Inspection device, inspection method, and program

Abstract

To solve the problem in which sentence inspection based on inspection rules per user could not be conducted conventionally.SOLUTION: An inspection device 1 comprises: an inspection information acceptance unit 121 which accepts inspection information containing a sentence in association with a user identifier; a rule acquisition unit 131 which acquires one or more inspection rules corresponding to a user identifier corresponding to inspection information from a rule storage unit 112 which stores one or more inspection rules for sentence inspection in association with user identifiers per user identifier; an inspection unit 132 which inspects a sentence using one or more inspection rules for the inspection information; an inspection result configuration unit 133 which configures output inspection results using the inspection results in the inspection unit 132; and a result output unit 141 which outputs the inspection results. The inspection device can inspect sentences based on the inspection rules per user.SELECTED DRAWING: Figure 2

Description

[Technical Field]

[0001] The present invention relates to an inspection device or the like that checks documents or the like. [Background technology]

[0002] Conventionally, editors for creating documents have had a proofreading function (see, for example, Non-Patent Document 1). [Prior art documents] [Non-patent literature]

[0003] [Non-Patent Document 1] Microsoft (registered trademark), "Check spelling and grammar in Office", [online], [searched April 28, 2021], Internet [URL: https: / / support.microsoft.com / ja-jp / topic / office-%E3%81%A7%E3%82%B9%E3%83%9A%E3%83%AB-%E3%83%81%E3%82%A7%E3%83%83%E3%82%AF%E3%81%A8%E6%96%87%E7%AB%A0%E6%A0%A1%E6%AD%A3%E3%82%92%E8%A1%8C%E3%81%86-5cdeced7-d81d-47de-9096-efd0ee909227] Summary of the Invention [Problem to be solved by the invention]

[0004] However, in the prior art, it was not possible to inspect text based on appropriate inspection rules. [Means for solving the problem]

[0005] The inspection device of the first invention is an inspection device comprising an inspection information receiving unit that receives inspection information including a sentence in association with a user identifier, a rule acquisition unit that acquires one or more inspection rules corresponding to a user identifier corresponding to the inspection information from a rule storage unit in which one or more inspection rules for inspecting sentences are stored in association with each user identifier, an inspection unit that performs an inspection on the inspection information using each of the one or more inspection rules, an inspection result construction unit that constructs an inspection result to be output using the inspection results in the inspection unit, and a result output unit that outputs the inspection result.

[0006] With this configuration, it is possible to inspect text based on inspection rules for each user.

[0007] Furthermore, the inspection device of the second invention differs from the first invention in that the inspection rules have a correct term dictionary containing one or more correct terms for detecting typos, the inspection unit comprises: morphological analysis means for performing morphological analysis on sentences contained in the inspection information and acquiring two or more morphemes; comparison string acquisition means for acquiring two or more comparison strings, which are strings of one or two or more consecutive morphemes from the two or more morphemes acquired by the morphological analysis means; difference degree information acquisition means for acquiring difference degree information that specifies the degree of difference between each of the two or more comparison strings and one or more correct terms; and judgment means for determining that the corresponding comparison string is a typo if the degree of difference specified by the difference degree information is not 0 but is within a threshold or less than the threshold, and the inspection result construction unit constructs an inspection result that specifies the comparison strings that have been judged to be typos.

[0008] With this configuration, it is possible to check for errors in the correct terms registered by the user.

[0009] In addition, the inspection device of the third invention is an inspection device in which, compared to the second invention, the difference degree information acquisition means acquires the number of characters in the comparison string or the correct term to be compared with the comparison string, and acquires difference degree information according to the number of characters.

[0010] With this configuration, it is possible to appropriately check for errors in the correct terms registered by the user.

[0011] Furthermore, in comparison with the first invention, the inspection device of the fourth invention further comprises a learning device storage unit in which a learning device obtained by performing a machine learning learning process on two or more pieces of training data having positive example sentences, which are correct sentences, and two or more negative example sentences, which are sentences containing errors generated from the positive example sentences using two or more error patterns, is stored; the rule acquisition unit acquires the learning device; the inspection unit performs a machine learning prediction process using the learning device and one or more sentences contained in the inspection information, and comprises prediction means for identifying locations of typos or omissions for each of the one or more sentences; and the inspection result composition unit composes an inspection result having location information identifying locations of typos or omissions.

[0012] This configuration allows for proper checking of typographical errors and omissions.

[0013] Furthermore, the testing device of the fifth invention is an testing device in which, compared to the fourth invention, the teacher data has a pattern identifier that identifies an error pattern, the prediction means performs a machine learning prediction process using one or more sentences contained in the testing information and a learning device, and also obtains a pattern identifier for each of the one or more sentences, and the testing result composition unit composes testing results that also have a type identifier corresponding to the pattern identifier.

[0014] With this configuration, the type of error can also be known.

[0015] Furthermore, the inspection device of the sixth invention is an inspection device in which, compared to any one of the first to fifth inventions, the inspection rule includes a positive example image which is a correct image, the inspection unit acquires one or more images contained in the inspection information, acquires image difference information regarding the differences between each of the one or more images and the positive example image, and is equipped with an image judgment means for judging whether the information regarding the differences identified by the image difference information satisfies a pseudo-image condition for detecting a pseudo-image relative to the positive example image, and the inspection result construction unit constructs an inspection result indicating that the image is an error if the image contained in the inspection information satisfies the pseudo-image condition.

[0016] This configuration also makes it possible to check for errors in the pseudo image.

[0017] Furthermore, the inspection device of the seventh invention is an inspection device that, compared to any one of the first to sixth inventions, comprises a selection interface configuration unit that configures a selection interface in which the inspection results have location information that specifies the location of an error in the inspection information and a type identifier that identifies the type of error, and which has one or more type identifiers resulting from unique processing of the type identifiers that the inspection results have as selection items, and which does not have type identifiers that the inspection results do not have as selection items, a selection interface output unit that outputs the selection interface, and a selection receiving unit that receives selections for the selection items that the selection interface has, and the inspection result configuration unit compares errors corresponding to one or more location information that are paired with the type identifier corresponding to the selection item for the selection with errors corresponding to other location information to configure visually different inspection results.

[0018] With this configuration, the inspection results can be viewed with high operability using a selection interface according to the type of error.

[0019] Furthermore, the inspection device of the eighth invention is an inspection device in which, compared to any one of the first to sixth inventions, the inspection result has location information that specifies the location of an error in the inspection information and a type identifier that identifies the type of error, and the inspection result composition unit composes the inspection result in a manner that visually indicates the location information and the type identifier.

[0020] With this configuration, the location and type of the error can also be displayed.

[0021] Furthermore, the inspection device of the ninth invention is an inspection device in which, with respect to any one of the first to eighth inventions, the one or more inspection rules include an inspection rule that checks whether two or more terms are placed in a position that satisfies a positional condition.

[0022] This configuration makes it possible to check for errors of inappropriate expressions resulting from the combination of two or more terms.

[0023] Furthermore, the learning device of the tenth invention is a learning device comprising: a pattern information storage unit that stores one or more pattern information related to error patterns; a positive example sentence acquisition unit that acquires one or more positive example sentences that are correct sentences; a negative example sentence acquisition unit that acquires one or more negative example sentences that are sentences containing errors using one or more pattern information for each of the one or more positive example sentences; a learning unit that performs machine learning learning processing on two or more teacher data sets that have positive example sentences and one or more negative example sentences, and acquires a learning device; and a storage unit that stores the learning device.

[0024] With this configuration, a learning device can be configured that checks for typos, omissions, etc. [Effects of the Invention]

[0025] The inspection device according to the present invention can inspect text based on appropriate inspection rules. [Brief explanation of the drawings]

[0026] [Figure 1] Conceptual diagram of information system A in embodiment 1 [Figure 2] Block diagram of Information System A [Figure 3] A flowchart illustrating an example of the operation of the inspection device 1 [Figure 4] A flowchart illustrating an example of the inspection process [Figure 5] Flowchart illustrating an example of a same sentence check process [Figure 6] Flowchart illustrating an example of a same sentence check process [Figure 7] Flowchart illustrating an example of a same sentence check process [Figure 8] A flowchart illustrating an example of the image inspection process [Figure 9] A flowchart illustrating an example of the test result configuration process. [Figure 10] A flowchart illustrating an example of the selection IF configuration process [Figure 11] A flowchart illustrating an example of the selected test result configuration process [Figure 12] A flowchart illustrating an example of the operation of the terminal device 2 [Figure 13] A diagram showing the rule management table [Figure 14] Figure showing an example of the output [Figure 15] Figure showing an example of the output [Figure 16] A diagram showing an example of the same examination information [Figure 17] Figure showing an example of the output [Figure 18] Block diagram of a learning device 3 according to a second embodiment [Figure 19] A flowchart illustrating an example of the operation of the learning device 3. [Figure 20] A flowchart illustrating an example of the teacher data acquisition process [Figure 21] Figure showing the teacher data management table [Figure 22] Figure showing an example of the test results [Figure 23] Overview of the computer system in the above embodiment [Figure 24] Block diagram of the computer system DETAILED DESCRIPTION OF THE INVENTION

[0027] Hereinafter, embodiments of an inspection device and the like will be described with reference to the drawings. Note that components with the same reference numerals in the embodiments perform similar operations, and therefore repeated description may be omitted.

[0028] (Embodiment 1) In this embodiment, an inspection device will be described that manages inspection rules for documents and the like in association with users and inspects documents and the like based on the inspection rules corresponding to the users.

[0029] In this embodiment, a checking device for detecting errors in registered correct terms will be described.

[0030] In this embodiment, an inspection device for detecting typographical errors and omissions will be described.

[0031] In this embodiment, an inspection device that also acquires and outputs the type of error will be described.

[0032] In this embodiment, an inspection device that detects a pseudo image for a registered image will be described.

[0033] In this embodiment, an inspection device will be described that configures and outputs a selection interface such as a menu based on the type of error.

[0034] Furthermore, in this embodiment, an inspection device that outputs inspection results in a manner that allows the type of error to be visually grasped will be described.

[0035] 1 is a conceptual diagram of an information system A according to this embodiment. The information system A includes an inspection device 1 and one or more terminal devices 2.

[0036] The inspection device 1 is a device that inspects inspection information including text. The inspection device 1 is usually a so-called server, such as a cloud server or an ASP server. The type of the inspection device 1 is not important. The inspection device 1 may also operate standalone.

[0037] The terminal device 2 is a terminal used by a user. The user is, for example, a user who obtains test results of test information. The terminal device 2 is, for example, a so-called personal computer, a tablet terminal, a smartphone, or the like, and the type does not matter.

