Text consistency confirmation system, text consistency confirmation device, text consistency confirmation method and text consistency confirmation program

Abstract

A text consistency confirmation system, a text consistency confirmation device, a text consistency confirmation method and a text consistency confirmation program are provided. A business terminal sends a file including text data for text consistency confirmation (step S404). The text consistency confirmation device subdivides the received text data (step S407), and sends publishing data including the subdivided text data to the user terminal (step S408). The user terminal performs a game of text consistency confirmation on the subdivided text data included in the publishing data, and sends the game result to the text consistency confirmation device (step S409). The text consistency confirmation device sends a result file including the text consistency confirmation result based on the received game result to the business terminal (step S418).

Description

Technical Field

[0001] The present invention relates to a text consistency confirmation system, a text consistency confirmation device, a text consistency confirmation method and a text consistency confirmation program using game applications. Background Art

[0002] At present, as technologies for digitizing paper documents, OCR (Optical Character Recognition) and ICR (Intelligent Character Recognition) are widely known. For example, in Patent Document 1, in order to distinguish the difference between printed characters in a plurality of document images obtained by OCR, it is determined which character image the input character image is close to by referring to the character pattern data, and the corresponding character code is obtained.

[0003] Prior art literature

[0004] Patent Literature

[0005] Patent Document 1: Japanese Patent Application Publication No. 2013-90262 Summary of the invention

[0006] Problems to be solved by the invention

[0007] The processing in Patent Document 1 is based on the premise that the text is correctly read. However, although the technology for electronicizing paper documents has improved, misreading and unreadable characters occur at a fixed rate, and there is no technology that can completely recognize all characters. Therefore, it is ultimately necessary to confirm the consistency of the text through human visual observation. At this time, in order to achieve cost reduction, the confirmation of text consistency is sometimes performed through external entrustment. However, when performing external entrustment, since personal information is provided to the outside, there is a risk of information leakage.

[0008] Therefore, an object of the present invention is to provide a text consistency confirmation system or the like that can ensure safety when visual confirmation of text consistency is performed by external outsourcing.

[0009] Solutions to Solve Problems

[0010] The text consistency confirmation system comprises: a business terminal operated by a client of text consistency confirmation; a text consistency confirmation device for controlling text consistency confirmation; and a user terminal for performing text consistency confirmation using a game application. The business terminal sends a file including text data to be confirmed for text consistency. The text consistency confirmation device subdivides the received text data and sends publishing data including the subdivided text data to the user terminal. The user terminal performs a game for text consistency confirmation on the subdivided text data included in the publishing data and sends a game result to the text consistency confirmation device. The text consistency confirmation device sends a result file including a text consistency confirmation result based on the received game result to the business terminal.

[0011] Effects of the Invention

[0012] Among the inventions disclosed in this application, the effects obtained by representative inventions are briefly described below.

[0013] That is, according to the representative embodiment of the present invention, it is possible to ensure safety when visual confirmation of text consistency is performed by external outsourcing. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 This is an overall configuration diagram showing an example of a character consistency checking system according to an embodiment of the present invention.

[0015] Figure 2 This is a block diagram showing a configuration example of each device in the character consistency checking system.

[0016] Figure 3 This is a diagram showing the configuration of an example of a user host.

[0017] Figure 4 This is a diagram showing an example of the structure of an enterprise host.

[0018] Figure 5 It is a structural diagram showing an example of a character DB.

[0019] Figure 6 It is a structural diagram showing an example of the evaluation information DB.

[0020] Figure 7 This is a diagram showing an example of the structure of the payment history DB.

[0021] Figure 8 It is a structural diagram showing an example of the confirmation result DB.

[0022] Fig. 9 It is a structural diagram showing an example of a segmented character DB.

[0023] Fig.10This is a diagram showing an example of the structure of the delivery character DB.

[0024] Fig.11 It is a structural diagram showing an example of a game result DB.

[0025] Fig.12 It is a diagram showing an example of the structure of the AI ​​determination result DB.

[0026] Fig.13 It is a diagram showing an example of the structure of the majority voting result DB.

[0027] Fig.14 This is a flowchart showing an example of a character consistency confirmation process according to one embodiment of the present invention.

[0028] Fig.15 This is a flowchart showing a specific example of user registration processing.

[0029] Fig.16 This is a flowchart showing an example of character data acceptance / recognition processing.

[0030] Fig.17 This is a flowchart showing an example of the publishing process.

[0031] Fig.18 This is a diagram showing an example of a character consistency confirmation game.

[0032] Fig.19 This is a diagram showing another example of the character consistency confirmation game.

[0033] Fig. 20 This is a diagram showing another example of the character consistency confirmation game.

[0034] Fig.21 This is a flowchart showing an example of result collection processing.

[0035] Fig. 22 This is a flowchart showing an example of majority voting processing and AI determination result comparison processing.

[0036] Fig.23 This is a diagram showing an example of establishing a correspondence between the inability reason code and the inability reason.

[0037] Fig.24 This is a flowchart showing an example of evaluation / payment amount calculation processing.

[0038] Fig.25 This is a flowchart showing an example of result transmission processing. DETAILED DESCRIPTION

[0039] (One embodiment)

[0040] Below, the best mode for implementing the present invention is described in detail according to the accompanying drawings. In addition, in each of the accompanying drawings for describing the best mode for implementing the invention, the same reference numerals are used for components with the same functions, and repeated description thereof is omitted.

[0041] <Composition of the text consistency confirmation system>

[0042] Figure 1 FIG. 1 is a diagram showing the overall structure of an example of a text consistency confirmation system according to an embodiment of the present invention. Figure 1 As shown, the text consistency confirmation system 1 includes a text consistency confirmation device 10, a business terminal 20, a user terminal 30, and a network N. In the text consistency confirmation system 1, communication including data transmission and reception is performed between the text consistency confirmation device 10 and the business terminal 20 and between the text consistency confirmation device 10 and the user terminal 30 via the network N. Furthermore, communication can also be performed between a plurality of user terminals 30 via the network N depending on the usage situation.

[0043] The text consistency checking device 10 is a device for controlling text consistency checking using a game application. When there is a text consistency checking request from a service terminal 20, the text consistency checking device 10 issues text data for text consistency checking to a user terminal 30 of a registered user who has installed a dedicated game application.

