Information processing system, information processing device, information processing method and computer program

Abstract

To provide an information processing system capable of effectively improving work efficiency.SOLUTION: An information processing system S includes: ambiguous expression detection and presentation units (101 and 201) that, when at least one of a term and a phrase with ambiguity is detected in a part or all of text included in document data, present at least one of the detected term and phrase with ambiguity; and logical check units (103 and 203) that compare text included in one of a plurality of document data associated with each other with text included in other document data, and issue a warning when a contradictory relation is detected between the text included in the one of the plurality of document data and the text included in the other document data.SELECTED DRAWING: Figure 2

Description

[Technical Field]

[0001] The present invention relates to an information processing system, an information processing device, an information processing method, and a computer program for improving work efficiency. [Background technology]

[0002] The use of computers can significantly improve the efficiency of work performed by humans. Needless to say, for example, the efficiency of communication or reporting work can be improved by creating text data using a computer and sending it to a desired party, compared to manual work.

[0003] In the field of computer processing, research and development into machine learning has been active in recent years, and natural language processing models in particular have made remarkable progress. Natural language processing models are becoming more widespread in various fields. The widespread use of such natural language processing models is expected to further improve the efficiency of human tasks in the future.

[0004] A natural language processing model is a learning model that can understand the natural language used by humans and perform processing corresponding to that understanding. For example, such a natural language processing model can understand questions and inquiries made by humans in natural language, generate answers to those questions, and provide them to users.

[0005] The accuracy of sentence understanding by natural language processing models is improving day by day as learning progresses. However, natural language processing models currently do not perform appropriate sentence understanding and may not provide desired processing. One of the reasons for this situation is the ambiguity of sentences written by humans. Patent Document 1, for example, is an example of a technology in the field of machine learning that focuses on the ambiguity of sentences. [Prior art documents] [Patent documents]

[0006] [Patent Document 1] Japanese Patent Publication No. 2023-021636 Summary of the Invention [Problem to be solved by the invention]

[0007] For example, when a person creates a text in an operation such as communication, reporting, questioning, or inquiry, ambiguous words or phrases can affect the understanding of the person to whom the text is provided.

[0008] In other words, both human understanding and natural language processing model understanding are affected by ambiguous words and phrases. The outputs (judgments, answers, etc.) of humans and natural language processing models based on inappropriate understanding may be inappropriate. If work proceeds based on inappropriate outputs, extensive revisions may be required, significantly reducing work efficiency and potentially resulting in significant losses.

[0009] For example, in project management, it is necessary to simultaneously manage many items such as personnel, materials, and schedules. Judgments and decisions must be made sequentially for many items, and each task generally proceeds quickly. In this type of project management, for example, mistakes in judgment and decision-making are likely to occur, and the accumulation of these mistakes can result in huge losses.

[0010] The present invention has been made in consideration of the above background, and aims to provide an information processing system, information processing device, information processing method, and computer program that can effectively improve work efficiency by reducing or eliminating ambiguity in sentences that are created by humans during work and that serve as a premise for human and / or computer understanding, such as learning models, and encouraging appropriate output (judgments, answers, etc.) from humans and computers that occurs during work based on understanding of the sentences. [Means for solving the problem]

[0011] The present invention relates to the following aspects.

[0012] an ambiguous expression detection and presentation unit that, when at least one of ambiguous terms and phrases is detected in a part or all of a sentence included in the document data, presents at least one of the detected ambiguous terms and phrases; An information processing system comprising: a logic check unit that compares a sentence contained in one of a plurality of mutually associated document data with a sentence contained in the other document data, and issues a warning if a contradictory relationship is detected between a sentence contained in one of the plurality of document data and a sentence contained in the other document data.

[0013] an ambiguous expression detection and presentation unit that, when at least one of ambiguous terms and phrases is detected in a part or all of a sentence included in the document data, presents at least one of the detected ambiguous terms and phrases; An information processing device comprising: a logic check unit that compares a sentence contained in one of a plurality of mutually associated document data with a sentence contained in the other document data, and presents a warning if a contradictory relationship is detected between a sentence contained in one of the plurality of document data and a sentence contained in the other document data.

[0014] an ambiguous expression detection and presentation step of presenting at least one of the detected ambiguous terms and phrases when at least one of the ambiguous terms and phrases is detected in a part or all of the sentences included in the document data; An information processing method comprising a logic check step of comparing a sentence contained in one of a plurality of mutually associated document data with a sentence contained in another document data, and presenting a warning if a contradictory relationship is detected between a sentence contained in one of the plurality of document data and a sentence contained in another document data.

[0015] an ambiguous expression detection and presentation step of presenting at least one of the detected ambiguous terms and phrases when at least one of the ambiguous terms and phrases is detected in a part or all of the sentences included in the document data; A computer program that causes a computer to execute a logic check step of comparing a sentence included in one of a plurality of mutually associated document data with a sentence included in another document data, and presenting a warning if a contradictory relationship is detected between a sentence included in one of the plurality of document data and a sentence included in another document data. [Effects of the Invention]

[0016] According to the present invention, work efficiency can be effectively improved. [Brief explanation of the drawings]

[0017] [Figure 1] 1 is a diagram schematically illustrating a configuration of an information processing system according to a first embodiment. [Figure 2] 2 is a block diagram showing the functional configuration of each unit constituting the information processing system shown in FIG. 1. FIG. [Figure 3] 10 is a flowchart illustrating the flow of an ambiguity check process performed in the information processing system shown in FIG. [Figure 4] 10 is a flowchart illustrating the flow of a logic check process performed in the information processing system shown in FIG. [Figure 5] 2 is a diagram specifically illustrating the ambiguity check process performed in the information processing system shown in FIG. 1. FIG. [Figure 6] 2 is a diagram specifically illustrating a logic check process performed in the information processing system shown in FIG. 1. FIG. [Figure 7] 2 is a diagram specifically illustrating a logic check process performed in the information processing system shown in FIG. 1. FIG. [Figure 8] 2 is a diagram illustrating an example of learning data of a learning model used in the information processing system shown in FIG. 1. [Figure 9]FIG. 10 is a diagram schematically illustrating a configuration of an information processing system according to a second embodiment. [Figure 10] FIG. 10 is a diagram schematically illustrating a configuration of an information processing system according to a third embodiment. [Figure 11] FIG. 10 is a diagram schematically illustrating a configuration of an information processing system according to a fourth embodiment. [Figure 12] FIG. 2 is a diagram illustrating a hardware configuration of an information processing device that can constitute a terminal device and a server that constitute the information processing system according to each embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0018] Each embodiment will be described below.

