Re-learning system, re-learning method, and program

Abstract

To realize a large-scale language model useful for business support systems. [Solution] A large-scale language model acquisition unit (101) of a re-learning system (1) acquires a pre-trained large-scale language model. A training data acquisition unit (102) acquires training data for re-learning the large-scale language model, which is created based on activity data indicating a user's activity performed in a business support system that supports the user's business. A re-learning unit (103) creates a business support model specific to the business support system by re-learning the large-scale language model based on the training data.

Description

[Technical field]

[0001] The present disclosure relates to a re-learning system, a re-learning method, and a program. [Background technology]

[0002] Conventionally, a business support system that supports a user's business is known. For example, Patent Document 1 describes, as an example of a business support system, groupware that supports communication between users belonging to an organization. The business support system of Patent Document 1 inputs data indicating a usage pattern of the groupware by a user to a trained model that estimates whether or not a no-input period occurs during which the user does not perform an operation. The trained model estimates whether or not a no-input period occurs based on the data. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2023-091326 A Summary of the Invention [Problem to be solved by the invention]

[0004] However, large-scale language models used in the field of natural language processing are required to analyze a very large number of languages, unlike trained models that perform simple classification such as those in Patent Document 1. For this reason, conventional techniques such as those in Patent Document 1 have not been able to realize large-scale language models that are useful for business support systems.

[0005] One of the objectives of the present disclosure is to realize a large-scale language model that is useful for business support systems. [Means for solving the problem]

[0006] The re-learning system of the present disclosure includes a large-scale language model acquisition unit that acquires a pre-trained large-scale language model, a training data acquisition unit that acquires training data for re-learning the large-scale language model created based on activity data indicating a user's activity performed in a business support system that supports the user's business, and a re-learning unit that creates a business support model specific to the business support system by performing re-learning of the large-scale language model based on the training data. Effect of the Invention

[0007] The present disclosure makes it possible to realize a large-scale language model that is useful for business support systems. [Brief description of the drawings]

[0008] [Figure 1] FIG. 2 is a diagram illustrating an example of a hardware configuration of a re-learning system. [Diagram 2] FIG. 13 is a diagram illustrating an example of a screen displayed on a user terminal. [Diagram 3] FIG. 2 is a diagram illustrating an example of a relationship between a large-scale language model and a task support model. [Figure 4] FIG. 2 is a diagram illustrating an example of functions realized by the re-learning system. [Diagram 5] FIG. 2 is a diagram illustrating an example of a training database. [Figure 6] FIG. 4 is a diagram illustrating an example of a user database. [Figure 7] FIG. 2 is a diagram illustrating an example of a process executed in the re-learning system. [Figure 8] FIG. 13 is a diagram illustrating an example of multi-stage relearning. [Figure 9] FIG. 13 is a diagram illustrating an example of a function realized in Modification 4. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0009] [1. Hardware configuration of the re-learning system] An example of an embodiment of a relearning system, a relearning method, and a program according to the present disclosure will be described. FIG. 1 is a diagram showing an example of a hardware configuration of a relearning system. For example, the relearning system 1 includes a relearning terminal 10 and a business support system 2. The business support system 2 includes a business support server 20 and a user terminal 30. Each of the relearning terminal 10, the business support server 20, and the user terminal 30 is connected to a network N such as the Internet or a LAN.

[0010] The relearning terminal 10 is a computer that executes relearning, which will be described later. For example, the relearning terminal 10 is a personal computer, a server computer, or a tablet. For example, the relearning terminal 10 includes a control unit 11, a memory unit 12, a communication unit 13, an operation unit 14, and a display unit 15. For example, the control unit 11 includes at least one processor. The memory unit 12 includes at least one of a volatile memory such as a RAM and a non-volatile memory such as a flash memory. The communication unit 13 includes at least one of a communication interface for wired communication and a communication interface for wireless communication. The operation unit 14 is an input device such as a mouse, a touch panel, or a keyboard. The display unit 15 is a liquid crystal or organic EL display.

[0011] The business support server 20 is a server computer. For example, the business support server 20 includes a control unit 21, a storage unit 22, and a communication unit 23. The hardware configurations of the control unit 21, the storage unit 22, and the communication unit 23 may be similar to those of the control unit 11, the storage unit 12, and the communication unit 13, respectively.