[0044] The business terminal 20 is a processing device for processing character consistency confirmation requested by a client such as a company. The user terminal 30 is a processing device for receiving published data including character data used for playing a game from the character consistency confirmation device 10 and transmitting a game result.

[0045] The text consistency checking device 10 and the service terminal 20 are information processing devices such as PC terminals and server terminals that can communicate via a network. The user terminal 30 is an information processing device such as a smart phone terminal, a mobile phone terminal, a PDA terminal, a tablet terminal, a PC terminal, etc.

[0046] 《Text consistency confirmation device 10》

[0047] Figure 2 FIG. 1 is a block diagram showing an example of the configuration of each device of the text consistency confirmation system. Figure 2As shown, the text consistency confirmation device 10 includes a control unit 11, a game content unit 12, a storage unit 13, a communication unit 14, etc. The control unit 11 performs various processes related to text consistency confirmation as follows: receiving a file including text data for which text consistency confirmation is requested from a client; publishing data including text data to a game user and receiving a game result of text consistency recognition from the game user; sending a result file including text consistency confirmation results to the client; etc.

[0048] The control unit 11 has a text data acceptance / recognition process 111, a segmentation / randomization process 112, a user registration process 113, a publishing process 114, a result collection process 115, a majority voting process 116, an evaluation / payment amount calculation process 117, a restoration process 118, and a result sending process 119.

[0049] The control unit 11 includes a processor such as a CPU, and the character consistency checking program is executed in the processor to realize each functional block of the control unit 11. In addition, each functional block of the control unit 11 may be realized by cooperation between software and hardware other than the processor.

[0050] The text data acceptance / recognition processing 111 receives a document including text data for text consistency confirmation sent by the client via the business terminal 20, and performs text recognition of the text data using OCR, ICR, AI, etc.

[0051] The segmentation / randomization process 112 segments the received character data and performs randomization while ensuring the security of the user terminal 30. Specifically, the segmentation / randomization process 112 segments and randomly distributes characters recognized from character data of multiple clients, thereby eliminating the association with the original paper document.

[0052] The user registration process 113 receives registration information required for a user who owns the user terminal 30 to play a game application and registers the user as a game user.

[0053] The publishing process 114 matches the text data segmented / randomized by the segmentation / randomization process 112 to an appropriate game user from the registered game users, and publishes the text data to the matched game user. For example, the publishing process 114 matches the game user to be published by referring to the attribute information of the registered game users.

[0054] The result collection process 115 collects the game results of the game application played by the game user. The majority vote process 116 performs a majority vote process on the collected results and determines the consistency of the text.

[0055] The evaluation / payment amount calculation processing 117 evaluates and ranks each game user based on the collected game results, and calculates the reward corresponding to the evaluation and rank of the game user. The restoration processing 118 performs the restoration of the subdivided and randomized text data to confirm the consistency of the text. The result transmission processing 119 transmits the result file including the confirmation result of the text consistency to the service terminal 20.

[0056] The game content unit 12 is a processing unit that performs operation management of the game application on the server side. The game content unit 12 is operated and executed according to a well-known or existing game technology.

[0057] The storage unit 13 is a functional block that stores various information required for the operation of the text consistency confirmation system 1. The storage unit 13 has storage areas corresponding to the user host 131, the enterprise host 132, the text DB 133, the subdivided text DB 137, the release text DB 138, the evaluation information DB 134, the payment history DB 135, the confirmation result DB 136, the game result DB 140, the AI ​​judgment result DB 141, and the majority voting result DB 142, respectively.

[0058] The user host 131 stores the attributes of the registered game users. Figure 3 FIG. 1 is a diagram showing an example of a user host. Figure 3 As shown, the user host 131 has storage areas corresponding to, for example, user ID 131A, registration date 131B, user name 131C, nationality 131D, gender 131E, account information 131F, age 131G, language 131H, 131I, etc. Each attribute information stored in the user host 131 is used to match the game user who is the delivery destination of the delivery data.

[0059] The user ID 131A stores a unique user ID assigned when the game user registers. The registration date 131B stores the registration time of the game user. The time also includes information about the date. The user name 131C stores the name of the game user. The nationality 131D stores the nationality of the game user. The gender 131E stores the gender of the game user. The account information 131F stores the transfer destination account information for paying rewards. The age 131G stores the age of the game user. The languages ​​131H and 131I store the names of the available languages ​​that the game user can use.

[0060] The company host 132 stores company information of a company (client) that owns the business terminal 20 and requests consistency check of text. Figure 4 FIG. 1 is a diagram showing an example of the structure of an enterprise host. Figure 4As shown, the enterprise host 132 has storage areas corresponding to the enterprise ID 132A, the enterprise name 132B, the address 132C, the responsible department 132D, and the like.

[0061] The company ID 132A stores the company ID unique to the company (client) assigned at the time of registration. The company name 132B stores the registered company name corresponding to the company ID. The address 132C stores the address of the registered company. The responsible department 132D stores the responsible department of the registered company using the business terminal 20.

[0062] The character DB 133 stores information such as character data received from the service terminal 20 . Figure 5 : is a diagram showing an example of a text DB. Figure 5 As shown, the character DB 133 has storage areas corresponding to the company ID 133A, the acceptance date 133B, the document number 133C, the language 133D, and the like.

[0063] The enterprise ID 133A stores the enterprise ID of the enterprise that sent the text data. The acceptance date 133B stores the time when the text data was received. The time also includes the date information. The file number 133C stores the file number for identifying the received text data. The file number can be a file number assigned in advance in the business terminal 20 on the sending side, or a file number assigned in the text consistency confirmation device 10. The language 133D stores the language used in the text data.

[0064] The evaluation information DB 134 stores evaluation results of game users. Figure 6 : is a diagram showing an example of the structure of the evaluation information DB. Figure 6 As shown, the evaluation information DB 134 has storage areas corresponding to the user ID 134A, the evaluation date 134B, the number of answers 134C, the correct answer rate 134D, the level 134E, and the like.