[0038] 2 is a block diagram of an information system A according to the present embodiment. The inspection device 1 includes a storage unit 11, a reception unit 12, a processing unit 13, and an output unit 14. The storage unit 11 includes a user information storage unit 111, a rule storage unit 112, and a learning device storage unit 113. The reception unit 12 includes an inspection information reception unit 121 and a selection reception unit 122. The processing unit 13 includes a rule acquisition unit 131, an inspection unit 132, an inspection result configuration unit 133, and a selection interface configuration unit 134. The inspection unit 132 includes a morphological analysis unit 1321, a comparison string acquisition unit 1322, a difference degree information acquisition unit 1323, a judgment unit 1324, a prediction unit 1325, and an image judgment unit 1326. The output unit 14 includes a result output unit 141 and a selection interface output unit 142.

[0039] The terminal device 2 includes a terminal storage unit 21, a terminal reception unit 22, a terminal processing unit 23, a terminal transmission unit 24, a terminal reception unit 25, and a terminal output unit 26.

[0040] Various types of information are stored in the storage unit 11 that constitutes the inspection device 1. The various types of information include, for example, user information (to be described later), inspection rules (to be described later), a learning device (to be described later), and positive example images.

[0041] A positive example image is a correct image. A positive example image is usually an image registered by a user. A positive example image is, for example, a company's graphic trademark, image trademark, or logo.

[0042] The user information storage unit 111 stores one or more pieces of user information. The user information is information about a user who uses the inspection device 1. The user information usually has a user identifier. The user identifier is information that identifies a user, such as an ID, email address, telephone number, or name. The user may also be an organization. The organization may be, for example, a company or a local government. If the user is an organization, the user identifier may be an organization identifier.

[0043] The rule storage unit 112 stores one or more inspection rules for each user identifier. An inspection rule is a rule for inspecting an inspection target. An inspection target is an object to be inspected for errors. An inspection target includes a text. A text is made up of one or more sentences. An inspection target may also include an image.

[0044] Inspection rules are information for detecting errors in the inspection target. Inspection rules may be inspection programs. Examples of inspection rules include (1) misspelled term rules, (2) misspelled or omitted words rules, (3) inappropriate term set rules, and (4) pseudo-image rules. (1) Rules for mistyped terms

[0045] The misspelled term rule is information for detecting an incorrect term from a correct term. The misspelled term rule has a correct term dictionary containing one or more correct terms. A correct term is, for example, a term registered by a user. A correct term is a single word or a string of two or more consecutive words. The misspelled term rule includes, for example, a condition regarding the distance between a correct term and an incorrect term. The distance condition is, for example, "0<distance<=threshold X" or "0<distance<threshold X." Furthermore, the distance is, for example, the edit distance, but it may also be the Levenshtein distance, the minimum edit distance, or the like. (2) Rules for typos and omissions

[0046] The typographical error and omission rules are information for detecting typos, omissions, etc. The typographical error and omission rules are information based on positive example sentences and negative example sentences, and are, for example, a set of sets of a learning device, correct terms, and typographical error and omission terms acquired by the learning device 3 described later. A typographical error and omission term is a term that has a typo or omission in comparison to a correct term. (3) Inappropriate term set rules

[0047] An inappropriate term set rule is information for checking whether two or more terms are positioned in a position that satisfies a position condition. A position condition is a condition regarding the relative positions of two or more terms. For example, an inappropriate term set rule is, "Term 1 and Term 2 must be within N words (N is a natural number)" or "Term 2 must be within N words (N is a natural number) after Term 1." For example, an inappropriate term set rule is, "'slim' must be within five words after 'absolute'."

[0048] The inappropriate term set rule can detect sentences that contain inappropriate expressions for sentences created by the user, such as exaggerated expressions and expressions that may cause misunderstandings. (4) Pseudo-image rules

[0049] A pseudo image rule includes pseudo image conditions for determining whether an image being inspected is a pseudo image compared to a positive example image. A pseudo image rule includes one or more positive example images. A positive example image is an image that the user considers correct. Examples of positive example images include a company logo or a graphic trademark owned by a company. A pseudo image is an image that resembles a positive example image. An image that is completely different from a positive example image does not satisfy the pseudo image conditions.

[0050] The pseudo image conditions are, for example, conditions regarding the similarity between the image to be inspected and the positive example image, such as "threshold Y<similarity<1" or "threshold Y<=similarity<1" (if the similarity is 1, they match).

[0051] The pseudo-image conditions are, for example, conditions related to one or more feature quantities of an image. The feature quantities are, for example, feature quantities related to shape, feature quantities related to color, and feature quantities related to composition. The feature quantities related to shape are, for example, aspect ratio. The feature quantities related to color are, for example, whether the image is monochrome or color, and the proportion of each color in the composition. The feature quantities related to composition are, for example, the presence or absence of a character string, or the presence or absence of a specific shape (for example, a triangle or a rectangle).

[0052] It is preferable that the pseudo-image conditions include both conditions regarding the similarity and conditions regarding the feature amount.

[0053] The learning device storage unit 113 stores one or more learning devices. A learning device may also be stored for each user. In other words, a learning device may be associated with a user identifier.

[0054] The learning device is information acquired by the learning device 3, which will be described later. The learning device may also be called a learning model, model, or the like. The learning device here is information acquired by performing a machine learning learning process on two or more pieces of training data each having a positive example sentence and one or more negative example sentences corresponding to the positive example sentence. The positive example sentence is a correct sentence. The negative example sentence is a sentence that includes an error in part of the positive example sentence. It is also preferable that the training data have a pattern identifier that identifies an error pattern. The pattern identifier may be a type identifier.

[0055] The reception unit 12 receives various instructions and information. The various instructions and information include, for example, test information and selection. Here, reception usually refers to reception from the terminal device 2, but the concept also includes reception of information input from an input device such as a keyboard, mouse, or touch panel, and reception of information read from a recording medium such as an optical disk, magnetic disk, or semiconductor memory.

[0056] The examination information receiving unit 121 receives the examination information. The examination information receiving unit 121 normally receives the examination information from the terminal device 2. The examination information is normally associated with a user identifier.

[0057] The selection receiving unit 122 receives a selection for a selection item that a selection interface (hereinafter referred to as "selection IF" as appropriate) has. The selection receiving unit 122 typically receives a selection from the terminal device 2. The selection has an item identifier of the selected item. The item identifier may be any information that identifies the selected item.

[0058] The processing unit 13 performs various types of processing. The various types of processing are performed by a rule acquisition unit 131, an inspection unit 132, an inspection result configuration unit 133, and a selection interface configuration unit .

[0059] The rule acquisition unit 131 acquires, from the rule storage unit 112, one or more inspection rules corresponding to the user identifier corresponding to the inspection information received by the inspection information reception unit 121.

[0060] The rule acquiring unit 131 acquires, for example, a learning device from the learning device storage unit 113. Note that the learning device may also be used as information constituting the inspection rule.

[0061] The inspection unit 132 inspects the inspection information received by the inspection information receiving unit 121 using one or more inspection rules acquired by the rule acquiring unit 131 .

[0062] The inspection unit 132 preferably acquires inspection results including error information related to the error. The error information includes location information and a type identifier. The location information is information that identifies the location of the error. The location information is, for example, an offset within the inspection target, a start address and an end address, a character, a character string, an image ID, etc. The type identifier is information that identifies the type of error.

[0063] The inspection unit 132 detects inappropriate expressions using an inspection rule that inspects whether two or more terms are arranged in positions that satisfy a positional condition, for example.

[0064] The morphological analysis means 1321 performs morphological analysis on the sentence included in the test information to obtain two or more morphemes. The morphological analysis means 1321 typically performs morphological analysis on one or more sentences included in the sentence included in the test information to obtain two or more morphemes. The morphological analysis technique is a well-known technique, and therefore a detailed description thereof will be omitted.

[0065] The comparison string acquisition means 1322 acquires two or more comparison strings from the two or more morphemes acquired by the morphological analysis means 1321. A comparison string is a single morpheme or a string of two or more consecutive morphemes.

[0066] When the morphological analysis means 1321 divides the sentence "high customizability achieved" into five morphemes, "high | customizability | | achieved |," the comparison string acquisition means 1322 acquires, for example, the comparison strings "high," "high customization," "high customizability," "high customizability achieved," "customization," "customizability," "customization," "achieve customizability," "quality," "quality," "achieve quality," "achieve," "achieve," "realize," "realize," "realize," "realize."

[0067] The difference degree information acquisition means 1323 acquires difference degree information between the comparison string and the correct term. The difference degree information acquisition means 1323 acquires difference degree information between each of the two or more comparison strings acquired by the comparison string acquisition means 1322 and one or more correct terms. The difference degree information is information that specifies the degree of difference. The difference degree information is, for example, distance. The edit distance is preferable as the difference degree information, but Levenshtein distance, minimum edit distance, etc. may also be used. The difference degree information may also be, for example, the number of characters that differ.

[0068] It is preferable that the difference degree information acquisition means 1323 does not acquire difference degree information between the comparison string and the correct term when the difference between the number of characters in the comparison string and the correct term is equal to or greater than a threshold. In other words, by not calculating difference degree information between two strings with a large difference in the number of characters, it is possible to increase the processing speed. It is preferable to proceed with the processing assuming that two strings with a large difference in the number of characters are unrelated strings, rather than a typo, etc.

[0069] The difference degree information acquisition means 1323 preferably acquires the number of characters of the comparison string or the correct term to be compared with the comparison string, and acquires difference degree information according to the number of characters. The difference degree information according to the number of characters generally indicates that the larger the number of characters, the less the influence of a single character difference on the difference degree information. The difference degree information according to the number of characters generally indicates that the smaller the number of characters, the greater the influence of a single character difference on the difference degree information.

[0070] The determination means 1324 determines that the corresponding comparison character string is a typographical error if the degree of difference specified by the difference degree information is not 0 but is within or less than the threshold value. Note that a typographical error here refers to an error, and includes typos and the like.

[0071] The prediction means 1325 performs machine learning prediction processing using one or more sentences included in the test information and a learning device, and acquires location information that identifies the location of a typo or omission for each of the one or more sentences. The location information is information that identifies a location within a sentence. Examples of location information include an offset within the test object, a start address and an end address within the test object, a character, a character string, etc. It is preferable that the prediction means 1325 acquires error information that includes location information and type information.

[0072] The prediction means 1325 preferably performs machine learning prediction processing using one or more sentences included in the test information and a learning device, and also acquires a pattern identifier for each of the one or more sentences. The algorithm for the machine learning prediction processing is not important. For example, deep learning is preferable as the machine learning, but random forests, decision trees, etc. may also be used.