[0012] The user terminal 30 is a user's computer. For example, the user terminal 30 is a personal computer, a tablet, a smartphone, or a wearable terminal. For example, the user terminal 30 includes a control unit 31, a storage unit 32, a communication unit 33, an operation unit 34, and a display unit 35. The hardware configurations of the control unit 31, the storage unit 32, the communication unit 33, the operation unit 34, and the display unit 35 may be similar to those of the control unit 11, the storage unit 12, the communication unit 13, the operation unit 14, and the display unit 15, respectively.

[0013] In addition, the program stored in at least one of the storage unit 12, the storage unit 22, and the storage unit 32 may be supplied to at least one of the relearning terminal 10, the business support server 20, and the user terminal 30 via the network N. In addition, the hardware configuration of at least one of the relearning terminal 10, the business support server 20, and the user terminal 30 is not limited to the example of FIG. 1. For example, at least one of the relearning terminal 10, the business support server 20, and the user terminal 30 may include at least one of a reading unit (e.g., a memory card slot) that reads an information storage medium and an input / output unit (e.g., a USB terminal) that connects to the information storage medium. In this case, the program stored in the information storage medium may be supplied to at least one of the relearning terminal 10, the business support server 20, and the user terminal 30 via at least one of the reading unit and the input / output unit.

[0014] Furthermore, the computer included in the re-learning system 1 is not limited to the example in FIG. 1. The re-learning system 1 may include at least one computer. For example, the re-learning system 1 may include only the re-learning terminal 10. In this case, the business support system 2 exists outside the re-learning system 1. For example, the re-learning system 1 may include a computer not shown in FIG. 1.

[0015] [2. Overview of the Re-learning System] In this embodiment, the case where the business support system 2 is groupware is taken as an example. For example, the business support system 2 supports the business of a user who belongs to an organization such as a company or a government agency. The user uses a business support function possessed by the business support system 2. The business support function is a function possessed by a program developed for business support (for example, a program stored in the business support server 20 or a program stored in the user terminal 30).

[0016] The business support function may be various known functions. For example, the business support function may be a thread function for users to communicate in a thread, a schedule function for users to manage a schedule, a database function for users to manage a database, a file management function for users to manage files, or an email management function for users to manage email. The business support function may be other functions implemented in known groupware.

[0017] The business support system 2 may be a system that is not classified as groupware. For example, the business support system 2 may be a system that supports business unrelated to an organization, a system that does not have a function for linking with other products, or other systems. The business support system 2 may be any system that supports some business of the user.

[0018] 2 is a diagram showing an example of a screen displayed on the user terminal 30. For example, when a user operates the user terminal 30 to log in to the business support system 2, the user terminal 30 displays a business support screen SC for supporting the user's business on the display unit 35. The user can use any business support function from the business support screen SC. The example of FIG. 2 shows the business support screen SC when the user uses the thread function. The user can post any message to the thread displayed on the business support screen SC.

[0019] In addition to the threads shown in FIG. 2, various texts are registered in the business support system 2. If the business support system 2 could perform language analysis of the text registered in itself, it would be able to effectively support the user's business. For example, if the business support system 2 could analyze the text registered in a thread and create a summary of the thread, the user would be able to easily understand the topic in the thread. Similarly, language analysis of text is considered to be useful for business support functions other than the thread function. However, it would be very time-consuming for the operating company of the business support system 2 to create a large-scale language model for language analysis from scratch.

[0020] Therefore, the re-learning system 1 creates a business support model specific to the business support system 2 by executing re-learning of the pre-trained large-scale language model. The pre-trained large-scale language model may be a known model used in the field of natural language processing. For example, the large-scale language model may be a recurrent neural network type model, a long short-term memory network type model, or a Transformer type model.

[0021] In this embodiment, the large-scale language model is a Transformer-type model. For example, the large-scale language model may be a GPT (Generative Pre-trained Transformer) or BERT (Bidirectional Encoder Representations from Transformers) model. PaLM (Pathways Language Model) or LLaMA (Large Language Model Meta AI), which may be classified as a Transformer-type model, may be used as the large-scale language model. The large-scale language model may be another model that uses a machine learning technique.