[0065] The user ID 134A stores the user ID of the game user for identifying the evaluation object. The evaluation date 134B stores the time when the evaluation is performed. The time also includes the date information. The number of answers 134C stores the number of answers from the previous evaluation to the current evaluation. The correct answer rate 134D stores the correct answer rate specified by the ratio of the number of correct answers to the number of answers. The correct answer rate can be the total correct answer rate since registration, or the correct answer rate from the previous evaluation to the current evaluation. The level 134E stores the level of the game user.

[0066] The payment history DB 135 stores the payment history of rewards for game users. Figure 7 : is a diagram showing an example of a payment history DB. Figure 7As shown, the payment history DB 135 has storage areas corresponding to the user ID 135A, the payment date 135B, the payment amount 135C, the accumulated amount 135D, and the like.

[0067] The user ID 135A stores the user ID of the game user who paid the reward. The payment date 135B stores the date on which the reward was paid. Multiple dates can also be stored in the payment date 135B. The payment amount 135C stores the amount of the reward paid. The cumulative amount 135D stores the cumulative amount of the reward paid.

[0068] The confirmation result DB 136 stores the character consistency confirmation results of the character data and the like. Figure 8 : is a diagram showing an example of the structure of the confirmation result DB. Figure 8 As shown, the confirmation result DB 136 has storage areas corresponding to the enterprise ID 136A, the file number 136B, the result file management number 136C, the result determination date 136D, the result report date 136E, the transmission FLG 136F, the text ID 136G, the failure reason code 136H, and the like.

[0069] The enterprise ID 136A stores the enterprise ID of the destination enterprise of the result file including the text consistency confirmation result. The file number 136B stores the file number of the original file for which the text consistency confirmation has been performed. The result file management number 136C stores the management number of the result file including the text consistency confirmation result sent to the business terminal 20. The management number of the result file is assigned in the text consistency confirmation device 10. The result determination date 136D stores the date on which the result determination is performed. The result report date 136E stores the report date on which the text consistency confirmation result is reported to the client's enterprise. Specifically, the result report date refers to the date on which the result file is sent.

[0070] The sending FLG136F stores a sending flag indicating that a result report has been made to the enterprise. The sending FLG136F may be composed of, for example, a register. The text ID136G may store, for example, the ID of text data with a low correct answer rate in the text consistency confirmation result. Alternatively, the text ID136G may also store the ID of text data that is inconsistent between the arrangement based on the majority voting process described later and the arrangement based on AI. The inability reason code 136H stores the inability reason code corresponding to various inability reasons when the text consistency cannot be confirmed.

[0071] The segmented character DB 137 stores character data segmented and randomized by the segmentation / randomization process 112 . Fig. 9 : is a diagram showing an example of a subdivided text DB. Fig. 9As shown, the subdivided character DB 137 has storage areas corresponding to the subdivided character number 137A, the company ID 137B, the file number 137C, the restoration code 137D, the character image 137E, and the like.

[0072] The subdivided character number 137A stores the number assigned to the subdivided character data. The company ID 137B stores the company ID of the company that sent the file including the character data for character consistency confirmation. The file number 137C stores the ID assigned to the file including the character data received from the business terminal 20. The restoration code 137D stores the code when restoring the subdivided character. The character image 137E stores the image (Image) including the subdivided character.

[0073] The distribution text DB 138 stores information in which the segmented text and / or image are associated with the matched game user. Fig.10 1 is a diagram showing an example of the structure of the release text DB 138. Fig.10 As shown, the posting character DB 138 has storage areas for user ID 138A, subdivided character number 138B, character image 138C, and the like.

[0074] The user ID 138A stores the user ID of the matched game user. The subdivided character number 138B stores the number of the posting character posted to the matched game user. The character image 138C stores an image including the character corresponding to the subdivided character number.

[0075] The game result DB 140 stores the game results for which the text consistency is confirmed according to the game type and the game scene. Fig.11 : is a diagram showing an example of a game result DB. Fig.11 As shown, the game result DB 140 has storage areas corresponding to a game ID 140A, a game progress ID 140B, a start date 140C, an end date 140D, a required number of answers 140E, a number of answers 140F, and the like.

[0076] Game ID 140A stores the ID of each game. For example, text consistency confirmation is performed for one or more text data in each game. Game progress ID 140B stores the game progress of each game. Start date 140C stores the start date of each game. End date 140D stores the end date of each game. The release end date is a pre-set end date, but for example, when the required number of answers are obtained before the set release end date and the release is ended earlier than scheduled, it is rewritten as a new release end date. The start date and end date of each game are, for example, the release start date and the release end date. The required number of answers 140E stores the number of answers required for majority voting processing. The number of answers 140F stores the actual number of answers for each game.

[0077] The AI ​​determination result DB 141 stores the character consistency determination results based on AI. Fig.12 : is a diagram showing an example of the structure of the AI ​​determination result DB. Fig.12 As shown, the AI ​​determination result DB 141 has storage areas corresponding to a company ID 141A, a file number 141B, a result file management number 141C, a result determination date 141D, a character ID 141G, a sequence 141H, and the like.

[0078] The enterprise ID 141A stores the enterprise ID of the enterprise that has entrusted the text consistency confirmation. The file number 141B stores the file number of the file including the text data that has been confirmed based on AI. The result file management number 141C stores the management number of the result file including the result of the confirmation of the text consistency based on AI. The result determination date 141D stores the date when the text consistency confirmation based on AI was performed. The text ID 141G stores the text ID of the candidate text used in the text consistency confirmation based on AI. The order 141H stores the order of each candidate text in the arrangement generated by AI.

[0079] The majority vote result DB 142 stores the character consistency confirmation results of the majority vote process based on the game results. Fig.13 : is a diagram showing an example of the structure of the majority voting result DB. Fig.13 As shown, the majority voting result DB 142 has storage areas corresponding to the enterprise ID 142A, the file number 142B, the result file management number 142C, the result determination date 142D, the character ID 142G, the sequence 142H, and the like.

[0080] The enterprise ID 142A stores the enterprise ID of the enterprise that has entrusted the text consistency confirmation. The file number 142B stores the file number of the file including the text data that has been confirmed by the majority voting process. The result file management number 142C stores the management number of the result file including the confirmation result of the text consistency based on the majority voting process. The result judgment date 142D stores the date when the text consistency confirmation based on the majority voting process is performed. The text ID 142G stores the text ID of the candidate text used when confirming the text consistency based on the majority voting process. The order 142H stores the order of each candidate text in the arrangement generated by the majority voting process.