[0073] The image determination means 1326 acquires one or more images included in the inspection information and acquires image difference information relating to the differences between each of the one or more images and a positive example image. The image determination means 1326 then determines whether the information relating to the differences identified by the image difference information satisfies the pseudo-image conditions. The image difference information is, for example, the similarity between the two images and the differences between each feature of the two images. The feature may be, for example, the aspect ratio, color distribution, or the presence or absence of text, but is not limited thereto.

[0074] The image determining unit 1326 calculates, for example, the similarity between the positive example image and the test image. The test image is an image to be inspected.

[0075] The image determination means 1326 then determines whether the similarity satisfies the similarity conditions included in the pseudo-image conditions. If the similarity conditions are satisfied, the image determination means 1326 acquires, for example, one or more types of feature amounts of the positive example image and one or more types of feature amounts of the test image. The image determination means 1326 then compares the two feature amounts for each type and determines whether the test image is a pseudo-image.

[0076] If the inspection image is a pseudo image, the image determination unit 1326 acquires, for example, location information that identifies the location of the image in the inspection information. Also, the image determination unit 1326 acquires, for example, a type identifier "pseudo image."

[0077] The inspection result composing unit 133 uses the inspection results from the inspecting unit 132 to compose the inspection results to be output.

[0078] The inspection result construction unit 133 constructs an inspection result that identifies, for example, a comparison character string that has been determined to be a typographical error. The inspection result construction unit 133 constructs an inspection result that allows visual recognition of a portion that identifies a comparison character string that has been determined to be a typographical error.

[0079] The inspection result construction unit 133 constructs an inspection result having location information that identifies locations of typos or omissions, for example.

[0080] The test result composing unit 133 composes test results that also have a type identifier corresponding to the pattern identifier, for example.

[0081] For example, when an image included in the inspection information satisfies the pseudo-image condition, the inspection result constructing unit 133 constructs an inspection result indicating that the image is an error.

[0082] For example, the test result construction unit 133 compares errors corresponding to one or more pieces of location information paired with a type identifier corresponding to a selected item for selection with errors corresponding to other piece of location information to construct a visually different test result. Also, a visually different test result may be a test result that includes only errors corresponding to one or more pieces of location information paired with the type identifier.

[0083] The inspection result composing unit 133 composes the inspection result in a form that visually indicates the location information and type identifier, for example.

[0084] The selection interface configuration unit 134 configures a selection interface. The selection interface is an interface having one or more selection items. The selection interface has one or more type identifiers resulting from unique processing of type identifiers contained in the test results as selection items, and does not have type identifiers not contained in the test results as selection items. The selection interface is, for example, a menu, a collection of buttons, a collection of check boxes, etc. A selection item is an item that can be selected. The selection item is, for example, a menu item, a button, a check box, etc.

[0085] The selection interface configuration unit 134 typically acquires all type identifiers contained in the test results, performs unique processing on the type identifiers, and configures a selection interface in which one or more different type identifiers are used as selection items. The selection interface may have selection items (e.g., "All") corresponding to all type identifiers contained in the test results. The selection interface does not have selection items corresponding to type identifiers not contained in the test results.

[0086] The output unit 14 outputs various types of information. The various types of information are, for example, test results. Here, "output" usually means transmission to the terminal device 2, but it may also be a concept that includes display on a display, projection using a projector, printing on a printer, sound output, storage on a recording medium, and delivery of processing results to other processing devices or other programs.

[0087] The result output unit 141 outputs the test result constructed by the test result construction unit 133. The result output unit 141 transmits the test result to the terminal device 2, for example.

[0088] The selection interface output unit 142 outputs the selection interface.

[0089] Various types of information are stored in the terminal storage unit 21 constituting the terminal device 2. The various types of information include, for example, a user identifier and examination information.

[0090] The terminal reception unit 22 receives various instructions and information. The various instructions and information include, for example, examination information and selection instructions. Note that a selection instruction is an instruction to select a selection item.

[0091] The device processing unit 23 performs various types of processing. For example, the various types of processing are processing for converting instructions and information received by the terminal receiving unit 22 into instructions and information with a data structure to be transmitted. For example, the various types of processing are processing for converting information received by the terminal receiving unit 25 into information with a data structure to be output.

[0092] The terminal transmitting unit 24 transmits various instructions and information, such as test information, a user identifier, and an instruction to select an option.

[0093] The terminal receiving unit 25 receives various types of information, such as test results and selection interfaces.

[0094] The terminal output unit 26 outputs various types of information, such as test information, test results, and selection interfaces.

[0095] The storage unit 11, the user information storage unit 111, the rule storage unit 112, the learning device storage unit 113, and the terminal storage unit 21 are preferably non-volatile recording media, but can also be realized by volatile recording media.

[0096] There is no restriction on the process by which information is stored in the storage unit 11 etc. For example, information may be stored in the storage unit 11 etc. via a recording medium, information transmitted via a communication line etc. may be stored in the storage unit 11 etc., or information input via an input device may be stored in the storage unit 11 etc.

[0097] The reception unit 12, rule acquisition unit 131, inspection unit 132, inspection result configuration unit 133, selection interface configuration unit 134, and terminal reception unit 25 are usually realized by wireless or wired communication means.

[0098] The processing unit 13, rule acquisition unit 131, inspection unit 132, inspection result configuration unit 133, selection interface configuration unit 134, morphological analysis means 1321, comparison string acquisition means 1322, difference degree information acquisition means 1323, judgment means 1324, prediction means 1325, image judgment means 1326, and terminal processing unit 23 can typically be realized by a processor, memory, etc. The processing procedures of the processing unit 13, etc. are typically realized by software, and the software is recorded on a recording medium such as a ROM. However, they may also be realized by hardware (dedicated circuit). The processor may be an MPU, CPU, GPU, etc., and the type is not important.

[0099] The output unit 14, the result output unit 141, the selection interface output unit 142, and the terminal transmission unit 24 are typically realized by wireless or wired communication means.

[0100] The terminal reception unit 22 can be realized by a device driver for an input means such as a touch panel or a keyboard, or control software for a menu screen.

[0101] The terminal output unit 26 may or may not be considered to include an output device such as a display, a speaker, etc. The terminal output unit 26 may be realized by driver software for an output device, or a combination of driver software for an output device and the output device, etc.

[0102] Next, a description will be given of an example of the operation of the information system A. First, a description will be given of an example of the operation of the inspection device 1 with reference to the flowchart of FIG.

[0103] (Step S301) The test information receiving unit 121 determines whether test information, etc. has been received. If test information, etc. has been received, the process proceeds to step S302; if test information, etc. has not been received, the process returns to step S301. Note that the reception here is usually reception from the terminal device 2, but may also be reception of user input. Furthermore, the test information, etc. is, for example, test information and a user identifier. However, the test information and the user identifier do not have to be received at the same time.

[0104] (Step S302) The processing unit 13 performs an inspection process on the inspection information. An example of the inspection process will be described with reference to the flowchart of FIG.

[0105] (Step S303) The inspection result composing unit 133 performs an inspection result composing process. An example of the inspection result composing process will be described with reference to the flowchart of FIG.

[0106] (Step S304) The result output unit 141 outputs the test result constructed in step S303. Note that the output here is usually transmitted to the terminal device 2, but may also be displayed on a display, stored in a recording medium, or the like.

[0107] (Step S305) The reception unit 12 determines whether or not a selected IF output instruction has been received for the test result output in step S304. If a selected IF output instruction has been received, the process proceeds to step S306, and if a selected IF output instruction has not been received, the process proceeds to step S313. Note that the reception here is usually reception from the terminal device 2, but may also be reception of a user input.

[0108] (Step S306) The selection interface configuration unit 134 performs a selection interface configuration process. An example of the selection IF configuration process will be described with reference to the flowchart of FIG.

[0109] (Step S307) The selection interface output unit 142 outputs the selection interface configured in step S306. The output here is usually a transmission to the terminal device 2.

[0110] (Step S308) The selection receiving unit 122 determines whether one or more selection items have been selected in the selection interface output in step S307. If a selection item has been selected, the process proceeds to step S309, and if no selection item has been selected, the process proceeds to step S311. Note that the reception here is usually reception from the terminal device 2, but may also be reception of a user input.

[0111] (Step S309) The test result composing unit 133 composes test results corresponding to the selected selection items. An example of the selected test result composing process will be described with reference to the flowchart of FIG.

[0112] (Step S310) The result output unit 141 outputs the test results acquired in step S309. Note that the output here is usually transmitted to the terminal device 2, but may also be displayed on a display, stored in a recording medium, or the like.

[0113] (Step S311) The reception unit 12 determines whether or not an instruction to end the output of the test results has been received. If an instruction to end the output has been received, the process proceeds to step S312, and if an instruction to end the output has not been received, the process returns to step S308.

[0114] (Step S312) The processing unit 13 performs processing to terminate the output of the test result. The process returns to step S301. Note that this processing is, for example, disconnecting communication with the terminal device 2, logging out the user, etc.

[0115] (Step S313) The reception unit 12 determines whether or not an instruction to end the output of the test results has been received. If an instruction to end the output has been received, the process proceeds to step S314, and if an instruction to end the output has not been received, the process returns to step S305.

[0116] (Step S314) The processing unit 13 performs processing to terminate the output of the test results, and then returns to step S301.

[0117] In the flowchart of FIG. 3, the processes from steps S305 to S314 may be performed by the terminal device 2, as will be described later.

[0118] In the flowchart of FIG. 3, the process ends when the power is turned off or an interrupt occurs to end the process.

[0119] Next, an example of the inspection process in step S302 will be described with reference to the flowchart in FIG.

[0120] (Step S401) The rule obtaining unit 131 obtains a user identifier corresponding to the received test information.

[0121] (Step S402) The rule acquisition unit 131 acquires one or more inspection rules paired with the user identifier acquired in step S401 from the rule storage unit 112. Note that the rule acquisition unit 131 may also acquire one or more inspection rules that are commonly used by two or more users from the rule storage unit 112.

[0122] (Step S403) The inspection unit 132 assigns 1 to the counter i.

[0123] (Step S404) The inspection unit 132 determines whether the i-th sentence is present in the received inspection information. If the i-th sentence is present, the process proceeds to step S405, and if the i-th sentence is not present, the process proceeds to step S411.

[0124] (Step S405) The checking unit 132 acquires the i-th sentence from the check information.

[0125] (Step S406) The inspection unit 132 assigns 1 to the counter j.

[0126] (Step S407) The inspection unit 132 determines whether or not there is an inspection rule corresponding to the j-th sentence from among the one or more inspection rules acquired in step S402. If there is an inspection rule corresponding to the j-th sentence, the process proceeds to step S408; if there is no inspection rule corresponding to the j-th sentence, the process proceeds to step S410.