[0019] <<First embodiment>> <Outline of information processing system configuration> 1 shows a schematic configuration of an information processing system S according to the first embodiment. The information processing system S includes a terminal device 10 as an information processing device, a first server 20, and a second server 30.

[0020] The terminal device 10, the first server 20, and the second server 30 are connected to each other via a network 40 so as to be able to transmit and receive information to and from each other. The network 40 is a communication network including, for example, the Internet, a Wide Area Network (WAN), a Local Area Network (LAN), etc. The terminal device 10, the first server 20, and the second server 30 are connected to the network 40 wirelessly or via a wired connection, and can communicate with external devices. The functional configuration of the information processing system S will be described in detail below.

[0021] <Functional configuration of information processing system> 2 is a block diagram showing the functional configuration of each unit (terminal device 10, first server 20, and second server 30) that constitutes the information processing system S. The information processing system S performs specific processing on document data held in the terminal device 10. Specifically, the specific processing performed by the information processing system S is processing to check ambiguity in a sentence (ambiguity check processing (including, in this example, ambiguous expression detection and presentation processing and corrected expression presentation processing)) and processing to detect contradictory sentences (logic check processing).

[0022] In the information processing system S according to this embodiment, when a user makes a request to execute the ambiguity check process on a sentence included in document data on the terminal device 10, the first server 20 receives this request via the network 40. The first server 20 requests an answer to the received request from the second server 30 via the network 40. The second server 30 derives an answer to the request and provides it to the first server 20. The first server 20 provides the provided answer to the terminal device 10. The logic check process is also performed in the same procedure as the ambiguity check described above.

[0023] That is, the information processing system S according to this embodiment is essentially a system that provides so-called SaaS (Software as a Service). However, the specific form of the information processing system S is not particularly limited. In the information processing system S, for example, the functions performed by the second server 30 may be incorporated into the terminal device 10, or the functions performed by the second server 30 may be incorporated into the first server 20. Each part will be described in detail below.

[0024] (Terminal Device) The terminal device 10 is a device operated by a user, such as a personal computer, a smartphone, a tablet, etc. The terminal device 10 includes a terminal-side ambiguous expression detection and presentation unit 101, a terminal-side corrected expression presentation unit 102, and a terminal-side logic check unit 103.

[0025] The terminal-side ambiguous expression detection and presentation unit 101 presents at least one of the relevant terms and phrases when at least one of ambiguous terms and phrases is detected in part or all of a sentence contained in document data stored in the terminal device 10.

[0026] More specifically, the terminal-side ambiguous expression detection and presentation unit 101 in this embodiment transmits a request to execute an ambiguity check process for part or all of the sentences included in the document data to the first server 20 in response to a user operation. A response to this request to execute the ambiguity check process is acquired through cooperation between the first server 20 and the second server 30 and provided to the terminal device 10, more specifically, to the terminal-side ambiguous expression detection and presentation unit 101. Then, when the response to the request to execute the ambiguity check process is provided, the terminal-side ambiguous expression detection and presentation unit 101 presents information corresponding to the response on a display on the terminal device 10, for example.

[0027] The request for executing the ambiguity check process includes information about the sentence to be checked and an instruction for the ambiguity check process. The response to the request for executing the ambiguity check process includes information about ambiguous terms and / or phrases, or information indicating that such ambiguity does not exist. Specifically, the information about the detection of such ambiguity corresponds to the response to the request for executing the ambiguous expression detection and presentation process, which is included in the ambiguity check process request. As will be described later, in this embodiment, the response to the request for execution of the ambiguity check process may also include information on at least one of one or more corrective terms, one or more corrective phrases, and one or more modifiers that correct the ambiguity of the expression that has been determined to be an ambiguous term and / or phrase. This information on correction corresponds, in detail, to the response to the request for execution of the corrected expression presentation process, which is included in the request for execution of the ambiguity check process. This response to the request for execution of the corrected expression presentation process is processed by the terminal-side corrected expression presentation unit 102, as will be described later.

[0028] The manner in which the terminal-side ambiguous expression detection and presentation unit 101 presents information corresponding to the answer is not particularly limited, but may be, for example, a manner in which ambiguous terms and / or phrases are displayed in a space different from the display space of the sentence in the document data on the display, and text notifying the user that the ambiguity exists is displayed. In addition to displaying the text, ambiguous terms and / or phrases in the sentence in the document data on the display may be highlighted.

[0029] The terminal-side corrected expression presentation unit 102 is a functional unit for presenting at least one of one or more corrective terms, one or more corrective phrases, and one or more modifiers that correct the ambiguity of at least one of the terms and phrases detected by the terminal-side ambiguous expression detection and presentation unit 101. In this embodiment, when a request to execute an ambiguity check process for part or all of the sentences included in the document data is transmitted from the terminal-side ambiguous expression detection and presentation unit 101 to the first server 20, a request to propose corrective expressions for the ambiguous terms and / or phrases (a request to execute a corrected expression presentation process) is also transmitted at the same time. A response to this request to propose corrective expressions is also provided to the terminal device 10, more specifically, to the terminal-side corrected expression presentation unit 102, via the first server 20 and the second server 30.

[0030] When a response to the request for suggesting a corrected expression is provided, the terminal-side corrected expression presenting unit 102 presents information corresponding to the response on the display of the terminal device 10. The response to the suggestion request here refers to information such as corrective terms and phrases (alternative expressions) for ambiguous terms and / or phrases, or the addition of modifiers. Modifiers are a concept that includes nouns, collocations (phrases) of nouns and particles, adjectives, and adverbs.