[0081] The communication unit 14 is a communication interface that communicates with the service terminal 20 and the user terminal 30 according to a command from the control unit 11 or the like.

[0082] 《Business Terminal 20》

[0083] The service terminal 20 is as follows Figure 2 The figure shows a control unit 21 , a storage unit 22 , a communication unit 23 , etc. The control unit 21 has a scanning process 211 , a character data transmission process 212 , and a result reception process 213 .

[0084] The scanning process 211 acquires text data for character recognition of paper documents using OCR, ICR, or AI. The acquired text data also includes text data of handwritten text. The text data transmission process 212 transmits a file including the text data acquired in the scanning process 211 to the text consistency verification device 10. The result reception process 213 receives a result file including the text consistency verification result from the text consistency verification device 10.

[0085] The storage unit 22 has storage areas corresponding to the character data DB 221 and the confirmed data DB 222. The character data DB 221 stores the character data acquired in the scanning process 211. The confirmed data DB 222 stores the character consistency check result included in the result file received from the character consistency check device 10.

[0086] The communication unit 23 is a communication interface for communicating with the character consistency checking device 10 in accordance with a command from the control unit 21 or the like.

[0087] 《User terminal 30》

[0088] The user terminal 30 includes a game application 31, a communication unit 32, etc. The game application 31 is a program downloaded and installed from the text consistency checking device 10. The communication unit 32 is a communication interface for communicating with the text consistency checking device 10 in accordance with a command from a control unit or the like in the user terminal 30 (not shown).

[0089] in addition, Figure 2 The structure is an example of the function for realizing the present invention, and the present invention is not limited to such a structure. For example, a part of the processing performed by the text consistency confirmation device 10 may be performed by the user terminal 30, or text data for text consistency confirmation may be distributed to a PC terminal used by an operator who performs professional text consistency confirmation, etc., not limited to game users, and executed as a text consistency confirmation service.

[0090] 《Text consistency confirmation processing》

[0091] Next, the text consistency confirmation process related to this embodiment is described. Fig.14 FIG. 1 is a flowchart showing an example of text consistency confirmation processing according to an embodiment of the present invention. Fig.14 In the figure, the text consistency checking device 10 is centered, and the business terminal 20 and the user terminal 30 are shown on the left and right. Fig.14 The text consistency confirmation process includes steps S401 to S419.

[0092] First, the text consistency confirmation device 10 discloses / provides a dedicated game application for checking text consistency through a game. Game users can freely download the game application (step S401). The environment for providing the game application can be inside the text consistency confirmation device 10 (such as the game content unit 12, etc.), and can also be installed in other download-only environments.

[0093] The game user holding the game terminal 30 downloads and installs the game application, and then enters necessary items to apply for game user registration (step S402).

[0094] 《User registration process》

[0095] The text consistency checking device 10 examines the application content from the new user, stores the user who meets the predetermined registration conditions in the user host 131, and registers the user (step S403). If the registration conditions are not met, the text consistency checking device 10 sends a rejection to the user terminal 30 and notifies that the registration conditions are not met (not shown).

[0096] Here, a specific example of the user registration process will be described. Fig.15 is a flowchart showing a specific example of user registration processing. Fig.15The user registration process includes steps S70 to S76. In step S70, it is determined whether there is a user registration application. When the user terminal 30 applies for registration of the game user in step S70 (yes), the user registration process 113 of the control unit 11 confirms the information of the application and examines the registration qualification (step S71). When the user terminal 30 does not apply for registration of the game user (no), the process returns to step S70, and the user registration process 113 waits until there is a registration application from the user.

[0097] In step S72, it is determined whether the application items are incomplete. If the result of the confirmation and review of the application items is that there are incomplete information or return items (yes), the user registration process 113 notifies the user terminal 30 of the situation (step S73) and returns to step S70. On the other hand, if the application items are not incomplete (no), the user registration process 113 assigns a unique identifiable user ID to the user (step S74). The user registration process 113 registers the user in the user host 131 (step S75) and notifies the user terminal 30 that the user registration is complete (step S76).

[0098] Return to Fig.14 The scanning process 211 of the business terminal 20 scans the paper document for which the text consistency confirmation is requested and obtains the text data, and stores the obtained text data in the text data DB 221. Furthermore, the text data sending process 212 sends the obtained text data to the text consistency confirmation device 10 (step S404). In addition, the timing of step S404 can be before step S403 or at the same time as step S403.

[0099] 《Text data acceptance / recognition processing》

[0100] The text consistency checking device 10 receives a file including text data from the business terminal 20. The text data receiving / recognition processing 111 checks the information of the enterprise that sent the file with the enterprise information registered in the enterprise host 132, and then associates the received text data with the enterprise ID and the like and stores them in the text DB 133 (step S405). The text consistency checking device 10 performs text recognition of the text data stored in the text DB 133 using existing well-known techniques (step S406).

[0101] Here, a specific example of text data acceptance / recognition processing is described. Fig.16 1 is a flowchart showing an example of text data acceptance / recognition processing. Fig.16The text data receiving / recognition process 111 includes steps S50 to S54. In step S50, it is determined whether a request for text consistency confirmation has been made by the enterprise. If no request for text consistency confirmation has been made from the business terminal 20 in step S50 (No), the text data receiving / recognition process 111 returns to step S50 and waits until a request for text consistency confirmation has been made from the enterprise.

[0102] On the other hand, when a request for character consistency check is made from the business terminal 20 (Yes), the character data acceptance / recognition processing 111 collates the company information included in the character data received from the business terminal 20 with the company information stored in the company host 132 (step S51).

[0103] In step S52, it is determined whether the received enterprise information is registered in the enterprise host 132. In the case where the received enterprise information is registered as the client (yes), the text data acceptance / recognition processing 111 uses OCR, ICR, AI, etc. to perform text recognition processing on the received text data. The text data acceptance / recognition processing 111 associates the recognized text with the enterprise information and stores it in the text DB 133 (step S54). On the other hand, in the case where the received enterprise information is not registered as the client (no), the text data acceptance / recognition processing 111 sends a notification urging the retransmission of the text data to the business terminal 20 that sent the text data (step S53).