[0022] For example, when a text is input, the large-scale language model divides the text into multiple tokens. The tokens are elements for the large-scale language model to understand the text. For example, the tokens are words, morphemes, phrases, or characters. The large-scale language model converts each of the multiple tokens into an embedding representation. The large-scale language model understands the context of the text input thereto based on the sequence of the embedding representations of each of the multiple tokens. The large-scale language model performs output according to its own task based on the context of the text. The large-scale language model executes this series of processes based on the parameters of each layer, such as the intermediate layer and the output layer. The large-scale language model may have hundreds of millions to trillions of parameters.

[0023] For example, the pre-learning of the large-scale language model is performed by another computer that is not included in the re-learning system 1. The operating company of the business support system 2 acquires the large-scale language model that has been pre-learned by another computer, either for a fee or free of charge. The large-scale language model can support the user's business to some extent through pre-learning. However, since the pre-learning of the large-scale language model is performed based on general-purpose data that is unrelated to business support, it is not possible to perform language analysis specific to business support. Therefore, the re-learning system 1 re-learns the large-scale language model based on data specific to business support.

[0024] Re-learning is a process for changing at least a part of a large-scale language model. For example, re-learning is adjustment of parameters of a large-scale language model. Re-learning may be replacement of some layers included in a large-scale language model. In this embodiment, a case where fine-tuning corresponds to re-learning is taken as an example. For example, fine-tuning is adjustment of parameters of a large-scale language model so that the parameters are adapted to a specific task. The meaning of fine-tuning may be a well-known meaning with various theories. Fine-tuning is not limited to the example of this embodiment.

[0025] Note that the re-learning is not limited to fine tuning. Re-learning may be any process for changing at least a part of the large-scale language model. For example, the re-learning may be transfer learning or distillation. There are various theories on the meaning of transfer learning or distillation, but transfer learning or distillation may have a known meaning among various theories. The re-learning may be a process other than fine tuning, transfer learning, and distillation. Hereinafter, the large-scale language model after re-learning is referred to as a business support model. The part simply describing the large-scale language model means a pre-trained large-scale language model (a large-scale language model before re-learning).

[0026] FIG. 3 is a diagram showing an example of the relationship between a large-scale language model and a business support model. For example, pre-learning of the large-scale language model M1 is performed so that the large-scale language model M1 can process a specific task. The task of this embodiment is different from a general task and is a task for business support. Hereinafter, the task of this embodiment is referred to as a business support task. In the example of FIG. 3, creating a summary of a thread corresponds to the business support task.

[0027] The business support task is not limited to creating a summary of a thread. For example, the business support task may be creating a summary of text other than the message of a thread (for example, the body of an email, a comment posted by a comment function of a database, or text of a website such as a help page), automatically creating some kind of text, machine translation, answering a user's question, extracting data registered in the business support system 2, classifying data registered in the business support system 2, analyzing user behavior, predicting user behavior, or other tasks.

[0028] For example, suppose that thread data indicating a thread posted to the business support system 2 is input to the large-scale language model M1. The large-scale language model M1 analyzes the text of the thread indicated by the thread data based on parameters adjusted by pre-learning. The large-scale language model M1 outputs summary data indicating a summary of the thread based on the analysis result of the text. Since the pre-learning of the large-scale language model M1 is performed based on general-purpose data, the large-scale language model M1 can create summaries with a certain degree of accuracy, but may not be able to create summaries specific to business support.

[0029] For example, in a general summary, assume that the beginning and end of the text to be summarized are important. In this case, pre-learning of the large-scale language model M1 is performed so that the large-scale language model M1 creates a summary based on the beginning and end of the text. The large-scale language model M1 creates a summary of the thread based on the beginning and end of the thread indicated by the thread data input to the large-scale language model M1.

[0030] On the other hand, in a summary for business support, the part including the date and time of the meeting and the text entered by the high-ranking user are more important than the beginning and end of the text. In this case, if these contents are not included at the beginning and end of the text, the summary created by the large-scale language model M1 may not be very useful. For example, the summary created by the large-scale language model M1 may not include the date and time of the meeting and may not include the text entered by the high-ranking user.