[0127] (Step S408) The inspection unit 132 inspects the i-th sentence using the inspection rule corresponding to the j-th sentence. An example of such sentence inspection processing will be described using the flowcharts of Figures 5, 6, and 7. It is preferable that the inspection unit 132 performs inspections in all of the flowcharts of Figures 5, 6, and 7.

[0128] (Step S409) The inspection unit 132 increments the counter j by 1. The process returns to step S407.

[0129] (Step S410) The inspection unit 132 increments the counter i by 1. The process returns to step S404.

[0130] (Step S411) The inspection unit 132 assigns 1 to a counter k.

[0131] (Step S412) The inspection unit 132 determines whether the kth inspection image is present in the inspection information. If the kth inspection image is present, the process proceeds to step S413; if not, the process returns to the upper process.

[0132] (Step S413) The inspection unit 132 assigns 1 to the counter l.

[0133] (Step S414) The inspection unit 132 determines whether or not an inspection rule corresponding to the lth positive example image exists. If such an inspection rule exists, the process proceeds to step S415; if not, the process proceeds to step S417. Note that there may be one or more positive example images. There may also be one or more inspection rules.

[0134] (Step S415) The inspection unit 132 performs image inspection processing on the kth inspection target image using the inspection rule corresponding to the lth positive example image. An example of the image inspection processing will be described with reference to the flowchart in FIG.

[0135] (Step S416) The inspection unit 132 increments the counter 1 by 1. The process returns to step S414.

[0136] (Step S417) The inspection unit 132 increments the counter k by 1. The process returns to step S412.

[0137] Next, a first example of the sentence checking process in step S408 will be described with reference to the flowcharts of Figures 5, 6 and 7. The first example (sentence check 1) is a check for spelling errors in a correct term.

[0138] (Step S501) The morphological analysis unit 1321 performs morphological analysis on the target sentence and divides the sentence into morphemes. Note that here, the inspection unit 132 may divide the target sentence into one or more words.

[0139] (Step S502) The comparison character string obtaining unit 1322 assigns 1 to a counter i.

[0140] (Step S503) The comparison character string acquisition unit 1322 determines whether the i-th morpheme exists among the one or more morphemes acquired in step S501. If the i-th morpheme exists, the process proceeds to step S504; if the i-th morpheme does not exist, the process returns to the upper process.

[0141] (Step S504) The comparison character string obtaining unit 1322 assigns 1 to a counter j.

[0142] (Step S505) The comparison string acquisition means 1322 determines whether or not a jth comparison string that starts with the ith morpheme exists. If the jth comparison string exists, the process proceeds to step S506; if the jth comparison string does not exist, the process proceeds to step S519. The jth comparison string is a string that starts with the ith morpheme and is made up of one or more consecutive morphemes.

[0143] (Step S506) The comparison character string obtaining unit 1322 obtains the j-th comparison character string that begins with the i-th morpheme.

[0144] (Step S507) The difference degree information acquisition means 1323 acquires the number of characters (N1) of the comparison character string acquired in step S506.

[0145] (Step S508) The difference degree information acquiring means 1323 assigns 1 to a counter k.

[0146] (Step S509) The difference degree information acquiring means 1323 determines whether the kth correct term exists in the correct term dictionary in the rule storage unit 112. If the kth correct term exists, the process proceeds to step S510; if not, the process proceeds to step S518.

[0147] (Step S510) The difference degree information obtaining means 1323 obtains the k-th correct term from the correct term dictionary.

[0148] (Step S511) The difference degree information obtaining means 1323 obtains the number of characters (N2) of the correct term obtained in step S510.

[0149] (Step S512) The difference degree information acquisition means 1323 determines whether N1 and N2 satisfy a predetermined condition. If the condition is satisfied, the process proceeds to step S513, and if the condition is not satisfied, the process proceeds to step S517. The predetermined condition is, for example, a condition regarding the difference between N1 and N2, and typically, the difference must be small, for example, "|N1-N2|<threshold" or "|N1-N2|<=threshold."

[0150] (Step S513) The difference degree information acquiring unit 1323 acquires difference degree information between the k-th correct term and the comparison string acquired in step S506. Note that the difference degree information is, for example, the edit distance between the two strings.

[0151] (Step S514) The determining means 1324 acquires the conditions. It is preferable that the inspecting unit 132 acquires the conditions according to N1, N2, or N1 and N2.

[0152] (Step S515) The determination means 1324 determines whether or not the difference degree information satisfies the condition acquired in step S514. If the condition is satisfied, the process proceeds to step S516, and if the condition is not satisfied, the process proceeds to step S517.

[0153] (Step S516) The determination means 1324 acquires the location information of the j-th comparison character string and a type identifier that identifies the type of error (here, for example, "fluctuation in expression"), and stores them in a buffer (not shown). The process proceeds to step S518.

[0154] (Step S517) The difference degree information acquiring means 1323 increments the counter k by 1. The process returns to step S509.

[0155] (Step S518) The comparison character string acquisition unit 1322 increments the counter j by 1. The process returns to step S503.

[0156] (Step S519) The comparison character string acquisition unit 1322 increments the counter i by 1. The process returns to step S505.

[0157] Next, a second example of the sentence inspection process in step S408 will be described with reference to the flowcharts in FIGS. 6 and 7. The second example (Sentence Inspection 2) is an inspection for inappropriate expressions using multiple terms. The inspection rule here is a rule having the structure of "Term 1", "Relative Position", "Term 2". For example, the inspection rule here is "Term 2 must be present within N terms after Term 1". For example, the inspection rule is "'slimming' must be present within 5 words after 'absolute'".

[0158] (Step S601) The morphological analysis unit 1321 performs morphological analysis on the target sentence and divides the sentence into morphemes.

[0159] (Step S602) The inspection unit 132 assigns 1 to the counter i.

[0160] (Step S603) The inspection unit 132 determines whether the i-th morpheme exists among the one or more morphemes acquired in step S601. If the i-th morpheme exists, the process proceeds to step S504; if the i-th morpheme does not exist, the process returns to the upper process.

[0161] (Step S604) The inspection unit 132 assigns 1 to the counter j.

[0162] (Step S605) The inspection unit 132 determines whether or not the j-th inspection rule exists. If the j-th inspection rule exists, the process proceeds to step S606, and if the j-th inspection rule does not exist, the process proceeds to step S615.

[0163] (Step S606) The inspection unit 132 acquires the i-th morpheme. The i-th morpheme is assumed to be "Term 1."

[0164] (Step S607) The inspection unit 132 determines whether or not "Term 1" acquired in step S606 matches "Term 1" of the j-th inspection rule. If they match, the process proceeds to step S608, and if they do not match, the process proceeds to step S614.

[0165] (Step S608) The inspection unit 132 acquires one or more "Term 2"s corresponding to the relative position of the j-th inspection rule with respect to "Term 1." If the inspection rule is "'soreru' (to lose weight) is present within five words after 'zettai' (absolute)," the inspection unit 132 acquires the first to fifth morphemes after Term 1 as the relative positions.

[0166] (Step S609) The inspection unit 132 assigns 1 to a counter k.

[0167] (Step S610) The inspection unit 132 determines whether the kth morpheme exists in "Term 2" acquired in step S608. If the kth morpheme exists, the process proceeds to step S607, and if the kth morpheme does not exist, the process proceeds to step S614.

[0168] (Step S611) The inspection unit 132 determines whether the k-th morpheme (term 2) matches "term 2" of the j-th inspection rule. If they match, the process proceeds to step S612, and if they do not match, the process proceeds to step S613.

[0169] (Step S612) The inspection unit 132 acquires the location information and type identifier (here, for example, "inappropriate expression") corresponding to "Term 1" and "Term 2", and stores them in a buffer (not shown).

[0170] (Step S613) The inspection unit 132 increments the counter k by 1. The process returns to step S610.

[0171] (Step S614) The inspection unit 132 increments the counter j by 1. The process returns to step S605.

[0172] (Step S615) The inspection unit 132 increments the counter i by 1. The process returns to step S603.

[0173] Next, a third example of the sentence checking process in step S408 will be described with reference to the flowchart in Fig. 7. The third example (sentence check 3) is a check for typos and omissions.

[0174] (Step S701) The prediction means 1325 acquires a learning device from the learning device storage unit 113.

[0175] (Step S702) The prediction means 1325 acquires the sentence to be checked.

[0176] (Step S703) The prediction means 1325 performs a prediction process using the learning device and the sentence, and obtains a prediction result.

[0177] (Step S704) The prediction means 1325 assigns 1 to the counter i.

[0178] (Step S705) The prediction means 1325 determines whether or not the i-th error exists in the prediction result acquired in step S703. If the i-th error exists, the process proceeds to step S706, and if not, the process returns to the upper process.

[0179] (Step S706) The prediction means 1325 obtains location information corresponding to the i-th error and a type identifier included in the i-th error, and stores them in a buffer (not shown).

[0180] (Step S707) The prediction means 1325 increments the counter i by 1. The process returns to step S705.

[0181] Next, an example of the image inspection process in step S415 will be described with reference to the flowchart in FIG.

[0182] (Step S801) The image determination unit 1326 adjusts the size of the test image or the positive example image. Note that adjusting the size means, for example, adjusting the vertical or horizontal size.

[0183] (Step S802) The image determination unit 1326 calculates the similarity between the test image and the positive example image. Note that the similarity is an example of image difference information.

[0184] (Step S803) The image determination means 1326 determines whether the similarity acquired in step S802 satisfies the pseudo image condition. If the condition is satisfied, the process proceeds to step S804, and if the condition is not satisfied, the process returns to the upper process.

[0185] (Step S804) The image determination means 1326 acquires one or more feature amounts of the positive example image.

[0186] (Step S805) The image determination means 1326 acquires one or more feature amounts of the inspection image.

[0187] (Step S806) The image determination means 1326 assigns 1 to the counter i.

[0188] (Step S807) The image determination means 1326 determines whether or not the i-th type of feature exists. If the i-th type of feature exists, the process proceeds to step S808, and if not, the process returns to the upper process.

[0189] (Step S808) The image determination means 1326 determines whether the i-th type of feature of the positive example image and the i-th type of feature of the test image satisfy a condition. If the condition is satisfied, the process proceeds to step S809; if the condition is not satisfied, the process proceeds to step S810.

[0190] (Step S809) The image determination unit 1326 acquires location information specifying the position of the inspection image and a type identifier corresponding to the feature amount of the ith type, and stores them in a buffer (not shown).

[0191] (Step S810) The image determination means 1326 increments the counter 1 by 1. The process proceeds to step S807.