[0104] Return to Fig.14 Then, the segmentation / randomization processing 112 performs segmentation and randomization of the characters recognized in step S406 to generate data for distribution to the game user (step S407). The segmentation and randomization of the characters use existing known technologies. By segmenting and randomizing the characters, the characters recognized from the same character data are dispersed, so that even if the data for distribution is distributed to the game user, security can be guaranteed.

[0105] Release Processing

[0106] Then, the publishing process 114 analyzes the publishing data generated in step S407. Furthermore, the character consistency checking device 10 refers to the attributes of the registered users stored in the user host 131, matches the registered users suitable for the analysis results of the publishing data, and performs publishing processing of the publishing data for the matched appropriate game users (step S408).

[0107] Specifically, the character consistency checking device 10 sends character data obtained by segmenting the same character data to a plurality of game users. In addition, the character consistency checking device 10 sends the segmented same character data to a plurality of game users.

[0108] Furthermore, for example, when checking the consistency of idiomatic expressions and text with coherence, a game user who can use the language of the text data or a game user who has the nationality of a country where the language is the mother tongue, second language, or third language is used will be matched. Thus, the text consistency check can be performed efficiently.

[0109] In order to ensure security during the publishing process, the segmentation / randomization process 112 adds an identifier to each character of the file during segmentation and randomization. During randomization, for example, the number of character data with the same identifier in the published data is limited to a predetermined number. This makes it difficult to restore the published data and obtain the original character data, so security is guaranteed.

[0110] The publishing process 114 matches the game user who sent the text data with reference to the attributes of each text data and the user host 131. Specifically, the published data is associated with the user IDs of the matched multiple game users and published. In this way, by distributing the segmented data according to predetermined conditions and publishing it to multiple game users, it becomes difficult to restore the original file, thereby ensuring security.

[0111] Here, a specific example of the publishing process is described. Fig.17 is a flowchart showing an example of publishing processing. Fig.17 The step S80 to S86 are included. In step S80, it is determined whether the subdivided character DB is updated as a result of the subdivided / random processing in step S407. When the subdivided character DB is not updated in step S80 (No), the character consistency confirmation device 10 returns to step S80 and waits until the subdivided character DB 137 is updated. On the other hand, when the subdivided character DB is updated (Yes), the character consistency confirmation device 10 refers to the subdivided character to determine the language used by the character (step S81).

[0112] Furthermore, the text consistency checking device 10 determines the difficulty of the text data when performing consistency checking using judgment items such as the number of characters in the text data, accumulated information of past judgment results, idioms, and whether the text data is coherent (step S82). Furthermore, the text consistency checking device 10 matches the game user of the publishing destination using the difficulty and the evaluation result stored in the user host 131 (step S83).

[0113] Supplement the difficulty level of text data. For example, text data similar to text data whose discrimination results based on AI in the past and discrimination results based on the majority voting process described below are judged to be significantly different, text data that is difficult to judge as text such as signatures / attributions, or text data that requires confirmation of text consistency based on combinations with surrounding text, such as the Chinese character "一" for Chinese numerals, the elongated sound "-", the number "1", and the English letter "l", etc., are determined to have a greater difficulty level.

[0114] The text consistency confirmation device 10 associates the text data with multiple game users that match it (step S84), stores it in the publishing text DB 138 (step S85), and publishes it to these multiple game users (step S86).

[0115] Moreover, in these publishing processes, the text consistency confirmation device 10 can also include in the published data the text data obtained by subdividing for text consistency confirmation and the abstracted candidate text that is the comparison object for the text corresponding to this text data. The abstracted candidate text is, for example, a notation and / or an image, etc.

[0116] Furthermore, the text consistency confirmation device 10 can also subdivide a word or signature composed of multiple characters into one block for the received text data. At this time, the text consistency confirmation device 10 matches the game users who send the published data with reference to the attributes of the game users. Specifically, the text consistency confirmation device 10 compares the language used for the block with the available languages registered for the attributes of the game users, and matches the game users whose languages are the same. Or, the text consistency confirmation device 10 can also compare the language used for the block with the nationality registered for the attributes of the game users, and match the game users with the nationality of a country that has the language used for the block as the common language.

[0117] "Execution of the Text Consistency Confirmation Game"

[0118] Here, return to Fig.14 the description. The game user uses the user terminal 30 to play a game based on the published data including the text data published by the text consistency confirmation device 10, thereby performing text consistency confirmation of the text data (step S409).

[0119] [Game Example (1)]

[0120] Explain Game Example (1). Fig.18 is a diagram showing an example of the text consistency confirmation game. First, for Fig.18An example will be described. The game user operates the user terminal 30 in his or her possession to play a game for confirming text consistency. An image 302 based on the text data received from the text consistency confirmation device is displayed on the display screen 301 of the user terminal 30. The images of the characters included in the image 302 are recognized by OCR, ICR, AI, etc. Candidate characters 303 to 305 similar to the characters are displayed on the display screen 301. In Fig.18 "Tian" 303, "You" 304, and "Jia" 305 are respectively displayed as candidate characters.

[0121] In addition, instead of these candidate characters, abstracted candidate characters can be used. Specifically, when the characters for which text consistency is to be confirmed are Chinese characters, marks and / or images obtained by abstracting the aforementioned "Tian", "You", and "Jia" can be used as candidate characters. Thereby, it becomes possible to match game users who do not recognize Chinese characters, and the number of answers can be effectively increased.

[0122] The game user clicks and selects the candidate character determined to be consistent with the character in the image 302, and by clicking the OK button 308, the selected candidate character is determined. When the candidate character is determined, the process proceeds to the next screen (not shown).

[0123] In addition, when it is not possible to determine text consistency based on the displayed candidate characters, the game user clicks the Skip button 306 to proceed to the next screen and can try the next game.

[0124] [Game Example (2)]

[0125] Next, Game Example (2) will be described. Fig.19 is a diagram showing another example of the text consistency confirmation game. On the display screen 301 of the user terminal 30, instead of the character considered to be the most consistent with the character in the image 302, other similar characters are displayed as candidate characters. Specifically, in Fig.19 , instead of "Tian" 303 ( Fig.18 ) considered to be the most consistent with the character in the image 302, the similar character "Shen" 309 is displayed as a candidate character, and the game user determines the text consistency. By incorporating such a configuration, it becomes possible to easily evaluate the correctness and accuracy of the text consistency judgment for the game user.

