 may select one or a plurality of target information labels from the target information labels included in the plurality of target information based on the calculated target information similarity.

According to the present embodiment, a relevance database that stores a plurality of basic information in which the basic content is divided into chunk structures and a plurality of target information in which the target content classified by attribute is divided into chunk units. It is provided with a plurality of.

Thereby, according to the present embodiment, the target information corresponding to the first basic information is extracted as the first target information by referring to each relevance database classified for each attribute. Therefore, when new information is included in the specific basic information or when there is a change, the part extracted as the first target information corresponds to the edited part in the target content such as the manual. Can easily grasp the edited part in the target content. Therefore, when editing the target content such as a manual, it is only necessary to edit the target information extracted as the first target information, and the editing work of the target content can be performed in a short time.

In particular, when editing each target content as a manual, advertisement, or FAQ from the specifications acquired as basic content, the part extracted as the first target information is edited in each target content as manual, advertisement, or FAQ. It will correspond to the place. Therefore, the user can easily grasp the edited part in each target content for each attribute such as the manual, the advertisement, and the FAQ.

<Third example of information providing device 1>
The third example of the information providing device 1 is different from the first example in that it further includes an access control unit. The access control unit is realized, for example, by the CPU 101 executing a program stored in the storage unit 104 or the like with the RAM 103 as a work area.

The access control unit controls access to various databases and target contents. Access includes full access, read and write access, comment-only access, read-only access, and line-of-sight access prohibition. The access control unit is performed based on the access control information. The access control information includes a user name and the access assigned to each user name. The access control information is stored in, for example, the storage unit 104.

When a user is assigned full access, the user has full read and write access to various databases and target content, and the user can use any aspect of the user interface. For example, for full access, the user can change the database configuration. If the user has read and write access, the user has read and write to the target content, but cannot change the database configuration. In the case of comment-only access, the user can insert comments into the target content, but cannot change various databases or target content. With read-only access, the user can view the target content, but cannot make changes to the various databases or the target content, and cannot insert any comments.

For example, suppose that a new target content is generated based on specific basic information and the generated new target content is edited. At this time, according to the present embodiment, an access control unit is further provided. Thereby, a specific one or a plurality of users among the plurality of users can perform a predetermined access based on the access control information. That is, for multiple users who use various databases and target contents, the control of the editing type such as read-only and full access is linked with the authority based on the user's attributes, and for each of various databases and target contents. Can be managed. In particular, it is possible to prevent unintended editing by allowing only authorized users to edit such as writing while making only browsing accessible at the same time.

Although the embodiment of the present invention has been described, the embodiment is presented as an example and is not intended to limit the scope of the invention. These novel embodiments can be implemented in various other embodiments, and various omissions, replacements, and changes can be made without departing from the gist of the invention. These embodiments and modifications thereof are included in the scope and gist of the invention, and are also included in the scope of the invention described in the claims and the equivalent scope thereof.

1: Information providing device 5: User terminal 6: Server 7: Public communication network 10: Housing 15: Input unit 16: Output unit 17: Storage unit 18: Control unit 31: Basic information acquisition unit 32: Basic information comparison unit 33 : Basic information similarity calculation unit 34: Target information extraction unit 35: Target information similarity calculation unit 100: Information providing system 101: CPU
102: ROM
103: RAM
104: Preservation unit 105: I / F
106: I / F
107: I / F
108: Input part 109: Output part 110: Internal bus S31: Basic information acquisition unit Step S32: Basic information comparison step S33: Basic information similarity calculation step S34: First target information extraction step S35: Target information similarity calculation step S36 : Second target information extraction step

Claims

A relevance database in which a plurality of basic information in which the basic content is divided into chunk structures and a plurality of target information in which target content having an attribute different from the basic content is divided into chunk structures are stored in association with each other.
A database for calculating basic information similarity constructed by machine learning using a plurality of the basic information,
Basic information acquisition means to acquire specific basic information,
A basic information comparison means for comparing the basic information with the specific basic information,
When the basic information and the specific basic information do not match by the basic information comparison means, the basic information similarity calculation database is referred to, and the similarity between the basic information and the specific basic information is shown. Basic information similarity calculation means for calculating basic information similarity,
Based on the basic information similarity, the first basic information is selected from the plurality of basic information, the relevance database is referred to, and the target information corresponding to the first basic information is extracted as the first target information. An information providing system characterized by having a target information extraction means.
A database for target information similarity estimation processing constructed by machine learning using a plurality of the target information,
Target information similarity that refers to the target information similarity estimation processing database and calculates the target information similarity that indicates the similarity between the target information and the first target information extracted by the target information extraction means. With a calculation means,
The information providing system according to claim 1, wherein the target information extracting means further extracts a second target information different from the first target information based on the target information similarity.
A claim characterized by including a plurality of the relevance databases stored in association with a plurality of the basic information and a plurality of the target information in which the target content classified by an attribute is divided into chunk structures. The information providing system described in Item 2.