[0192] Next, an example of the inspection result configuration process in step S303 will be described with reference to the flowchart in FIG.

[0193] (Step S901) The test result configuration unit 133 acquires test information.

[0194] (Step S902) The inspection result configuration unit 133 assigns 1 to the counter i.

[0195] (Step S903) The inspection result configuration unit 133 determines whether the i-th error information exists in a buffer (not shown). If the i-th error information exists, the process proceeds to step S904; if not, the process returns to the upper processing. The error information includes location information and a type identifier.

[0196] (Step S904) The inspection result configuration unit 133 acquires location information from the i-th error information.

[0197] (Step S905) The inspection result configuration unit 133 acquires the type identifier from the i-th error information.

[0198] (Step S906) The inspection result composing unit 133 displays the information corresponding to the location information in the inspection result including the inspection information in a manner that allows the type of error corresponding to the type identifier to be identified.

[0199] (Step S907) The inspection result configuration unit 133 increments the counter i by 1. The process returns to step S903.

[0200] Next, an example of the selection IF configuration process in step S306 will be described with reference to the flowchart in FIG.

[0201] (Step S1001) The selection interface configuration unit 134 acquires type identifiers from all the error information in a buffer (not shown).

[0202] (Step S1002) The selection interface configuration unit 134 acquires the number of errors for each type identifier. The number of errors for each type identifier is the number of pieces of error information for each type identifier.

[0203] (Step S1003) The selection interface configuration unit 134 performs unique processing on the type identifier acquired in step S1001.

[0204] (Step S1004) The selection interface configuration unit 134 sorts the type identifiers. Note that the selection interface configuration unit 134 sorts the type identifiers in descending order, for example, using the number of errors acquired in step S1002 as a key. The selection interface configuration unit 134 also sorts the type identifiers, for example, according to a predetermined priority order of the type identifiers.

[0205] (Step S1005) The selection interface configuration unit 134 configures a selection interface having, as selection items, character strings or images corresponding to the type identifier, and returns to the upper processing.

[0206] In the flowchart of FIG. 10, the selection interface configuration unit 134 may configure a selection interface that also has predetermined selection items (for example, "all").

[0207] Next, an example of the selected test result configuration process in step S309 will be described with reference to the flowchart of FIG.

[0208] (Step S1101) The test result configuration unit 133 acquires test information.

[0209] (Step S1102) The inspection result structuring unit 133 acquires the type identifier corresponding to the selection.

[0210] (Step S1103) The inspection result configuration unit 133 assigns 1 to the counter i.

[0211] (Step S1104) The inspection result configuration unit 133 determines whether the i-th error information including the type identifier acquired in step S1102 exists in a buffer (not shown). If the i-th error information exists, the process proceeds to step S1105; if not, the process returns to the upper processing.

[0212] (Step S1105) The inspection result configuration unit 133 acquires location information contained in the i-th error information.

[0213] (Step S1106) In the inspection result including the inspection information, the inspection result configuration unit 133 displays the information corresponding to the location information acquired in step S1105 in a different way from the others. Displaying the information in a different way from the others means, for example, highlighting the information, not displaying errors of other type identifiers, etc.

[0214] (Step S1107) The inspection result structuring unit 133 increments the counter i by 1. The process returns to step S1104.

[0215] Next, an example of the operation of the terminal device 2 will be described with reference to the flowchart of Fig. 12. In the flowchart of Fig. 12, the description of the same steps as those in the flowchart of Fig. 3 will be omitted.

[0216] (Step S1201) The terminal reception unit 22 determines whether or not the examination information has been received. If the examination information has been received, the process proceeds to step S1202, and if the examination information has not been received, the process returns to step S1201.

[0217] (Step S1202) The device processing unit 23 acquires the user identifier from the terminal storage unit 21. The device processing unit 23 also associates the user identifier with the test information. The terminal transmitting unit 24 transmits the user identifier and the test information to the test device 1.

[0218] (Step S1203) The terminal receiving unit 25 determines whether or not the test results have been received from the test device 1. If the test results have been received, the process proceeds to step S1204, and if the test results have not been received, the process returns to step S1203.

[0219] (Step S1204) The terminal processing unit 23 is configured to output the inspection result received in step S1203. The terminal output unit 26 outputs the inspection result. The process proceeds to step S305.

[0220] In the flowchart of FIG. 12, the process ends when the power is turned off or an interrupt occurs to end the process.

[0221] A specific example of the operation of the information system A in this embodiment will be described below.

[0222] Assume that two or more users, including user A with user identifier "U001", input their own company's specific inspection rules into terminal device 2 and register them in inspection device 1. As a result, the rule management table shown in Fig. 13 is stored in rule storage unit 112 of inspection device 1. Note that any interface, etc., through which users input inspection rules into terminal device 2 may be used.

[0223] 13 is a table for managing inspection rules for each user. The rule management table manages one or more records each having an "ID," "inspection rule," "type identifier," and "user identifier."

[0224] The learning device storage unit 113 stores learning devices acquired by the learning device 3, which will be described later. Here, the learning devices are, for example, learning device 1, learning device 2, etc., and are associated with user identifiers. However, the learning devices may be information common to two or more users.

[0225] 13 is a rule for detecting pseudo images, and includes the condition "<aspect ratio> matches, <character string> exists, <color> matches" for the positive example image 1301. "<aspect ratio> matches, <character string> exists, <color> matches" indicates that any image other than the one that matches the aspect ratio of the positive example image 1301, has not only an image but also a character string (here, "MIRASENSES"), and also matches the color, is deemed to be a pseudo image and an error is detected.

[0226] In this situation, the following four specific examples will be explained. Specific Example 1 is a case where various sentences are inspected. Specific Example 2 is a case where a selection interface is constructed from the inspection results and output. Specific Example 3 is a case where only errors corresponding to selected items are output using the selection interface. Specific Example 4 is a case where pseudo-images are detected.

[0227] (Example 1) Now, assume that user A with user identifier "U001" inputs test information for which he or she wishes to be tested into terminal device 2. Then, the terminal device 2 transmits the test information and user identifier "U001" to the test device 1.

[0228] Next, the test information receiving unit 121 of the test device 1 receives the test information and the user identifier "U001".

[0229] Next, the processing unit 13 performs an inspection process on the inspection information as follows: First, the rule acquisition unit 131 acquires one or more inspection rules paired with the received user identifier "U001" from the rule management table of FIG.

[0230] Next, the inspection unit 132 first applies the inspection rule "ID=1" to the inspection information, and detects spelling variations of "interface" by the operation described using the flowchart in Figure 5, and accumulates error information having the location information and the type identifier "spelling variation" in a buffer not shown.

[0231] Furthermore, the inspection unit 132 applies the inspection rule "ID=2" to the inspection information, and detects spelling variations of "database" through the operation described using the flowchart in Figure 5, and stores error information having the location information and the type identifier "spelling variation" in a buffer (not shown).

[0232] Furthermore, the inspection unit 132 applies the inspection rule of "ID=28" to the inspection information, and determines whether or not an expression including the term 2 "lose weight" exists within five words following the term 1 "absolutely" by the operation described using the flowchart in Fig. 6. If an expression exists, the inspection unit 132 stores error information including the location information of the term 1 and the term 2 and the type identifier "exaggerated expression" in a buffer (not shown).

[0233] Similarly, the inspection unit 132 applies the inspection rule of "ID=29" to the inspection information, and determines whether or not an expression including the term 2 "disappear" exists within three words following the term 1 "complete" by the operation described using the flowchart in Fig. 6. If an expression exists, the inspection unit 132 stores error information having the location information of the term 1 and the term 2 and the type identifier "exaggerated expression" in a buffer (not shown).

[0234] Furthermore, the inspection unit 132 applies the inspection rule for "ID=56" to the inspection information, and acquires the learning device 1 paired with the user identifier "U001" from the learning device storage unit 113 through the operation described using the flowchart in Figure 7. The inspection unit 132 also performs a prediction process using the learning device 1 and each sentence of the inspection device, and acquires the prediction result. The inspection unit 132 also accumulates one or more pieces of error information included in the prediction result in a buffer (not shown). Note that this prediction result includes one or more pieces of error information having information on the location of the typo or omission and the type identifier "typo / omission."

[0235] After the above-described inspection process is completed, the inspection result composing unit 133 performs the inspection result composing process described using the flowchart in Fig. 9. Next, the result output unit 141 transmits the constructed inspection result to the terminal device 2.

[0236] Next, the terminal device 2 receives and outputs the test results. An example of such output is shown in FIG. 14. In FIG. 14, 1401 is text of the test information. Also in FIG. 14, the test information is displayed so that the location of the error and the type of error (type identifier) ​​can be visually recognized. In FIG. 14, the background of the text at the location of the error is a color corresponding to the type identifier. Also, in FIG. 14, 1402, a collection of error information is output in a manner that clearly indicates the content of the error for each piece of error information.

[0237] (Example 2) 14 displayed on the terminal device 2. Then, the terminal device 2 receives a selection IF output instruction. The terminal device 2 then transmits the selection IF output instruction to the inspection device 1.

[0238] Next, the reception unit 12 of the inspection device 1 receives a selection IF output instruction. Then, the selection interface configuration unit 134 performs the selection interface configuration process as follows. That is, the selection interface configuration unit 134 acquires type identifiers from all error information in a buffer (not shown). Here, it is assumed that the selection interface configuration unit 134 has acquired multiple "spelling variations," numerous "typographical errors," and a "spelling variation group display." It is also assumed that the error information in the buffer (not shown) (the error information for the error detected in Specific Example 1) did not include a "pseudo image."

[0239] Next, the selection interface configuration unit 134 obtains the number of errors for each of the obtained type identifiers.

[0240] Next, the selection interface configuration unit 134 performs unique processing on the acquired type identifiers and acquires "spelling variations," "misspellings," and "spelling variation group display." Next, the selection interface configuration unit 134 sorts the type identifiers in descending order, for example, using the acquired number of errors as a key, and acquires the type identifiers in the order of "misspellings," "spelling variations," and "spelling variation group display."

[0241] Next, the selection interface configuration unit 134 obtains the default selection items "All," "User-defined Rules," "User-defined Errata," and "Other Rules" from the storage unit 11. Then, the selection interface configuration unit 134 configures a selection interface (here, a menu) in which the selection items "Misspellings," "Spelling Variations," and "Show Spelling Variation Groups" are arranged in this order below the default selection items.

[0242] Next, the selection interface output unit 142 transmits the constructed menu to the terminal device 2.

[0243] Next, the terminal device 2 receives and outputs the menu. An example of such output is shown in Figure 15. An example of the menu that is a selection interface is shown as 1501 in Figure 15. 1501 does not have the menu item "pseudo image" corresponding to the type identifier that was not detected as an error.