[0126] [Game Example (3)]

[0127] Next, Game Example (3) will be described. Fig. 20 is a diagram showing another example of the text consistency confirmation game. It is displayed in Figure 18-19In the case where it is determined that there is no text that is identical or similar to the text in the image 302 among the candidate texts in the display screen 301, the game user clicks the input button 307 to enter the display screen 301. Fig. 20 Then, the game user inputs the characters that he thinks are correct by handwriting in the character input field 310. By incorporating such a structure, characters different from the candidate characters can be arbitrarily input.

[0128] These game screens are screens showing an example of basic actions of the game J, and the game content may also include playability and entertainment. For example, the user terminal 30 may communicate with user terminals owned by other game users via the network N to compete in the correctness and speed of answers.

[0129] 《Result Recovery Processing》

[0130] The game result is sent to the character consistency checking device 10 via the network N at any time or every time a predetermined time has passed, and is collected (step S410). Fig.21 is a flowchart showing an example of result recovery processing. Fig.21 The process includes steps S90 to S94.

[0131] In step S90, it is determined whether the game user has played the game. If the game user has not played the game (No), the text consistency confirmation device 10 (e.g., the result collection process 115) returns to step S90 and waits until the user plays the game. On the other hand, if the game user has registered the game application and has played the game (Yes), the text consistency confirmation device 10 monitors the battle status and the progress of the game within a predetermined time.

[0132] In step S91, it is determined whether a round of the game is over. A round mentioned here may be, for example, when the game user has completed at least one question or one step of the game or quits after completing one question or one or more steps.

[0133] If the round is not over (No), the character consistency confirmation device 10 returns to step S90 and waits. On the other hand, if the round is over (Yes), the character consistency confirmation device 10 collects the game results of the character consistency and stores the collected game results in the evaluation information DB 134 according to the game user (step S92).

[0134] In step S93, it is determined whether the user continues to play the game. When the game user continues to play the game in step S93 (yes), the text consistency confirmation device 10 returns to step S91. On the other hand, when the game user does not continue to play the game in step S93 (no), the text consistency confirmation device 10 performs the evaluation / payment amount calculation process 117 described later (step S94). In addition, the collection of game results from the game user may not be carried out for each round. For example, it may also be that the user terminal in the login state stores the game results of the text consistency of one or more rounds, and sends the game results at a predetermined time or whenever a game result sending request is received from the text consistency confirmation device 10.

[0135] 《Majority voting process, AI judgment result comparison process》

[0136] Back to Fig.14 Then, the text consistency confirmation device 10 (for example, majority voting process 116) performs judgment processing using majority voting as the main technique on the collected game results, thereby confirming the text consistency confirmation as the game result and judging the accuracy (step S411). Fig. 22 : is a flowchart showing an example of majority voting processing and AI judgment result comparison processing. Fig. 22 The process includes steps S100 to S109 and S140 to S142. Steps S100 to S103 are steps corresponding to the majority voting process, and steps S104 to S109 and S140 to S142 are steps corresponding to the AI ​​determination result comparison process.

[0137] In step S100, it is determined whether the game result is recycled. When the game result is not recycled in step S100 (No), the text consistency confirmation device 10 returns to step S100 and waits until the game result is recycled. On the other hand, when the game result has been recycled (Yes), the text consistency confirmation device 10 proceeds to step S101.

[0138] In step S101, it is determined whether the number of answers required for majority voting is ensured. When the number of answers does not meet the number required for majority voting for the same text data in step S101, the text consistency confirmation device 10 waits until the necessary number of times is ensured. However, when the number of answers does not meet the necessary number of times after a predetermined time has passed (No), the text consistency confirmation device 10 determines that the text consistency cannot be confirmed, stores the reason code for failure in the confirmation result DB 136 (step S102), and ends the majority voting process.

[0139] Fig.23 FIG. 1 is a diagram showing an example of establishing a correspondence between an inability reason code and an inability reason. Fig.23 The left side of the failure reason code table 139 shows a failure reason code column 139A in which failure reason codes are recorded, and a failure reason column 139B in which failure reasons corresponding to each failure reason code are recorded. The failure reason code when the text consistency cannot be confirmed is stored in the failure reason code 136H of the storage unit 13.

[0140] exist Fig.23 In the table 139, as the reasons for failure, examples include "cannot read", "cannot judge by majority vote", "large difference from AI", etc., and each reason for failure is assigned a reason code of "01" to "03". The reason code table 139 is stored in the storage unit 13 of the text consistency confirmation device 10, for example.

[0141] On the other hand, when the number of answers for the same text data in step S101 satisfies the number required for majority voting (yes), the text consistency confirmation device 10 generates an arrangement of candidate characters corresponding to the text data in descending order of text consistency based on majority voting, and stores the generated arrangement in the majority voting result DB142 (step S103).

[0142] The text consistency checking device 10 receives the judgment processing result of the majority vote based on the majority voting process 116, and compares the arrangement generated from the recognition result of OCR or ICR using AI with the arrangement generated by the majority voting process (step S104). In addition, the arrangement using AI is generated in the text consistency checking device 10 in parallel with the distribution of the text data to the game user. The generated arrangement is stored in the AI ​​judgment result DB 141.

[0143] In step S105, it is determined whether the arrangement generated by the majority voting process 116 is consistent with the arrangement generated by AI. When these arrangements are consistent in step S105 (yes), the text consistency confirmation device 10 determines that the reliability of text consistency confirmation based on majority voting is high, and stores the arrangement generated by the majority voting process 116 (stored in the majority voting result DB 142) in the confirmation result DB 136 (step S141), and ends the AI ​​determination result comparison process.

[0144] On the other hand, if the arrangements are inconsistent in step S105 (No), the text consistency checking device 10 compares the arrangements of both parties (step S106) and determines the inconsistent candidate text (step S107). Then, the text consistency checking device 10 determines the reason for the inconsistency (step S108). As an example of the reason for the inconsistency, for example, Fig. 20Of course, there are also cases where these arrangements are not consistent even if there is no handwritten text.