[0031] Furthermore, the terminal-side corrected expression presenting unit 102 in this embodiment monitors whether the user has corrected the sentence after presenting information such as corrective terms for ambiguous terms and / or phrases and the addition of modifiers. When the user has corrected the sentence, the terminal-side corrected expression presenting unit 102 transmits the correction details to the first server 20. Furthermore, document data in which the user has corrected the sentence is held in the terminal device 10 in response to the user's operation.

[0032] Terminal-side logic check unit 103 is a functional unit for comparing sentences included in multiple document data between the document data and checking (verifying) the existence of contradictory (logically contradictory) sentences. In detail, terminal-side logic check unit 103 compares a sentence included in one of multiple document data associated with each other and held in terminal device 10 with a sentence included in the other document data, and presents a warning if a contradictory relationship is detected between a sentence included in one of the multiple document data and a sentence included in the other document data.

[0033] More specifically, the terminal-side logic check unit 103 in this embodiment transmits a request to the first server 20 to execute a logic check process that verifies the presence of contradictory sentences among multiple pieces of document data that are related to each other, in response to a user operation. A response to this request to execute the logic check is obtained by the first server 20 and the second server 30 working together, as in the case of the ambiguity check process, and is provided to the terminal device 10, more specifically, to the terminal-side logic check unit 103. Then, when the terminal-side logic check unit 103 receives the response to the request to execute the logic check, it presents information corresponding to the response on a display on the terminal device 10, for example.

[0034] The request to execute the logic check process includes information on the document data (including sentences) to be checked and a command to execute the logic check process. The response to the request to execute the logic check process includes information on multiple sentences that may be contradictory to each other and information indicating the existence of multiple sentences that may be contradictory to each other. A contradictory relationship refers to a state in which mutually exclusive facts exist between two sentences. Simple examples of contradictions include the relationship between "A is B" and "A is not B" and the relationship between "raise A" and "lower A." As will be described later, in this embodiment, the logic check process is performed using a learning model. In this case, the learning model can detect contradictions by, for example, generating and classifying sentence vectors.

[0035] The manner in which the information corresponding to the above answer by the terminal-side logic check unit 103 is presented is not particularly limited, but for example, it may be such that text indicating the existence of multiple sentences that may be contradictory is displayed in a space different from the display space of the sentences in each of the multiple document data on the display.

[0036] Furthermore, the terminal-side logic checking unit 103 in this embodiment issues a notification when at least one of the plurality of document data to be compared has not been processed (ambiguity check processing) by the terminal-side ambiguous expression detection and presentation unit 101. More specifically, the terminal-side logic checking unit 103 is configured not to execute processing (logic check) when at least one of the plurality of document data to be compared has not been processed by the terminal-side ambiguous expression detection and presentation unit 101.

[0037] The accuracy of the logic check performed by the terminal-side logic check unit 103 may decrease if there is ambiguity in the sentences being compared. Therefore, in this embodiment, if at least one of the multiple document data compared by the terminal-side logic check unit 103 has not been processed by the terminal-side ambiguous expression detection and presentation unit 101, a notification is given to prompt the user to execute processing by the terminal-side ambiguous expression detection and presentation unit 101. If the user executes processing by the terminal-side ambiguous expression detection and presentation unit 101 in response to the notification and corrects the sentence, the accuracy of the logic check performed by the terminal-side logic check unit 103 on the corrected sentence may improve.

[0038] However, the terminal-side logic checking unit 103 may perform processing in accordance with user instructions, regardless of whether at least one of the multiple document data being compared has been processed by the terminal-side ambiguous expression detection and presentation unit 101.

[0039] Furthermore, in this embodiment, the terminal-side logic checking unit 103 presents multiple sentences that may be contradictory to each other as answers, and then accepts a questionnaire from the user regarding the correctness of the answers. When the terminal-side logic checking unit 103 accepts a questionnaire regarding the correctness of the answers from the user, it transmits the contents of the questionnaire to the first server 20. However, such a questionnaire acceptance function may be omitted.

[0040] (First server) The first server 20 is a device that accepts requests from the terminal device 10 and presents a response to the request to the terminal device 10. As described above, the terminal device 10 transmits a request to execute an ambiguity check process and / or a request to execute a logic check process to the first server 20. In detail, the first server 20 accepts a request to execute an ambiguity check process and / or a request to execute a logic check process from the terminal device 10. Then, the first server 20 obtains a response to the request to execute an ambiguity check process and / or a request to execute a logic check process from the second server 30 and presents it to the terminal device 10.

[0041] The first server 20 includes a server-side ambiguous expression detection and presentation unit 201 , a server-side corrected expression presentation unit 202 , and a server-side logic check unit 203 .

[0042] The server-side ambiguous expression detection and presentation unit 201 accepts a request to execute an ambiguity check process from the terminal device 10, and transmits information on the sentence to be checked and a command for the ambiguity check process to the second server 30. Then, the server-side ambiguous expression detection and presentation unit 201 presents a response to the request to execute the ambiguity check process provided by the second server 30 by transmitting it to the terminal device 10. As described above, the response to the request to execute the ambiguity check process means information on ambiguous terms and / or phrases, or information indicating that such information does not exist.

[0043] The server-side corrected expression presentation unit 202 transmits this information to the terminal device 10 when the response to the request to execute the ambiguity check process provided by the second server 30 includes at least one of one or more corrective terms, one or more corrective phrases, and one or more modifiers as correction suggestions.

[0044] In this embodiment, when a request to execute the ambiguity check process is transmitted from the terminal-side ambiguous expression detection and presentation unit 101 to the first server 20 as described above, a request to propose modified expressions for ambiguous terms and / or phrases is also transmitted at the same time. When the server-side modified expression presentation unit 202 acquires a response to the request to propose modified expressions from the second server 30, it transmits the response to the terminal device 10. As described above, the response to the proposal request means information such as a modified term, modified phrase, or addition of a modifier for the ambiguous term and / or phrase.

[0045] As described above, the terminal-side corrected expression presenting unit 102 of the terminal device 10 in this embodiment presents information such as a corrected term, a corrected phrase, or the addition of a modifier to an ambiguous term and / or phrase, and then monitors whether the user has corrected the sentence. When the user corrects the sentence, the terminal-side corrected expression presenting unit 102 transmits the correction details to the first server 20. The correction details are stored in the server-side corrected expression presenting unit 202. As will be described later, the ambiguity check process is performed using a pre-trained learning model stored in the second server 30. In this embodiment, the server-side corrected expression presenting unit 202 provides the information on the correction details provided by the terminal device 10 to the second server 30 for additional learning of the learning model.