[0244] The selection interface configuration process and the selection interface output process may be performed by the terminal device 2 instead of the inspection device 1. In this case, communication between the terminal device 2 and the inspection device 1 for the selection interface configuration process does not occur, which is preferable.

[0245] (Example 3) Next, assume that user A selects the selection item "typographical errors and omissions" from the menu 1501 in Fig. 15. Then, the terminal device 2 accepts the selection of the selection item "typographical errors and omissions". Next, the terminal device 2 transmits a selection identifying the selection item "typographical errors and omissions" to the inspection device 1.

[0246] Next, the selection receiving unit 122 of the inspection device 1 receives a selection instruction specifying the selection item "typographical error or omission."

[0247] Then, the inspection result configuration unit 133 acquires from the error information a large number of pieces of location information paired with the type identifier "typo" corresponding to the selected selection item "typo." The inspection result configuration unit 133 then configures an inspection result that indicates only the information corresponding to the acquired location information as the error location. Next, the result output unit 141 transmits the acquired inspection result to the terminal device 2.

[0248] The terminal device 2 receives and outputs the inspection result, which indicates only errors corresponding to "typographical errors and omissions."

[0249] (Example 4) 16 to the terminal device 2. The terminal device 2 then transmits the test information to the test device 1 in combination with the user identifier "U001."

[0250] Next, the test information receiving unit 121 of the test device 1 receives the user identifier "U001" and the test information.

[0251] Next, the rule acquisition unit 131 acquires one or more check rules paired with the received user identifier "U001" from the rule management table of FIG.

[0252] Next, the checking unit 132 performs the above-described check on each sentence in the check information in Fig. 16. Here, it is assumed that the checking unit 132 does not obtain any error information as a result of checking each sentence in the check information in Fig. 16.

[0253] Next, the inspection unit 132 inspects the pseudo image using the inspection rule of "ID=57" in FIG. 13 as follows.

[0254] That is, first, inspection unit 132 acquires inspection image 1601 from the inspection information in Fig. 16. Next, inspection unit 132 enlarges or reduces inspection image 1601 so that the vertical size of inspection image 1601 becomes the same as the vertical size of the positive example image and so that the aspect ratio of inspection image 1601 does not change.

[0255] Next, it is assumed that the inspection unit 132 calculates the similarity "S1" between the inspection image 1601 and the positive example image. Next, it is assumed that the inspection unit 132 determines that the similarity "S1" satisfies the pseudo image condition (here, "0.8<=similarity (S)<1").

[0256] Next, it is assumed that the inspection unit 132 acquires three feature amounts (<aspect ratio> AR0, <text> present, <color distribution> color distribution information 0) of the positive example image. Note that "AR0" is the numerical value of the aspect ratio, and "color distribution information 0" is information specifying the color distribution.

[0257] Next, it is assumed that the inspection unit 132 acquires three feature amounts of the inspection image 1601 (<aspect ratio> AR0 <character string> present <color distribution> color distribution information 1).

[0258] Next, the inspection unit 132 determines that the <aspect ratio> and <character string> satisfy (match) the conditions for the inspection image 1601. On the other hand, the inspection unit 132 determines that the "colors are different" for the inspection image 1601 because the <color distribution> does not match. In other words, the inspection unit 132 obtains an inspection result that the inspection image 1601 is a pseudo image and that the "colors are different."

[0259] Here, the inspection unit 132 acquires location information that identifies a location within the inspection information of the inspection image 1601, a type identifier "pseudo image," and error information having the error content "different color," and temporarily stores these in a buffer not shown.

[0260] Next, the inspection unit 132 acquires the inspection image 1602 in the inspection information in Fig. 16. Next, the inspection unit 132 enlarges or reduces the inspection image 1602 so that the vertical size of the inspection image 1602 becomes the same as the vertical size of the positive example image and so that the aspect ratio of the inspection image 1602 does not change.

[0261] Next, it is assumed that the inspection unit 132 calculates the similarity "S2" between the inspection image 1602 and the positive example image. Next, it is assumed that the inspection unit 132 determines that the similarity "S2" satisfies the pseudo image condition (here, "0.8<=similarity (S)<1").

[0262] Next, it is assumed that the inspection unit 132 acquires three feature amounts of the inspection image 1602 (<aspect ratio> AR1, <character string> present, <color distribution> color distribution information 0).

[0263] Next, the inspection unit 132 determines that the <color distribution> and <character string> satisfy the conditions for the inspection image 1602. On the other hand, the inspection unit 132 determines that the "aspect ratio" for the inspection image 1601 does not match, so that the "aspect ratio is different." In other words, the inspection unit 132 obtains an inspection result that the inspection image 1602 is a pseudo image and that the "aspect ratio is different."

[0264] Here, the inspection unit 132 acquires location information that identifies a location within the inspection information of the inspection image 1602, a type identifier "pseudo image," and error information having the error content "aspect ratio is different," and temporarily stores these in a buffer not shown.

[0265] Next, the inspection unit 132 acquires the inspection image 1603 in the inspection information in Fig. 16. Next, the inspection unit 132 enlarges or reduces the inspection image 1603 so that the vertical size of the inspection image 1603 becomes the same as the vertical size of the positive example image and so that the aspect ratio of the inspection image 1603 does not change.

[0266] Next, it is assumed that the inspection unit 132 calculates the similarity "S3" between the inspection image 1603 and the positive example image. Next, it is assumed that the inspection unit 132 determines that the similarity "S3" does not satisfy the pseudo-image condition (here, "0.8<=similarity (S)<1"). In other words, the inspection unit 132 determines that the inspection image 1603 is not a pseudo-image.

[0267] Next, the inspection unit 132 acquires the inspection image 1604 in the inspection information in Fig. 16. Next, the inspection unit 132 enlarges or reduces the inspection image 1604 so that the vertical size of the inspection image 1604 becomes the same as the vertical size of the positive example image and so that the aspect ratio of the inspection image 1604 does not change.

[0268] Next, it is assumed that the inspection unit 132 calculates the similarity "S4" between the inspection image 1604 and the positive example image. Next, it is assumed that the inspection unit 132 determines that the similarity "S4" satisfies the pseudo image condition (here, "0.8<=similarity (S)<1").

[0269] Next, it is assumed that the inspection unit 132 acquires three feature amounts (<aspect ratio> AR0, <character string> none, <color distribution> color distribution information 0) of the inspection image 1604.

[0270] Next, the inspection unit 132 determines that the <aspect ratio> and <color distribution> of the inspection image 1602 satisfy the conditions. On the other hand, the inspection unit 132 determines that "there is no character string" for the inspection image 1601 because the <character string> does not match. In other words, the inspection unit 132 obtains an inspection result that the inspection image 1602 is a pseudo image and that "there is no character string."

[0271] Here, the inspection unit 132 acquires location information that identifies a location within the inspection information of the inspection image 1604, a type identifier "pseudo image," and error information having the error content "no string," and temporarily stores these in a buffer not shown.

[0272] Next, the inspection result composing unit 133 composes inspection results including error information corresponding to the inspection images 1601, 1602, and 1604. Next, the result output unit 141 transmits the constructed inspection results to the terminal device 2.

[0273] Next, the terminal device 2 receives and outputs the test results. An example of such output is shown in FIG. 17. In FIG. 17, three pseudo images for the positive example image are clearly indicated. It is preferable that errors in the pseudo images are clearly indicated in the test information of FIG. 16. In other words, it is preferable that the test result configuration unit 133 obtains test results that clearly indicate the pseudo images in the test information.

[0274] As described above, according to this embodiment, it is possible to inspect text based on inspection rules for each user.

[0275] Furthermore, according to this embodiment, it is possible to check for errors in the correct terms registered by the user.

[0276] Furthermore, according to this embodiment, typographical errors and omissions can be checked appropriately.

[0277] Furthermore, according to this embodiment, the location and type of the error can also be presented.

[0278] Furthermore, according to this embodiment, errors in the pseudo image can also be inspected.

[0279] Furthermore, according to this embodiment, the inspection results can be viewed with high operability using a selection interface according to the type of error.

[0280] In this embodiment, the inspection device 1 may be a standalone device. Also, in this embodiment, the inspection rule does not have to be managed for each user. Also, in this embodiment, the inspection rule may be a common rule for two or more users.

[0281] Furthermore, the processing in this embodiment may be implemented by software. This software may be distributed by software download or the like. Furthermore, this software may be recorded on a recording medium such as a CD-ROM and distributed. This also applies to other embodiments in this specification. The software implementing the information processing device in this embodiment is the following program. That is, this program causes a computer to function as: a test information receiving unit that receives test information including a sentence in association with a user identifier; a rule acquisition unit that acquires, for each user identifier, one or more test rules corresponding to the user identifier that corresponds to the test information from a rule storage unit in which one or more test rules for inspecting sentences are stored in association with the user identifier; an inspection unit that performs an inspection on the test information using the one or more inspection rules; an inspection result configuration unit that configures an inspection result to be output using the inspection result of the inspection unit; and a result output unit that outputs the inspection result.

[0282] (Embodiment 2) In this embodiment, a learning device that constitutes the above-mentioned learning device will be described.

[0283] 18 is a block diagram of a learning device 3 according to this embodiment. The learning device 3 includes a learning storage unit 31, a learning reception unit 32, and a learning processing unit 33. The learning storage unit 31 includes a positive example sentence storage unit 311 and a pattern information storage unit 312. The learning processing unit 33 includes a positive example sentence acquisition unit 331, a negative example sentence acquisition unit 332, a learning unit 333, and an accumulation unit 334.

[0284] Various types of information are stored in the learning storage unit 31. The various types of information include, for example, positive example sentences (to be described later), negative example sentences (to be described later), pattern information (to be described later), and a learning device.

[0285] One or more positive example sentences are stored in the positive example sentence storage unit 311. A positive example sentence is a correct sentence.

[0286] The pattern information storage unit 312 stores one or more pieces of pattern information. Pattern information is information relating to patterns of spelling errors. Pattern information is usually associated with a pattern identifier. Examples of pattern identifiers include "homonyms," "typos," and "similar characters." The pattern identifier may also be a type identifier.

[0287] The pattern information corresponding to the pattern identifier "homonym" is, for example, a program that generates homonyms for a correct term, or / and a homonym dictionary, which is one or more sets of a correct term and one or more homonyms. An example of a homonym dictionary set (correct term: homonym [incorrect term]) is (documents: various) (entry: introduction).

[0288] A program that generates homonyms is, for example, a program that refers to a kanji dictionary that has many pairs of kanji and readings, and generates homonyms by replacing one kanji character in a term with a kanji character that has the same reading.