[0145] In step S109, it is determined whether there is a handwritten input character. If there is no handwritten input character in step S109 (No), the character consistency checking device 10 stores an invalid reason code in the confirmation result DB 136 for the reason of inconsistency (step S142), and ends the AI ​​determination result comparison process.

[0146] On the other hand, in the case of handwritten input text (yes), the text consistency confirmation device 10 inserts the judgment result for the handwritten input text in the arrangement, and determines the reliability of the text consistency confirmation including the candidate text and the handwritten input text (step S140), stores the failure reason code in the confirmation result DB136 (step S142), and ends the AI ​​judgment result comparison processing.

[0147] Return to Fig.14 Based on the analysis result of step S411, the text consistency confirmation device 10 (e.g., the restoration process 118) performs the restoration process of the text data and stores the restored text data in the confirmation result DB 136 (step S412). In addition, the restoration process is a process that forms a pair with the segmentation / randomization process and is performed using an existing well-known technology.

[0148] 《Evaluation / payment amount calculation and processing》

[0149] The text consistency confirmation device 10 (e.g., evaluation / payment amount calculation processing 117) associates the restoration result with the game user information, and evaluates the game result in terms of correct answer rate, response rate, answer speed, transaction, etc. The evaluation result is stored in the evaluation information DB 134. For example, when evaluating the game user, the text consistency confirmation device 10 pre-sets evaluation points for evaluation items such as correct answer rate, response rate, answer speed, transaction, and usage frequency, and evaluates the game user based on the total of the evaluation points for each evaluation item.

[0150] In addition, the text consistency confirmation device 10 can also remove the answers of game users with extremely low evaluations during the majority voting process. This can improve the overall answer accuracy. These evaluation results are stored in the evaluation information DB 134 and / or the user host 131 and used as matching information of game users.

[0151] The text consistency checking device 10 feeds back the evaluation result of the game user to the game user as a rating, level, or reward information (step S413, step S414).

[0152] Fig.24 is a flowchart showing an example of evaluation / payment amount calculation processing. Fig.24 The process includes steps S110 to S117. In step S110, it is determined whether the majority voting process has been executed. If it is determined in step S110 that the majority voting process has not been executed (No), the text consistency checking device 10 returns to step S110 and waits until the majority voting process is executed.

[0153] On the other hand, when it is determined in step S110 that the majority voting process has been executed (yes), the text consistency confirmation device 10 performs sorting based on the accuracy of each game user (step S111), sorting based on the response rate (step S112), sorting based on the speed of answering (step S113), sorting based on the transaction processing (step S114), etc. according to the collected game results. In addition, the text consistency confirmation device 10 calculates the score according to the items such as accuracy, and determines the level, status, and rating of each game user (step S115). In addition, the text consistency confirmation device 10 can also use the items such as the response rate of the text with a greater degree of difficulty in text consistency confirmation, the handwritten text input rate, and the number of reasons for failure to answer for the evaluation of the game user.

[0154] The text consistency checking device 10 calculates the reward payment amount for each user based on the evaluation result using a predetermined ratio and coefficient (step S116). The text consistency checking device 10 stores the evaluation results of steps S111 to S115 and the calculation results of the reward payment amount in the evaluation information DB 134 (step S117).

[0155] Back to Fig.14 The game user can view his / her own rating and reward information etc. through the game screen etc. in the user terminal 30. The game user requests a reward from the text consistency checking device 10 as needed (step S415).

[0156] In response to the reward request from the user terminal 30 , the character consistency checking device 10 pays the reward to the corresponding game user, and stores the payment result in the payment history DB 135 (step S416 ).

[0157] The text consistency confirmation device 10 can also calculate the generated reward and notify the game user and the client's company as needed (step S417). In addition, step S417 can be performed simultaneously with step S414, or performed at a fixed time after step S414.

[0158] In addition, it is also possible to appropriately add various reward-related processes according to the usage situation. In other words, it is also possible not to perform reward-related processes.

[0159] 《Results sending and processing》

[0160] The character consistency checking device 10 (for example, the result transmission process 119 ) transmits the character consistency checking result and the like obtained by executing the series of processes and stored in the confirmation result DB 136 to the business terminal 20 (step S418 ).

[0161] Fig.25 FIG. 1 is a flowchart showing an example of result transmission processing. Fig.25 The process includes steps S130 to S132. In step S130, it is determined whether the confirmation result DB 136 is updated. If the confirmation result DB 136 is not updated (No), the text consistency checking device 10 returns to step S130 and waits until the confirmation result DB 136 is updated.

[0162] On the other hand, for example, when the restoration process based on the restoration process 118 is executed, Figure 8 When the result determination date 136D of the confirmation result DB 136 shown is updated (yes), the text consistency confirmation device 10 sends the processing result of the text data etc. on which the text consistency confirmation has been completed to the business terminal 20 (step S131), stores the result report date 136E, and performs the process of establishing and sending FLG 136F (step S132).

[0163] Back to Fig.14 When receiving the processing result from the character consistency checking device 10, the business terminal 20 stores the received processing result in the checked data DB 222 (step S419).

[0164] In addition, these series of processes are merely examples, and for example, the processes executed in the text consistency checking device 10 may sometimes be executed inside the user terminal 30, and the order of the processes may be appropriately switched within a range that does not affect the series of processes.

[0165] <Main Effects of the Present Embodiment>

[0166] According to the present embodiment, the received text data is segmented and randomized in the text consistency confirmation device 10. Specifically, the text consistency confirmation device 10 segments the text recognized by the text data of multiple clients and randomly distributes the segmented text data. According to this configuration, the association between the segmented and randomized text data and the original text data at the time of reception disappears, so the security of visual confirmation of text consistency based on external entrustment can be guaranteed.

[0167] Furthermore, according to the present embodiment, the character consistency checking device 10 sends the character data obtained by dividing the same character data to a plurality of game users. According to this configuration, it is difficult for a third party to restore the divided character data, so security can be guaranteed.

[0168] Furthermore, according to this embodiment, the character consistency checking device sends the subdivided identical character data to a plurality of game users. According to this configuration, since a plurality of character consistency checks are performed on the same character, the reliability of the character consistency check can be improved.