[0031] Therefore, the re-learning system 1 of this embodiment performs re-learning of the large-scale language model M1 based on activity data indicating the activity of a user in the business support system 2. The activity is an input by the user to the business support system 2. Since the activity data includes information useful for business support, a business support model M2 useful for business support is created by re-learning. In the example of FIG. 3, the thread data corresponds to the activity data. Examples of activity data other than the thread data will be described later.

[0032] For example, the business support model M2 has learned training data showing the relationship between a thread prepared for re-learning and a correct summary created from the thread. In this embodiment, a case is taken as an example in which the operating company of the business support system 2 creates a correct summary. For example, the training data is created so that, among messages posted to a thread, a message including the date and time of a meeting and a message from a user with a high position are included in the correct summary. By executing re-learning based on such training data, the business support model M2 becomes able to create summaries specific to business support.

[0033] As described above, the re-learning system 1 creates a business support model M2 specific to business support by re-learning the large-scale language model M1 that has been pre-learned based on general-purpose data. For example, it is not realistic for the operating company of the business support system 2 to learn the business support model M2 from scratch, but the re-learning system 1 can realize a business support model M2 useful for the business support system 2 by re-learning the large-scale language model M1. Details of the re-learning system 1 will be described below.

[0034] [3. Functions realized by the re-learning system] FIG. 4 is a diagram showing an example of functions realized by the re-learning system 1. As shown in FIG.

[0035] [3-1. Functions realized by re-learning terminal] For example, the relearning terminal 10 includes a data storage unit 100, a large-scale language model acquisition unit 101, a training data acquisition unit 102, and a relearning unit 103. The data storage unit 100 is realized by the storage unit 12. Each of the large-scale language model acquisition unit 101, the training data acquisition unit 102, and the relearning unit 103 is realized by the control unit 11.

[0036] [Data storage section] The data storage unit 100 stores data necessary for re-learning. For example, the data storage unit 100 stores a large-scale language model M1. The large-scale language model M1 includes parameters adjusted by pre-learning (parameters before re-learning) and a program for processing such as calculation of embedded representations.

[0037] The parameters of the large-scale language model M1 are referenced by the program of the large-scale language model M1. For example, the parameters are weighting coefficients and biases. The parameters may be various known parameters. The parameters are not limited to weighting coefficients and biases. For example, the parameters may be matrices referenced in the calculation of an embedding representation, positional encodings referenced in the encoding of token positions, or other parameters.

[0038] The program of the large-scale language model M1 indicates the internal processing of the large-scale language model M1. For example, the program of the large-scale language model M1 indicates the processing of an encoder that calculates an embedded representation, a decoder that creates an output specific to a business support task from the embedded representation, an output layer that performs a final output, and other layers. In these processes, parameters are referenced. There may be parameters specific to a certain business support task. The program of the large-scale language model M1 may be a publicly known program.

[0039] For example, when re-learning by the re-learning unit 103 is completed, the data storage unit 100 stores the re-learned business support model M2. In this embodiment, a case is taken as an example where the large-scale language model M1 and the re-learned business support model M2 are separate data. Instead of these being separate data, the large-scale language model M1 may be overwritten by the business support model M2. The data storage unit 100 stores not only the large-scale language model M1 and the business support model M2, but also the training database DB1.

[0040] FIG. 5 is a diagram showing an example of the training database DB1. The training database DB1 is a database in which training data necessary for re-learning the large-scale language model M1 is stored. For example, the training data includes an input part that is input to the large-scale language model M1 during re-learning, and an output part that becomes a correct answer during re-learning. The combination of the input part and the output part may be any combination that corresponds to the business support task. The training data may be prepared by manual input or by using a known tool.

[0041] The input portion of the training data may be the activity data itself, or may be other data created based on the activity data. Since the large-scale language model M1 performs language analysis, the activity data represents text entered by the user. The activity data may represent other activities besides the text. For example, the activity may represent a user's reaction (e.g., a like or emoji reaction), the number of views, the number of replies, the date and time of posting the text, or other content.

[0042] The output portion of the training data indicates content according to the business support task. For example, if the business support task is to output some kind of text such as a summary, the output portion of the training data is the correct text. If the business support task is to output some kind of classification (label), the output portion of the training data is the correct classification. If the business support task is to extract some kind of data, the output portion of the training data is the correct data (data that should be extracted from the input portion of the training data). If the business support task is to sort some kind of data, the output portion of the training data is the correct order.