[0289] The pattern information corresponding to the pattern identifier "Typo" is, for example, a program that generates correct characters and one or more typo characters, or / and a Typo dictionary, which is one or more sets of correct characters and one or more typo characters. An example of a set of a Typo dictionary (correct characters:one or more typo characters) is (ma:mma).

[0290] The program that generates the typo character is, for example, a program that refers to a Roman alphabet dictionary that has many pairs of hiragana characters and Roman alphabets (consonant + vowel), obtains the consonant of the correct character, and places the consonant before the correct character.

[0291] The pattern information corresponding to the pattern identifier "similar characters" is, for example, a program that generates similar characters for a correct character, or / and a similar character dictionary, which is one or more sets of correct characters and similar characters. An example of a set of similar characters (correct character:similar characters) is (mame:me).

[0292] The program for generating similar characters is, for example, a program that refers to a similar character dictionary and replaces any hiragana character in a sentence with a similar character that pairs with the hiragana character.

[0293] The learning receiving unit 32 receives various information and instructions. The various information and instructions are, for example, correct example sentences and learning instructions. A learning instruction is an instruction to start learning.

[0294] The learning processing unit 33 performs a learning process. The learning process is performed by, for example, a positive example sentence acquisition unit 331, a negative example sentence acquisition unit 332, a learning unit 333, and an accumulation unit 334. However, although it is preferable that the negative example sentences are automatically generated, they do not have to be automatically generated.

[0295] The positive example sentence acquisition unit 331 acquires one or more positive example sentences. For example, the positive example sentence acquisition unit 331 acquires one or more positive example sentences from the positive example sentence storage unit 311. For example, the positive example sentence acquisition unit 331 acquires the positive example sentences accepted by the learning acceptance unit 32.

[0296] The negative example sentence acquisition unit 332 acquires one or more negative example sentences. It is preferable that the negative example sentence acquisition unit 332 automatically generates one or more negative example sentences using positive example sentences. A negative example sentence is a sentence that includes a typographical error. A negative example sentence corresponds to a positive example sentence.

[0297] The negative example sentence acquiring unit 332 acquires one or more negative example sentences for each of one or more positive example sentences, for example, by using one or more pieces of pattern information. A detailed process example in which the negative example sentence acquiring unit 332 acquires negative example sentences will be described later.

[0298] The learning unit 333 performs machine learning learning processing on two or more pieces of training data to obtain a learning device. The training data includes a positive example sentence and one or more negative example sentences. It is preferable that the training data includes one positive example sentence and one negative example sentence. It is preferable that the training data includes a type identifier that identifies the type of error. The type identifier is, for example, "homonym," "typo," or "similar character."

[0299] The learning unit 333 preferably performs machine learning learning processing using two or more sets of training data using, for example, BERT to acquire a learning device. For information about BERT "Bidirectional Encoder Representations from Transformers," see the URL "https: / / arxiv.org / pdf / 1810.04805.pdf."

[0300] There is no restriction on the machine learning algorithm used by the learning unit 333. The learning unit 333 performs a learning process on two or more pieces of training data using a machine learning algorithm such as deep learning, random forest, or decision tree, to acquire a learning device.

[0301] The storage unit 334 stores the learning device acquired by the learning unit 333.

[0302] The learning storage unit 31, the positive example sentence storage unit 311, and the pattern information storage unit 312 are preferably non-volatile recording media, but can also be realized as volatile recording media.

[0303] There is no restriction on the process by which information is stored in the learning storage unit 31 etc. For example, information may be stored in the learning storage unit 31 etc. via a recording medium, information transmitted via a communication line etc. may be stored in the learning storage unit 31 etc., or information input via an input device may be stored in the learning storage unit 31 etc.

[0304] The learning acceptance unit 32 can be realized by a device driver for an input means such as a touch panel or keyboard, or control software for a menu screen.

[0305] The learning processing unit 33, positive example sentence acquisition unit 331, negative example sentence acquisition unit 332, learning unit 333, and accumulation unit 334 can typically be realized by a processor, memory, or the like. The processing procedures of the learning processing unit 33 and the like are typically realized by software, and the software is recorded on a recording medium such as a ROM. However, they may also be realized by hardware (dedicated circuitry). The processor may be an MPU, CPU, GPU, or the like, and the type is not important.

[0306] Next, an example of the operation of the learning device 3 will be described with reference to the flowchart of FIG.

[0307] (Step S1901) The learning receiving unit 32 determines whether or not a learning instruction has been received. If a learning instruction has been received, the process proceeds to step S1902, and if a learning instruction has not been received, the process returns to step S1901.

[0308] (Step S1902) The learning processing unit 33 assigns 1 to the counter i.

[0309] (Step S1903) The learning processing unit 33 determines whether or not the i-th positive example sentence exists in the positive example sentence storage unit 311. If the i-th positive example sentence exists, the process proceeds to step S1904; if not, the process proceeds to step S1907.

[0310] (Step S1904 ) The positive example sentence acquiring unit 331 acquires the i-th positive example sentence from the positive example sentence storage unit 311 .

[0311] (Step S1905) The learning processing unit 33 acquires training data using the i-th positive example sentence. An example of the training data acquisition process will be described with reference to the flowchart in FIG.

[0312] (Step S1906) The learning processing unit 33 increments the counter i by 1. The process returns to step S1903.

[0313] (Step S1907) The learning processing unit 33 performs a learning process on the two or more pieces of teacher data acquired in step S1905, and acquires a learning device.

[0314] (Step S1908) The storage unit 334 stores the learning device acquired in step S1907. The process returns to step S1901.

[0315] In the flowchart of FIG. 19, the process ends when the power is turned off or an interrupt occurs to end the process.

[0316] Next, an example of the teacher data acquisition process in step S1905 will be described with reference to the flowchart in FIG.

[0317] (Step S2001) The negative example sentence acquiring unit 332 performs morphological analysis on the acquired positive example sentences.

[0318] (Step S2002) The negative example sentence acquiring unit 332 assigns 1 to a counter i.

[0319] (Step S2003) The negative example sentence acquiring unit 332 determines whether the i-th morpheme exists. If the i-th morpheme exists, the process proceeds to step S2004; if not, the process returns to the upper process.

[0320] (Step S2004) The negative example sentence acquiring unit 332 acquires the i-th morpheme.

[0321] (Step S2005) The negative example sentence acquiring unit 332 assigns 1 to a counter j.

[0322] (Step S2006) The negative example sentence acquiring unit 332 acquires the j-th homonym corresponding to the i-th morpheme. If the j-th homonym is acquired, the process proceeds to step S2007; if not, the process proceeds to step S2010.

[0323] For example, the negative example sentence acquiring unit 332 acquires a homonym paired with the i-th morpheme from a homonym dictionary. For example, the negative example sentence acquiring unit 332 acquires a kanji that has the same pronunciation as an arbitrary kanji included in the i-th morpheme from a kanji dictionary, and acquires a homonym in which the arbitrary kanji is replaced with the acquired kanji.

[0324] (Step S2007) The negative example sentence acquiring unit 332 acquires a negative example sentence in which the i-th morpheme of the positive example sentence is replaced with the j-th homonym.

[0325] (Step S2008) The negative example sentence acquiring unit 332 creates training data including the positive example sentences, the negative example sentences acquired in step S2007, and the type identifier "homonym", and stores the training data in a buffer (not shown).

[0326] (Step S2009) The counter j is incremented by 1. The process returns to step S2006.

[0327] (Step S2010) The negative example sentence acquiring unit 332 assigns 1 to a counter k.

[0328] (Step S2011) The negative example sentence acquiring unit 332 determines whether the kth character is present in the i-th morpheme. If the kth character is present, the process proceeds to step S2012, and if the kth character is not present, the process proceeds to step S2023.

[0329] (Step S2012) The negative example sentence acquiring unit 332 assigns 1 to a counter l.

[0330] (Step S2013) The negative example sentence acquisition unit 332 acquires the lth typo for the kth character in the ith morpheme. If the jth typo can be acquired, the process proceeds to step S2014; if not, the process proceeds to step S2017.

[0331] The negative example sentence acquiring unit 332 acquires, for example, one or more typo characters that are paired with the k-th character from the Typo dictionary.

[0332] For example, the negative example sentence acquiring unit 332 acquires the first consonant of the k-th character from a Roman alphabet dictionary, and acquires two characters by adding the consonant before the k-th character.

[0333] (Step S2014) The negative example sentence acquiring unit 332 acquires a negative example sentence in which the kth character in the ith morpheme of the positive example sentence is replaced with the jth typo.

[0334] (Step S2015) The negative example sentence acquiring unit 332 creates training data including the positive example sentence, the negative example sentence acquired in step S2014, and the type identifier "Typo", and stores the training data in a buffer (not shown).

[0335] (Step S2016) The counter l is incremented by 1. The process returns to step S2013.

[0336] (Step S2017) The negative example sentence acquiring unit 332 assigns 1 to a counter m.

[0337] (Step S2018) The negative example sentence acquiring unit 332 acquires the mth pseudo character for the kth character in the i-th morpheme. If the mth pseudo character is acquired, the process proceeds to step S2019; if not, the process proceeds to step S2022.

[0338] The negative example sentence acquiring unit 332 acquires, for example, a pseudo character that pairs with the k-th character from the similar character dictionary.

[0339] (Step S2019) The negative example sentence acquiring unit 332 acquires a negative example sentence in which the k-th character in the i-th morpheme of the positive example sentence is replaced with the m-th pseudo character.

[0340] (Step S2020) The negative example sentence acquiring unit 332 creates training data including the positive example sentence, the negative example sentence acquired in step S2019, and the type identifier "pseudo character," and stores the training data in a buffer (not shown).

[0341] (Step S2021) The counter m is incremented by 1. The process returns to step S2018.

[0342] (Step S2022) The counter k is incremented by 1. The process returns to step S2011.

[0343] (Step S2023) The counter i is incremented by 1. The process returns to step S2003.

[0344] A specific example of the operation of the learning device 3 in this embodiment will be described below.

[0345] It is assumed that the correct example sentence "I filled out the form" is currently stored in the correct example sentence storage unit 311 of the learning device 3.

[0346] Furthermore, the pattern information storage unit 312 of the learning device 3 stores a homonym dictionary associated with the pattern identifier "homonym." The homonym dictionary also includes a set of correct terms and homonyms (Documents: Miscellaneous) (Entry: Entry).

[0347] The pattern information storage unit 312 also stores a Typo dictionary associated with the pattern identifier "Typo." The Typo dictionary also includes a set of one or more correct characters and typos.