[0169] The present invention is not limited to the above-described embodiment, and includes various modified examples. Furthermore, the above-described embodiment is described in detail to explain the present invention in an easily understandable manner, and is not limited to having all the described configurations.

[0170] In addition, a part of the structure of a certain embodiment may be replaced with a structure of another embodiment, and a part of the structure of a certain embodiment may be added with a structure of another embodiment. Moreover, for a part of the structure of each embodiment, other structures may be added, deleted, or replaced. In addition, the components and relative sizes described in the drawings are simplified and idealized for easy understanding of the present invention, and may become more complicated shapes depending on actual installation.

[0171] Description of symbols

[0172] 1 text consistency confirmation system; 10 text consistency confirmation device; 11 control unit; 13 storage unit; 20 business terminal; 30 user terminal.

Claims

1. A text consistency confirmation system, characterized in that: have: The business terminal is operated by the client whose text consistency is confirmed; a text consistency confirmation device for controlling the text consistency confirmation; and The user terminal uses the game application to confirm the consistency of the text. The service terminal sends a file including text data for text consistency confirmation, The text consistency checking device subdivides the received text data, thereby eliminating the association with the original file, and sends publishing data including the subdivided text data to the user terminal. The user terminal performs the character consistency confirmation game on the subdivided character data included in the published data, and sends the game result to the character consistency confirmation device. The text consistency confirmation device performs judgment processing on the recovered game results using majority voting as the main technology, thereby determining the text consistency confirmation as the game result, and sending a result file including the text consistency confirmation result based on the received game result to the service terminal.

2. The text consistency confirmation system according to claim 1, characterized in that: The character consistency checking device subdivides characters recognized from the character data of a plurality of clients, and randomly assigns character data corresponding to the subdivided characters.

3. The text consistency confirmation system according to claim 1, characterized in that: The character consistency confirmation device sends the character data obtained by subdividing the same character data to a plurality of game users.

4. The text consistency confirmation system according to claim 1, characterized in that: The text consistency confirmation device sends the subdivided identical text data to multiple game users.

5. A text consistency confirmation device, characterized in that: The text consistency confirmation device receives text data to be confirmed for text consistency sent from a service terminal operated by a client, Subdividing the received text data so as to eliminate the association with the original file containing the text data, and sending the published data including the subdivided text data to a user terminal that uses a game application to check the consistency of the text, receiving from the user terminal a game result of the character consistency confirmation performed on the subdivided character data included in the published data, The collected game results are judged by majority voting, thereby confirming the consistency of the text as the game results. And a result file including a text consistency confirmation result based on the received game result is sent to the service terminal.

6. The text consistency checking device according to claim 5, characterized in that: The character consistency checking device subdivides characters recognized from the character data of a plurality of clients, and randomly assigns character data corresponding to the subdivided characters.

7. The text consistency checking device according to claim 5, characterized in that: The character consistency confirmation device sends the character data obtained by subdividing the same character data to a plurality of game users.

8. The text consistency checking device according to claim 5, characterized in that: The text consistency confirmation device sends the subdivided identical text data to multiple game users.

9. A method for confirming text consistency, characterized in that: include: The first step is to receive text data for text consistency confirmation sent from a service terminal operated by a client; The second step is to segment the received text data so as to eliminate the association with the original file containing the text data, and to send the published data including the segmented text data to the user terminal that uses the game application to check the consistency of the text; The third step is to receive from the user terminal the game result of the text consistency confirmation performed on the subdivided text data included in the published data; as well as The fourth step is to perform a judgment process on the recovered game results using majority voting as the main technology, thereby determining the text consistency confirmation as the game result, The fifth step is to send a result file including a text consistency confirmation result based on the received game result to the service terminal.

10. A storage medium, characterized in that: A text consistency verification program is stored to cause the processor to perform the following steps: The first step is to receive text data for text consistency confirmation sent from a service terminal operated by a client; The second step is to segment the received text data so as to eliminate the association with the original file containing the text data, and to send the published data including the segmented text data to the user terminal that uses the game application to check the consistency of the text; The third step is to receive from the user terminal the game result of the text consistency confirmation performed on the subdivided text data included in the published data; The fourth step is to perform a judgment process on the recovered game results using majority voting as the main technology, thereby determining the text consistency confirmation as the game result, The fifth step is to send a result file including a text consistency confirmation result based on the received game result to the service terminal.

11. The storage medium according to claim 10, characterized in that The second step includes the steps of segmenting characters recognized from the character data of a plurality of clients and randomly assigning character data corresponding to the segmented characters.

12. The storage medium according to claim 10, characterized in that The second step includes the step of sending the text data obtained by subdividing the same text data to a plurality of game users.

13. The storage medium according to claim 10, characterized in that The second step includes the following steps: sending the subdivided identical text data to a plurality of game users.

14. The text consistency confirmation system according to claim 1, characterized in that: When determining the character consistency confirmation, the character consistency confirmation device generates a ranking in which the candidate characters corresponding to the character data are ranked in descending order of character consistency based on majority voting.

15. The text consistency confirmation system according to claim 14, characterized in that: When determining the character consistency confirmation, the character consistency confirmation device compares the arrangement generated by using AI with the arrangement generated by majority voting processing.

16. The text consistency checking device according to claim 5, characterized in that: When the character consistency confirmation is determined, a ranking is generated for each candidate character corresponding to the character data in descending order of character consistency based on majority voting.

17. The text consistency checking device according to claim 16, characterized in that: When determining the character consistency confirmation, a comparison is performed between the arrangement generated by AI and the arrangement generated by majority voting processing.

18. The text consistency confirmation method according to claim 9, characterized in that: In the fourth step, when the character consistency confirmation is finalized, a ranking is generated for each candidate character corresponding to the character data in descending order of character consistency based on majority voting.

19. The storage medium according to claim 10, characterized in that In the fourth step, when the character consistency confirmation is finalized, a ranking is generated for each candidate character corresponding to the character data in descending order of character consistency based on majority voting.

20. The storage medium according to claim 19, characterized in that In the fourth step, when the character consistency confirmation is determined, the arrangement generated by using AI is compared with the arrangement generated by majority voting processing.

Images