[0046] The server-side logic checking unit 203 accepts a request to execute a logic check process from the terminal device 10, and transmits information on the document data (text) to be checked and a command for the logic check process to the second server 30. The server-side logic checking unit 203 also presents a response to the request to execute a logic check process provided by the second server 30 by transmitting it to the terminal device 10. As described above, the response to the request to execute a logic check process means information on multiple texts that may be contradictory to each other and information indicating the existence of multiple texts that may be contradictory to each other.

[0047] As described above, the terminal-side logic checking unit 103 of the terminal device 10 in this embodiment presents information on multiple sentences that may be contradictory to each other as answers, and then accepts a questionnaire from the user regarding the correctness of the answers. When the terminal-side logic checking unit 103 accepts a questionnaire regarding the correctness of the answers from the user, it transmits the contents of the questionnaire to the first server 20. The contents of the questionnaire are stored in the server-side logic checking unit 203. Similar to the ambiguity check process, the logic check process is also performed using a pre-trained learning model stored in the second server 30. In this embodiment, the server-side logic checking unit 203 provides the information on the questionnaire contents provided from the terminal device 10 to the second server 30 for additional learning of the learning model.

[0048] In this embodiment, the first server 20 functions as a relay between the terminal device 10 and the second server 30. When the first server 20 is interposed between the terminal device 10 and the second server 30 in this way, there is an advantage that feedback information (such as the above-mentioned correction contents and questionnaire contents) can be efficiently collected from a plurality of terminal devices 10. However, the first server 20 may be omitted, and in this case, for example, a configuration may be adopted in which the terminal device 10 obtains a response to a request directly from the second server 30.

[0049] (Second server) The second server 30 is a device that receives a request from the terminal device 10 via the first server 20 and presents a response to the request. The second server 30 includes a first learning model 301, a second learning model 302, and a third learning model 303.

[0050] When the second server 30 receives a request to execute an ambiguity check process, it derives an answer using the first learning model 301 and the second learning model 302. When the second server 30 receives a request to execute a logic check process, it derives an answer using the third learning model 303.

[0051] The first learning model 301, the second learning model 302, and the third learning model 303 are all natural language processing models. That is, the first learning model 301, the second learning model 302, and the third learning model 303 all function to understand sentences provided from the terminal device 10 via the first server 20 and to verify ambiguity and logical contradictions. The first learning model 301, the second learning model 302, and the third learning model 303 may be natural language processing models based on BERT (Bidirectional Encoder Representations from Transformers).

[0052] In this embodiment, the terminal-side ambiguous expression detection and presentation unit 101, the terminal-side corrected expression presentation unit 102, and the terminal-side logicality check unit 103 of the terminal device 10, and the server-side ambiguous expression detection and presentation unit 201, the server-side corrected expression presentation unit 202, and the server-side logicality check unit 203 of the first server 20 perform corresponding processing using natural language processing based on a learning model.

[0053] The first learning model 301 is a learning model that uses training sentences as input data and pre-trains learning data that uses at least one of the ambiguous terms and phrases in the training sentences as correct answer data. As a result, when the first learning model 301 receives a sentence as input data, it can derive at least one of the ambiguous terms and phrases as an answer and provide it to the first server 20.

[0054] The ambiguous terms and phrases derived by the first learning model 301 are basically determined in advance by pre-learning. In this embodiment, the learning model is BERT, and ambiguity is evaluated from the context. This allows ambiguous terms and phrases to be detected with high accuracy. Furthermore, additional learning can improve the detection accuracy.

[0055] The second learning model 302 is a learning model that has pre-trained learning data using training sentences as input data and corrected sentences obtained by correcting the training sentences as correct answer data. As a result, when the second learning model 302 receives a sentence as input data, it can derive a corrected sentence as an answer. In more detail, in this embodiment, the second learning model 302 provides the first server 20 with the differences and changes between the sentence as input data and the corrected sentence as information such as corrective terms, corrective phrases, and addition of modifiers to ambiguous terms and / or phrases.

[0056] The third learning model 303 is a learning model that uses first training sentences and second training sentences different from the first training sentences as input data, and pre-trains learning data in which tags indicating logical agreement (consistency) or contradiction (divergence) between the first training sentences and the second training sentences are used as correct answer data. As a result, when the third learning model 303 receives multiple sentences as input data, it can derive contradictions between the sentences and provide them to the first server 20.

[0057] As described above, the server-side corrected expression presenting unit 202 provides information on the correction content of the sentence provided from the terminal device 10 to the second server 30 for additional learning of the learning model. The server-side logicality checking unit 203 provides information on the questionnaire content regarding logicality provided from the terminal device 10 to the second server 30 for additional learning of the learning model. The information on the correction content of the sentence can be provided from the first server 20 to the second learning model 302. In this case, the second learning model 302 can perform additional learning (fine tuning) based on the information on the correction content of the sentence. Similarly, information on the questionnaire content regarding logicality can be provided from the first server 20 to the third learning model 303. In this case, the third learning model 303 can perform additional learning (fine tuning) based on the information on the questionnaire content.

[0058] It is preferable that additional learning in the second learning model 302 and the third learning model 303 is performed for each terminal device 10 (user). When fine tuning is performed, in this embodiment, an additional layer is set in the neural network of the second learning model 302 and the third learning model 303. In this case, by setting the additional layer for each terminal device 10 (user), it is possible to create a learning model suitable for each user.

[0059] In this embodiment, the second server 30 has a first learning model 301, a second learning model 302, and a third learning model 303. However, the second server 30 may be configured to have a single learning model that can execute the processes performed by the first learning model 301, the second learning model 302, and the third learning model 303. Also, for example, a server having the first learning model 301 and the second learning model 302 and a server having the third learning model 303 may be provided separately.