[0043] In this embodiment, a case where thread summary creation corresponds to a business support task is taken as an example. Therefore, in the example of Fig. 5, the input part of the training data is thread data for re-learning. The thread data used as the input part of the training data may be thread data that actually exists in the business support system 2, or may be thread data showing the contents of a virtual thread prepared by the operating company of the business support system 2. In the example of Fig. 5, the output part of the training data is summary data showing a summary that is the correct answer created from the thread data that is the input part of the training data.

[0044] The training data is not limited to the example of this embodiment. The training data may indicate content corresponding to the business support task. The input portion of the training data may be activity data for learning, or data created based on activity data for learning. The output portion of the training data may indicate a correct answer as a processing result for the input portion of the training data. The input and output portions of the training data may indicate content corresponding to the business support task. An example of training data corresponding to a business support task is described below, but the training data is not limited to the example below.

[0045] For example, if the task is to create a summary of text other than messages in a thread (e.g., the body of an email, a comment posted through a comment function in a database, or text on a website such as a help page), the input portion of the training data indicates the text for which a summary is to be created, and the output portion of the training data in this case indicates a ground-truth summary created from that text.

[0046] For example, if the business support task is the automatic creation of some kind of text (e.g., a message posted to a thread, the body of an email, a comment posted via a comment function in a database, or text on a website such as a help page), the input portion of the training data indicates the conditions for the text to be created. In this case, the output portion of the training data indicates the text that is the correct answer according to the conditions.

[0047] For example, if the business support task is machine translation, the input portion of the training data is the text before translation (text in the original language). In this case, the output portion of the training data is the translated text that is the correct answer. For example, if the business support task is an answer to a user's question, the input portion of the training data indicates a question prepared for re-learning. In this case, the output portion of the training data indicates the answer that is the correct answer to the question.

[0048] For example, if the business support task is extraction of data registered in the business support system 2 (e.g., extraction of the sender address of an e-mail, extraction of the recipient address of an e-mail, extraction of the organization and name written in the signature of an e-mail, extraction of the organization and name from the image data of a business card), the input portion of the training data indicates the data to be extracted. In this case, the output portion of the training data indicates the correct content extracted from the data.

[0049] For example, if the business support task is classification of data registered in the business support system 2 (e.g., classification of topics on a thread, classification of the contents of an e-mail, or classification of files registered in the business support system 2), the input portion of the training data indicates data for relearning. The output portion of the training data in this case indicates the classification of the data. For example, if the business support task is analysis or prediction of a user's behavior in the business support system 2, the input portion of the training data indicates the user's behavior. The output portion of the training data in this case indicates the analysis or prediction results of the behavior.

[0050] The data stored in the data storage unit 100 is not limited to the above example. The data storage unit 100 may store any data. For example, the data storage unit 100 may store a re-learning program that indicates a series of processes in re-learning the large-scale language model M1. When a business support model M2 is created for each organization of a user, the data storage unit 100 may store a business support model M2 for each organization.

[0051] [Large-scale language model acquisition section] The large-scale language model acquisition unit 101 acquires the pre-trained large-scale language model M1. In this embodiment, since the pre-trained large-scale language model M1 is stored in the data storage unit 100, the large-scale language model acquisition unit 101 acquires the pre-trained large-scale language model M1 from the data storage unit 100. When the pre-trained large-scale language model M1 is stored in a computer or information storage medium other than the re-learning terminal 10, the large-scale language model acquisition unit 101 acquires the pre-trained large-scale language model M1 from the other computer or information storage medium.

[0052] [Training data acquisition section] The training data acquisition unit 102 acquires training data for relearning the large-scale language model M1, which is created based on activity data indicating a user's activity performed in the business support system 2 that supports the user's business. In this embodiment, since the training data is stored in the training database DB1, the large-scale language model acquisition unit 101 acquires the training data from the training database DB1. When the training data is stored in a computer or information storage medium other than the relearning terminal 10, the training data acquisition unit 102 acquires the training data from the other computer or information storage medium.

[0053] [Relearning section] The re-learning unit 103 re-learns the large-scale language model M1 based on the training data, thereby creating a business support model M2 specific to the business support system 2. In this embodiment, an example is given in which the re-learning unit 103 re-learns the large-scale language model M1 based on a supervised learning algorithm. The re-learning algorithm may be a known algorithm. For example, the re-learning unit 103 may re-learn the large-scale language model M1 based on a semi-supervised learning or unsupervised learning algorithm.