[0348] Furthermore, a similar character dictionary associated with the pattern identifier "pseudo character" is stored in the pattern information storage unit 312. The similar character dictionary also includes a set (ma:me) of correct characters and similar characters.

[0349] In this situation, it is assumed that the user inputs a learning instruction to the learning device 3. Then, the learning receiving unit 32 of the learning device 3 receives the learning instruction.

[0350] Next, the positive example sentence acquiring unit 331 acquires the positive example sentence “I filled out the form” from the positive example sentence storage unit 311 .

[0351] Next, the negative example sentence acquisition unit 332 acquires the homonym "shorui" (various categories) for the term "shorin" (document) included in the positive example sentence "I filled out the document" from the homonym dictionary, according to the flowchart in FIG. 20. Then, the negative example sentence acquisition unit 332 acquires the negative example sentence "I filled out the document" by replacing "shorin" (document) in the positive example sentence with the homonym "shorui." The negative example sentence acquisition unit 332 then acquires training data having the positive example sentence, the negative example sentence, and the type identifier "homonym," and stores the data in a buffer (not shown). This training data is the record with "ID=1" in the training data management table in FIG. 21.

[0352] Furthermore, the negative example sentence acquisition unit 332 acquires the homonym "kairyu" (return) for the term "kanji" (entry) included in the positive example sentence "kanji" (entry into the document) from the homonym dictionary, according to the flowchart of FIG. 20. Then, the negative example sentence acquisition unit 332 acquires the negative example sentence "kanji" (entry into the document) by replacing "kanji" in the positive example sentence with the homonym "kairyu." The negative example sentence acquisition unit 332 then acquires training data having the positive example sentence, the negative example sentence, and the type identifier "homonym," and stores the data in a buffer (not shown). This training data is the record with "ID=2" in the training data management table of FIG. 21.

[0353] 20, the negative example sentence acquisition unit 332 acquires a misspelling "mma" for the character "ma" included in the positive example sentence "I filled out the document" from the Typo dictionary. Then, the negative example sentence acquisition unit 332 acquires the negative example sentence "I filled out the document" by replacing the "ma" in the positive example sentence with the misspelling "mma". Then, the negative example sentence acquisition unit 332 acquires training data having the positive example sentence, the negative example sentence, and the type identifier "Typo", and stores it in a buffer not shown. This training data is the record with "ID=3" in the training data management table of FIG. 21.

[0354] 20, the negative example sentence acquiring unit 332 acquires the pseudo-character "me" corresponding to the character "ma" included in the positive example sentence "I filled out the document" from the pseudo-character dictionary.The negative example sentence acquiring unit 332 then acquires the negative example sentence "I showed you how to fill out the document" by replacing the "ma" in the positive example sentence with the pseudo-character "me."The negative example sentence acquiring unit 332 then acquires training data having the positive example sentence, the negative example sentence, and the type identifier "pseudo-character," and stores the data in a buffer (not shown).This training data is the record with "ID=4" in the training data management table of FIG.

[0355] It is assumed that the above processing is performed on other positive example sentences, and the negative example sentence acquiring unit 332 acquires a huge amount of training data.

[0356] Next, the learning processing unit 33 performs a learning process using BERT on two or more pieces of training data in Fig. 21 to acquire a learning device. Then, the accumulation unit 334 accumulates the acquired learning device in the learning device storage unit 113 of the inspection device 1. Note that this learning device is a learning device used by the inspection device 1 to detect typos and omissions.

[0357] Furthermore, when the inspection device 1 receives, for example, the inspection information "... Written in the various categories...", the inspection unit 132 performs machine learning prediction processing using the sentence "Written in the various categories" of the inspection information and the learning device, and obtains the inspection result shown in 2201 of Fig. 22. The inspection result here includes information indicating the probability of each type identifier for each character.

[0358] If the probability is equal to or greater than a threshold value (here, 0.50), the inspection unit 132 determines that the character is an error identified by the type identifier corresponding to the probability. In other words, the inspection unit 132 determines that the character "" with a probability of "0.86" is an error corresponding to the type identifier "homonym."

[0359] In addition, in Figure 22, for example, the inspection unit 132 uses the Fully Connected Layer of the BERT-based learning device 2202 to provide the sentence ``I filled out the various categories'' to a module that performs prediction processing using a deep learning algorithm, executes the module, and obtains the inspection result 2201.

[0360] Then, the test result construction unit 133 constructs a test result indicating that the character "various" in the test information "····. Written in various categories···" is an error corresponding to the type identifier "homonym."

[0361] Next, the result output unit 141 outputs the test results.

[0362] As described above, according to this embodiment, it is possible to configure a learning device that checks for typos, omissions, etc. Furthermore, according to this embodiment, it is possible to check for typos, omissions, etc. using the learning device.

[0363] The software that realizes the learning device 3 in this embodiment is the following program. That is, this program includes a computer that can access a pattern information storage unit that stores one or more pattern information related to error patterns, a positive example sentence acquisition unit that acquires one or more positive example sentences that are correct sentences, a negative example sentence acquisition unit that acquires one or more negative example sentences that are sentences containing errors, for each of the one or more positive example sentences, using the one or more pattern information, and The program performs machine learning learning processing on two or more teacher data sets each having the positive example sentence and the one or more negative example sentences, and functions as a learning unit that acquires a learning device and a storage unit that stores the learning device.

[0364] 23 shows the appearance of a computer that executes the programs described herein to realize the inspection device 1 and the like according to the various embodiments described above. The above-described embodiments can be realized by computer hardware and a computer program executed thereon. FIG. 23 is an overview of this computer system 300, and FIG. 24 is a block diagram of the system 300.

[0365] In FIG. 23, a computer system 300 includes a computer 301 including a CD-ROM drive, a keyboard 302, a mouse 303, and a monitor 304.

[0366] 24, computer 301 includes, in addition to CD-ROM drive 3012, MPU 3013, bus 3014 connected to CD-ROM drive 3012 etc., ROM 3015 for storing programs such as a boot-up program, RAM 3016 connected to MPU 3013 for temporarily storing instructions of application programs and providing temporary storage space, and hard disk 3017 for storing application programs, system programs, and data. Although not shown here, computer 301 may further include a network card for providing connection to a LAN.

[0367] A program that causes the computer system 300 to execute the functions of the inspection device 1 and the like of the above-described embodiment may be stored on a CD-ROM 3101, inserted into the CD-ROM drive 3012, and then transferred to the hard disk 3017. Alternatively, the program may be transmitted to the computer 301 via a network (not shown) and stored on the hard disk 3017. The program is loaded into the RAM 3016 when executed. The program may also be loaded directly from the CD-ROM 3101 or the network.

[0368] The program does not necessarily include an operating system (OS) or a third-party program that causes the computer 301 to execute the functions of the inspection device 1 of the above-described embodiment. The program only needs to include instructions that call appropriate functions (modules) in a controlled manner to achieve the desired results. How the computer system 300 operates is well known, and a detailed description thereof will be omitted.

[0369] In addition, in the above program, the steps of transmitting information and receiving information do not include processing performed by hardware, such as processing performed by a modem or interface card in the transmission step (processing that can only be performed by hardware).

[0370] The computer that executes the program may be a single computer or a plurality of computers, that is, it may perform centralized processing or distributed processing.

[0371] Furthermore, in each of the above embodiments, it goes without saying that two or more communication means present in one device may be physically realized by one medium.

[0372] Furthermore, in each of the above embodiments, each process may be realized by centralized processing in a single device, or may be realized by distributed processing in a plurality of devices.

[0373] The present invention is not limited to the above-described embodiment, and various modifications are possible, and it goes without saying that these modifications are also included within the scope of the present invention. [Industrial Applicability]

[0374] As described above, the inspection device according to the present invention has the effect of being able to inspect text based on inspection rules for each user, and is useful as an inspection device, etc. [Explanation of symbols]

[0375] 1. Inspection equipment 2. Terminal Device 3 Learning Device 11 Storage area 12 Reception 13 Processing section 14 Output section 21 Terminal storage section 22 Terminal Reception 23 Terminal processing section 24 Terminal transmitter 25 Terminal receiving section 26 Terminal Output Unit 31 Learning storage section 32 Learning Reception Department 33 Learning processing unit 111 User information storage section 112 Rule Storage 113 Learning Unit Storage 121 Test Information Reception Department 122 Selection Reception Department 131 Rule Acquisition Department 132 Inspection Department 133 Test result composition section 134 Selection interface configuration section 141 Result output section 142 Selective interface output section 311 Correct sentence storage section 312 Pattern information storage section 331 Positive Sentence Acquisition Unit 332 Negative Example Sentence Acquisition Unit 333 Learning Department 334 Storage Unit 1301 positive images 1321 Morphological analysis means 1322 Comparison string acquisition method 1323 Difference degree information acquisition means 1324 Judgment means 1325 Prediction Methods 1326 Image Judgment Method

Claims

1. an examination information receiving unit that receives examination information including text; a rule acquisition unit that acquires one or more inspection rules from a rule storage unit that stores one or more inspection rules for inspecting a sentence, the inspection rules including an inspection rule that a term 1 and a term 2 exist within N words (N is a natural number) in the sentence, or that a term 2 exists within N words (N is a natural number) after a term 1; an inspection unit that performs inspection on the inspection information using the inspection rule; an inspection result configuration unit that configures inspection results to be output using the inspection results of the inspection unit; and a result output unit that outputs the inspection results.

2. The inspection rule is:

2. The inspection device according to claim 1, wherein the rules are for inspecting exaggerated expressions or for inspecting expressions that may be misleading.

3. 3. The inspection device according to claim 2, wherein the term 1 is "absolute" or "complete."

4. The test results are location information for identifying the location of an error in the inspection information and a type identifier for identifying the type of the error; a selection interface configuration unit that configures a selection interface that has one or more type identifiers obtained by uniquely processing the type identifiers included in the test results as selection items and does not have type identifiers not included in the test results as selection items; a selection interface output unit that outputs the selection interface; a selection receiving unit that receives a selection for a selection item included in the selection interface; The test result configuration unit The inspection device of claim 1 , wherein errors corresponding to one or more location information paired with a type identifier corresponding to the selected item for the selection are compared with errors corresponding to other location information to produce visually different inspection results.

5. The test results are location information for identifying the location of an error in the inspection information and a type identifier for identifying the type of the error; The test result configuration unit The inspection device according to claim 1 , wherein the inspection result is generated in a manner that visually indicates the location information and the type identifier.

6. An inspection method for causing a computer to execute all of the processes performed by the inspection apparatus according to any one of claims 1 to 5.

7. Computer, A program for causing the inspection device according to any one of claims 1 to 5 to function as the inspection device.

Images