[0060] <Processing by information processing system> An example of processing by the information processing system S will be described below. Fig. 3 is a flowchart illustrating the flow of the ambiguity check processing performed by the information processing system S. Fig. 4 is a flowchart illustrating the flow of the logic check processing performed by the information processing system S. First, the ambiguity check processing will be described with reference to Fig. 3.

[0061] (Ambiguous check processing) 3, a flowchart showing the processing of the terminal device 10 is shown on the left side, and a flowchart showing the processing of the first server 20 is shown on the right side. The processing of the terminal device 10 starts with the activation of a dedicated browser that accepts instructions for the ambiguity check processing and the logic check processing. The first server 20 is always active, and performs processing in conjunction with the terminal device 10.

[0062] In terminal device 10, input of document data is first monitored in step S101. When document data is input (YES in step S101), whether or not a sentence has been specified by the user in the document data is monitored in step S102. If document data is not input in step S101 (NO in step S101), as shown by the transition "A" in the figure, the process proceeds to step S110, where an end instruction or the occurrence of another event is monitored. If an end instruction or the occurrence of another event is not input in step S101, input of document data is monitored again in step S101.

[0063] In step S102, it is monitored whether the user has specified all or part of the text included in the document data. The manner of specifying the text monitored in step S102 may be, for example, a manner in which the user specifies all or part of the text included in the document data to be displayed on the display using a mouse, a touch panel, or the like.

[0064] If a sentence is specified in step S102 (YES in step S102), it is monitored in step S103 whether or not an instruction for ambiguity check processing has been issued by the user. If no sentence is specified in step S102 (NO in step S102), the process proceeds to step S110, where an end instruction or the occurrence of another event is monitored. If no end instruction or the occurrence of another event is detected in step S110, the process returns to step S101 and monitors the specification of a sentence again in step S102.

[0065] When the instruction for the ambiguity check process is confirmed in step S103 (YES in step S103), in step S104, the specified sentence and a request to execute the ambiguity check process are sent to the first server 20. Note that the monitoring in steps S102 and S103 and the sending process in step S104 correspond to the processing by the terminal-side ambiguous expression detection and presentation unit 101.

[0066] When the specified sentence and a request to execute the ambiguity check process are sent to the first server 20 in step S104, the server-side ambiguous expression detection and presentation unit 201, which has been waiting to receive a sentence from the terminal device 10 in step S121, confirms the reception and proceeds to step S122. Then, in step S122, the server-side ambiguous expression detection and presentation unit 201 requests the first learning model 301 and the second learning model 302 of the second server 30 to perform an ambiguity check process on the received sentence, and then obtains the result of the ambiguity check process.

[0067] The results of the ambiguity check process from the first learning model 301 and the second learning model 302 correspond to a response to a request to execute the ambiguity check process. In this embodiment, the response includes at least one of information about ambiguous terms and / or phrases in the sentence, information indicating their absence, and information such as a corrective term for the ambiguous term and / or phrase or the addition of a modifier.

[0068] When the result (answer) of the ambiguity check process is acquired in step S122, in step S123, the server-side ambiguous expression detection and presentation unit 201 and the server-side corrected expression presentation unit 202 transmit corresponding information to the terminal device 10. After the transmission process is performed in step S123, the server-side corrected expression presentation unit 202 waits to receive information on the corrected sentence from the terminal device 10 in step S124.

[0069] Returning to the processing on the terminal device 10 side, when the result of the ambiguity check processing is sent to the terminal device 10 in step S123, the terminal side ambiguous expression detection and presentation unit 101 and the terminal side corrected expression presentation unit 102 each receive the corresponding information in step S105.

[0070] Then, in step S106, the terminal-side ambiguous expression detection and presentation unit 101 presents, for example, on a display, information corresponding to the response to the request to execute the ambiguity check process. Also, in step S106, the terminal-side corrected expression presentation unit 102 presents, for example, on a display, as a response to the request to execute the ambiguity check process, at least one of one or more corrective terms, one or more corrective phrases, and one or more modifiers that correct the ambiguity of the ambiguous term and / or phrase. Also, if there are no ambiguous terms and phrases, the terminal-side ambiguous expression detection and presentation unit 101 presents that fact.

[0071] Thereafter, in step S107, it is recorded that the document data has undergone the ambiguity check process (more specifically, the ambiguous expression detection and presentation process). Thereafter, in step S108, it is monitored whether the sentence designated as the target of the ambiguity check process has been corrected by the user. If the sentence has been corrected in step S108, the corrected sentence is recorded in terminal device 10 and transmitted to first server 20 in step S109. Thereafter, in step S110, it is monitored whether an end instruction or another event has occurred. If an end instruction or another event has not occurred, the designation of the sentence is monitored again in step S102 after step S101, and if an end instruction or another event has occurred, the ambiguity check process ends.

[0072] On the other hand, if the sentence is not corrected in step S108, the occurrence of an end instruction or other event is monitored in step S110. In this case, whether an end instruction or other event has not occurred and new sentence is specified in step S102, or the occurrence of an end instruction or other event is confirmed, information indicating that no correction has been made to the sentence that is the target of the ambiguity check process is transmitted to the first server 20.

[0073] If the corrected sentence is transmitted in step S109, the first server 20 confirms receipt of the corrected sentence in step S124 (YES in step S124), records the corrected sentence in step S125, and then returns the process to step S121. Also, if receipt of the corrected sentence is not confirmed in step S124 and information indicating that no correction has been made to the sentence that is the target of the ambiguity check process is received (NO in step S124), the first server 20 does not record the sentence and returns the process to step S121. Note that the corrected sentence recorded in step S125 can be used for additional learning, as described above.

[0074] (Logical check processing) Next, the logic check process will be described with reference to Fig. 4. In Fig. 4, a flowchart showing the process of the terminal device 10 is shown on the left side, and a flowchart showing the process of the first server 20 is shown on the right side.

[0075] In terminal device 10, first, in step S201, input of multiple pieces of document data associated with each other is monitored. When document data is input (YES in step S201), in step S202, it is monitored whether or not a user has issued an instruction for logic check processing. If no document data is input in step S201 (NO in step S201), as shown by the transition "B" in the figure, the process proceeds to step S209, where an end instruction or the occurrence of another event is monitored. If no end instruction or the occurrence of another event is detected in step S209, input of document data is again monitored in step S201.

[0076] If the instruction to perform the logic check process is confirmed in step S202 (YES in step S202), it is confirmed in step S203 whether the ambiguity check process has been performed on the input document data. This confirmation is performed based on the ambiguity check process completion flag of the document data in step S107 of FIG.

[0077] If at least one document data item for which the ambiguity check process has not been performed is confirmed in step S203, a notification that the ambiguity check process has not been performed is displayed, for example, on a display in step S204. After the notification in step S204, as shown by the transition "C" in the figure, the process proceeds to step S209, where an end instruction or the occurrence of another event is monitored. If an end instruction or the occurrence of another event is not detected in step S209, the input of document data is monitored again in step S201.

[0078] On the other hand, when it is confirmed in step S203 that the ambiguity check process has been performed, in step S205, the plurality of document data input in step S201 or parts of the sentences therein and a request to perform the logic check process are transmitted to the first server 20. The processes in steps S201 to S204 correspond to the processes performed by the terminal-side logic check unit 103.

[0079] In step S205, multiple document data or parts of the sentences therein and a request to execute a logic check process are sent to the first server 20. In step S221, the server-side logic check unit 203, which has been waiting to receive document data from the terminal device 10, confirms the reception (YES in step S221) and proceeds to step S222. Then, in step S222, the server-side logic check unit 203 requests the third learning model 303 of the second server 30 to execute a logic check process to verify the existence of contradictory sentences among the received multiple related document data, and then obtains the results of the logic check process.

[0080] The result of the logic check process from the third learning model 303 corresponds to a response to a request to execute the logic check process. In this embodiment, the response includes information on multiple sentences that may be contradictory to each other and information indicating the existence of multiple sentences that may be contradictory to each other.

[0081] When the result of the logic check process is acquired in step S222, in step S223 the server-side logic check unit 203 transmits corresponding information to the terminal device 10. After the transmission process is performed in step S223, the server-side logic check unit 203 waits to receive information on the validity of the answers (questionnaire contents) from the terminal device 10 in step S224.

[0082] Returning to the processing on the terminal device 10 side, when the result (answer) of the logic check processing is transmitted to the terminal device 10 in step S223, the terminal side logic check unit 103 receives information on the verification result (answer) of the logic check processing in step S206.

[0083] Then, in step S207, the terminal-side logic check unit 103 presents information corresponding to the response to the request to execute the logic check process, for example, on a display. Here, for example, if a potentially contradictory sentence is detected, a notification to that effect may be displayed on the display, and the corresponding sentence may be highlighted.

[0084] Thereafter, in step S208, the terminal-side logic check unit 103 accepts the contents of a questionnaire, for example, as to whether the verification results for the presented possibly contradictory sentences were valid, and transmits the contents to the first server 20. Thereafter, in step S209, the occurrence of an end instruction or another event is monitored. If an end instruction or another event has not occurred, the input of document data is monitored again in step S201, and if an end instruction or another event is confirmed to have occurred, the ambiguity check process ends.

[0085] If the survey content is transmitted in step S208, the first server 20 confirms receipt of the survey content in step S224 (YES in step S224), records the survey content (information on the sentences and whether contradictions between sentences are valid or invalid) in step S225, and then returns the process to step S221. The survey content recorded in step S225 can be used for additional learning, as described above. The flow of returning the process to step S221 without receiving the survey content in step S224 is the same as that described in FIG. 3, and therefore will not be described here.

[0086] (Specific image of the process) Hereinafter, a specific image of the ambiguity check process and the logic check process in this embodiment will be described with reference to FIGS.

[0087] First, FIG. 5 is a diagram specifically explaining the ambiguity check process. More specifically, FIGS. 5(A) to 5(C) show images of the interface displayed on the display of the terminal device 10 when the ambiguity check process is executed. Reference numeral 51 in FIG. 5(A) indicates the title of the input document data ("Minutes" in the illustrated example). Below the title 51, the text included in the document data is displayed. The box indicated by reference numeral 52 indicates that a portion of the text has been specified by the user. When such a specification is made, a pop-up 53 prompting the execution of the ambiguity check process is displayed. Then, when the pop-up 53 is operated with, for example, a hand or a cursor 54, a request to execute the ambiguity check process is generated, and the request to execute the ambiguity check process is transmitted to the first server 20. The process up to this point corresponds to steps S101 to S104 in FIG. 3.

[0088] FIG. 5(B) shows how a response to a request to execute an ambiguity check process is displayed on the display. In this example, under the title "Feedback" 55, the response to the ambiguity check for the sentence is displayed, which includes an ambiguous term and / or phrase, one or more corrective terms, one or more corrective phrases, and modifiers that correct the ambiguity of the term and / or phrase. In FIG. 5(B), information indicated by reference numerals 56A and 56B corresponds to information on an ambiguous term and / or phrase. Information indicated by reference numeral 57 corresponds to one or more corrective terms, one or more corrective phrases, and modifiers that correct the ambiguity. The processing up to this point corresponds to steps S105 and S106 in FIG. 3.

[0089] Figure 5(C) shows how the user corrects the sentence based on the information displayed in Figure 5(B). Figure 5(C) shows, as an example, how the user selects a modifier to correct the ambiguity using a hand or cursor 58. When such a selection operation is performed, the sentence is corrected as shown in the correction location indicated by reference numeral 59.

[0090] Fig. 6 is a diagram specifically explaining the logic check process. In detail, Figs. 6(A) and (B) show images of the interface displayed on the display of the terminal device 10 when the logic check process is executed. Reference numeral 61 in Fig. 6(A) indicates the title of the input document data (in the illustrated example, "E Corporation New Business Development"). As an example, the multiple document data are associated by the title 61. Below the title 61, the text contained in the document data, the date it was created, etc. are displayed.

[0091] More specifically, in column 62, sentences 62A and 62B regarding decisions in the first document data created on October 30, 2023 are displayed. In column 63, sentence 63A regarding decisions in the second document data created on March 30, 2024 is displayed. The interface also shows a logic check execution button 64. By operating logic check execution button 64, a request to execute a logic check process is generated, and this execution request is sent to first server 20. The processing up to this point corresponds to steps S201 and 202 in FIG. 4.

[0092] Figure 6(B) shows the response to the request to execute the logic check process displayed on the display. In this example, a notification 66 indicating the existence of a contradictory sentence is displayed at the top of the interface. Furthermore, the contradictory sentence is highlighted by being enclosed in a two-dot chain line. The above process corresponds to steps S205 and S206 in Figure 4.

[0093] Furthermore, in this embodiment, if at least one document data item that has not undergone ambiguity check processing is found among the document data that is the subject of logic check processing, a notification that the ambiguity check processing has not been performed is issued. Figure 7 shows an example of a notification that the ambiguity check processing has not been performed. In Figure 7, a notification 68 indicating that the ambiguity check has not been performed is displayed at the top of the interface. The above processing corresponds to steps S203 and 204 in Figure 4.

[0094] (Image of training data) FIG. 8 is a diagram illustrating an example of learning data for a learning model used in the information processing system S. FIG. 8(A) shows an example of learning data used for learning by the first learning model 301. FIG. 8(B) shows an example of learning data used for learning by the second learning model 302. FIG. 8(C) shows an example of learning data used for learning by the third learning model 303.

[0095] The training data 81 shown in Figure 8(A) is training data that uses training sentences as input data and sets at least one of the ambiguous terms and phrases in the training sentences as correct answer data. Figure 8(A) shows only a portion of the training data that the first learning model 301 learns. For example, for the input data "This week, the roles have been sorted out," "this week" and "role" are set as correct answer data for ambiguous terms. For example, the first learning model 301 can be constructed by pre-training a huge amount of such training data.

[0096] The learning data 82 shown in Figure 8(B) uses training sentences as input data and corrected sentences obtained by correcting the training sentences as correct answer data. Figure 8(B) shows only a portion of the learning data learned by the second learning model 302. For example, for input data such as "Roles have been organized this week," the corrected sentence "Roles of Department Y have been organized on A-month B-day, 20XX" is set as correct answer data. For example, the second learning model 302 can be constructed by pre-learning a huge amount of such learning data.

[0097] The training data 83 shown in FIG. 8(C) is training data in which first training sentences and second training sentences different from the first training sentences are used as input data, and tags indicating logical agreement (consistency) or contradiction (contradiction) between the first training sentences and the second training sentences are used as correct answer data. FIG. 8(C) shows only a portion of the training data trained by the third learning model 303. For example, for input data including "sentence A" as the first training sentence and "sentence B" as the second training sentence, the tag "contradiction" is set as correct answer data. For example, the third learning model 303 can be constructed by previously training a huge amount of such training data.

[0098] The learning model used in the information processing system S is not particularly limited, and a learning model created by another person may be used. Even in this case, a learning model that meets the intentions of the user or service provider can be created by performing additional learning.

[0099] The information processing system S according to the first embodiment described above includes an ambiguous expression detection and presentation unit (101, 201) that presents at least one of the detected ambiguous terms and phrases when at least one of the ambiguous terms and phrases is detected in part or all of the sentences included in the document data, and a logic check unit (103, 203) that compares a sentence included in one of a plurality of mutually associated document data with a sentence included in the other document data, and presents a warning when a contradictory relationship is detected between a sentence included in one of the plurality of document data and a sentence included in the other document data.

[0100] With this configuration, the ambiguity of a sentence can be resolved or reduced based on ambiguous terms and / or phrases presented by the ambiguous expression detection and presentation unit (101, 201). This can improve the accuracy of interpretation by a person who interacts with the corrected sentence. Furthermore, the accuracy of tasks such as judgments and decisions made by a person can be improved based on the detection of contradictory sentences presented by the logic check unit (103, 203). The information processing system S according to this embodiment has both a tool for resolving or reducing the ambiguity of a sentence and a tool for verifying the logic between sentences, thereby enabling logic verification of the sentence after the ambiguity of the sentence has been resolved or reduced. It is noteworthy about the information processing system S according to this embodiment that when logic verification is performed on the sentence after the ambiguity of the sentence has been resolved or reduced, the accuracy of interpretation by the logic check unit (103, 203), which is a computer that interacts with the corrected sentence, can also be improved. By utilizing this information processing system S according to this embodiment, the accuracy of tasks such as judgments and decisions made by a person can be further improved.

[0101] In this way, the information processing system S of this embodiment encourages appropriate human and computer output (judgments, responses, etc.) that occurs during work based on understanding the text, thereby effectively improving the efficiency of work that progresses based on the text.

[0102] The information processing system S according to the first embodiment further includes a modified expression presentation unit (102, 202) that presents at least one of one or more modified terms, one or more modified phrases, and one or more modifiers that correct the ambiguity of at least one of the terms and phrases detected by the ambiguous expression detection and presentation unit (101, 201). This allows the user to easily correct the document.

[0103] Furthermore, in the present embodiment, the logic check unit (103, 203) notifies the user and does not perform a logic check if at least one of the document data to be compared has not been processed by the ambiguous expression detection and presentation unit (101, 201). In this case, the logic check unit (103, 203) is effectively utilized, and work efficiency is effectively improved.

[0104] <<Second embodiment>> 9 is a diagram showing a schematic configuration of an information processing system Sr1 according to the second embodiment. In the following embodiments, components common to those of the first embodiment are assigned the same reference numerals, and redundant explanations will be omitted.

[0105] 9 includes a terminal device 10 and a second server 30, but does not include a first server 20. In this configuration, the terminal device 10 obtains a response to the ambiguity check process and / or the logic check process directly from the second server 30.

[0106] <<Third embodiment>> 10 is a diagram illustrating a schematic configuration of an information processing system Sr2 according to a third embodiment. The information processing system Sr2 includes a terminal device 10 and a first server 20′, but does not include a second server 30. The first server 20′ includes a first learning model 301, a second learning model 302, and a third learning model 303.

[0107] <<Fourth embodiment>> 11 is a diagram illustrating a schematic configuration of an information processing system Sr3 according to a fourth embodiment. The information processing system Sr3 is composed only of a terminal device 10′, and does not include a first server 20 or a second server 30. The terminal device 10′ includes a first learning model 301, a second learning model 302, and a third learning model 303.

[0108] <Example of hardware configuration of information processing device> 12 is a diagram showing the hardware configuration of an information processing device 400 that can constitute the terminal device 10, the first server 20, and the second server 30. The information processing device 400 includes a processor 401, a ROM 402, a RAM 403, a communication interface 404, an operation device 405, and a display device 406, which are interconnected by a bus.

[0109] The processor 401 may be a CPU, a GPU, or a computing device that combines a CPU and a GPU. The processor 401 executes, for example, a computer program stored in a ROM 402 on a RAM 403. The computer program is a program that realizes the above-mentioned functional configurations of the terminal device 10, the first server 20, and / or the second server 30. The computer program may be realized by a combination of multiple programs or scripts rather than a single program.

[0110] The ROM 402 is a non-transitory computer-readable storage medium that stores computer programs, data required for executing the computer programs, and data generated by executing the computer programs. These computer programs and data are read into the RAM 403 when the computer programs are executed. The ROM 402 may be, for example, but is not limited to, a hard disk, an optical disk, a flash memory, or a magnetic tape.

[0111] The computer program may be pre-installed in the information processing device 400, or may be stored in a non-transitory computer-readable storage medium such as a CD-ROM. The computer program may also be uploaded to the Internet. The RAM 403 is a non-transitory computer-readable storage medium, such as, but not limited to, a DRAM or an SRAM.

[0112] The communication interface 404 is a circuit for the terminal device 10, the first server 20, and / or the second server 30 to communicate wirelessly or wired with an external device. The operation device 405 is a circuit for inputting operation signals from input devices such as a keyboard, a mouse, and a touch panel to the terminal device 10, the first server 20, and / or the second server 30. The display device 406 displays, for example, data stored in the terminal device 10, the first server 20, and / or the second server 30. The display device 406 may be, for example, an LCD (liquid crystal display), an organic electroluminescence display, or the like.

[0113] Furthermore, the information processing device 400 may be configured as a single device, or may be configured as a system made up of multiple computer devices connected to each other.

[0114] The above-described embodiment shows only one example for realizing the present invention, and the present invention can be implemented in various other forms. For example, various modifications, substitutions, omissions, or combinations thereof are possible without departing from the spirit of the present invention. Such modifications, substitutions, omissions, etc. are also included within the scope of the present invention, as well as the scope of the inventions described in the claims and their equivalents. [Explanation of symbols]

[0115] S, Sr1, Sr2, Sr3...Information processing system 10...Terminal device 20...First server 30...Second server 40…Network 101... Terminal side ambiguous expression detection and presentation unit 102... Terminal side corrected expression presentation unit 103...Terminal side logic check unit 201...Server-side ambiguous expression detection and presentation unit 202...Server-side corrected expression presentation unit 203...Server-side logic check unit 301...First Learning Model 302…Second Learning Model 303…Third Learning Model

Claims

1. an ambiguous expression detection and presentation unit that, when at least one of ambiguous terms and phrases is detected in a part or all of a sentence included in the document data, presents at least one of the detected ambiguous terms and phrases; An information processing system comprising: a logic check unit that compares a sentence contained in one of a plurality of mutually associated document data with a sentence contained in the other document data, and issues a warning if a contradictory relationship is detected between a sentence contained in one of the plurality of document data and a sentence contained in the other document data.

2. The information processing system further comprises a modified expression presentation unit that presents at least one of one or more modified terms, one or more modified phrases, and one or more modifiers that correct ambiguity regarding at least one of the terms and phrases detected by the ambiguous expression detection and presentation unit.

3. 2. The information processing system according to claim 1, wherein the logic check unit issues a notification when at least one of the plurality of document data to be compared has not been processed by the ambiguous expression detection and presentation unit.

4. 2. The information processing system according to claim 1, wherein the logic check unit does not execute processing when at least one of the plurality of document data to be compared has not been processed by the ambiguous expression detection and presentation unit.

5. The information processing system according to claim 1 , wherein the ambiguous expression detection and presentation unit and the logic check unit perform corresponding processing by natural language processing using a learning model.

6. The information processing system of claim 1 , wherein the ambiguous terms and phrases are predetermined.

7. The information processing system of claim 1, wherein the ambiguous expression detection and presentation unit uses training sentences as input data and performs corresponding processing using a learning model that has been pre-trained with learning data in which at least one of the ambiguous terms and phrases in the training sentences is correct data.

8. The information processing system of claim 2, wherein the corrected expression presentation unit uses training sentences as input data and performs corresponding processing using a learning model that has pre-trained learning data in which corrected sentences obtained by correcting the training sentences are used as correct answer data.

9. 2. The information processing system of claim 1, wherein the logic checking unit uses first training sentences and second training sentences different from the first training sentences as input data, and performs corresponding processing using a learning model that has been pre-trained with learning data in which tags indicating logical agreement or contradiction between the first training sentences and the second training sentences are used as correct answer data.

10. an ambiguous expression detection and presentation unit that, when at least one of ambiguous terms and phrases is detected in a part or all of a sentence included in the document data, presents at least one of the detected ambiguous terms and phrases; An information processing device comprising: a logic check unit that compares a sentence contained in one of a plurality of mutually associated document data with a sentence contained in the other document data, and presents a warning if a contradictory relationship is detected between a sentence contained in one of the plurality of document data and a sentence contained in the other document data.

11. an ambiguous expression detection and presentation step of presenting at least one of the detected ambiguous terms and phrases when at least one of the ambiguous terms and phrases is detected in a part or all of the sentences included in the document data; An information processing method comprising a logic check step of comparing a sentence contained in one of a plurality of mutually associated document data with a sentence contained in another document data, and presenting a warning if a contradictory relationship is detected between a sentence contained in one of the plurality of document data and a sentence contained in another document data.

12. an ambiguous expression detection and presentation step of presenting at least one of the detected ambiguous terms and phrases when at least one of the ambiguous terms and phrases is detected in a part or all of the sentences included in the document data; A computer program that causes a computer to execute a logic check step of comparing a sentence included in one of a plurality of mutually associated document data with a sentence included in another document data, and presenting a warning if a contradictory relationship is detected between a sentence included in one of the plurality of document data and a sentence included in another document data.

Images