[0054] For example, the re-learning unit 103 performs re-learning of the large-scale language model M1 by adjusting parameters of the large-scale language model M1 based on a supervised learning algorithm so that an output portion of the training data is output from the large-scale language model M1 when an input portion of the training data is input to the large-scale language model M1. The re-learning unit 103 calculates a loss indicating an error between an output from the large-scale language model M1 when an input portion of the training data is input to the large-scale language model M1 and an output portion of the training data based on a known loss function. The re-learning unit 103 performs re-learning of the large-scale language model M1 by adjusting parameters of the large-scale language model M1 so that the calculated loss is small. The re-learning unit 103 may repeatedly perform re-learning of the large-scale language model M1 until the loss becomes less than a threshold value.

[0055] For example, the large-scale language model M1 during re-learning divides the text indicated by the input portion of the training data into multiple tokens. The large-scale language model M1 encodes each of the multiple tokens and converts it into a sequence of embedded representations. The large-scale language model M1 predicts the continuation as necessary based on the sequence of embedded representations, and calculates an output according to the business support task. The greater the difference between the output of the large-scale language model M1 and the output portion of the training data, the greater the loss. The smaller the difference between the output of the large-scale language model M1 and the output portion of the training data, the smaller the loss. This series of processes is executed based on the program of the large-scale language model M1 (e.g., the encoder, decoder, output layer, and other layers) and the current parameters of the large-scale language model M1.

[0056] For example, the input portion of the training data indicates text corresponding to the activity data. Furthermore, the output portion of the training data is a text that is a correct answer corresponding to the text. In this case, the business support task is a task of creating another text based on some text. When the re-learning unit 103 inputs the input portion of the training data to the large-scale language model M1, the large-scale language model M1 outputs the other text corresponding to the sequence of embedded expressions of the text indicated by the input portion. The re-learning unit 103 performs re-learning of the large-scale language model M1 based on the difference between the other text and the output portion of the training data.

[0057] For example, the input portion of the training data indicates text based on activity data. Furthermore, the output portion of the training data is a classification that is the correct answer for the text. In this case, the business support task is a task of classifying some text. When the re-learning unit 103 inputs the input portion of the training data to the large-scale language model M1, the large-scale language model M1 outputs an estimation result of the classification according to the sequence of the embedded expressions of the text indicated by the input portion. The re-learning unit 103 performs re-learning of the large-scale language model M1 based on the difference between the estimation result of the classification and the output portion of the training data.

[0058] In this embodiment, a case where thread summary creation corresponds to a business support task is taken as an example, so the re-learning unit 103 inputs thread data, which is an input portion of training data, to the large-scale language model M1. The large-scale language model M1 outputs summary data indicating a summary according to the sequence of embedded expressions of the thread indicated by the thread data. The re-learning unit 103 executes re-learning of the large-scale language model M1 based on the difference between the output summary data and the output portion of the training data. Similarly, for the various examples of training data described above, the re-learning unit 103 inputs the input portion of the training data to the large-scale language model M1, and executes re-learning of the large-scale language model M1 based on the difference between the output from the large-scale language model M1 and the output portion of the training data.

[0059] The algorithm used in the re-learning may be a known algorithm. For example, the re-learning unit 103 performs re-learning of the large-scale language model M1 based on the backpropagation method. The re-learning unit 103 may perform re-learning of the large-scale language model M1 based on an algorithm other than the backpropagation method. For example, the re-learning unit 103 may perform re-learning of the large-scale language model M1 based on the gradient descent method, the momentum method, the quasi-Newton method, the conjugate gradient method, the local search method, or other algorithms.

[0060] In this embodiment, the re-learning unit 103 executes a series of re-learning processes by executing a re-learning program stored in the data storage unit 100. The re-learning unit 103 records the large-scale language model M1 after re-learning as a business assistance model M2 in the data storage unit 100. The re-learning unit 103 transmits the business assistance model M2 to the business assistance server 20. The business assistance model M2 transmitted to the business assistance server 20 is made available for use by the user.

[0061] [3-2. Functions realized by the business support server] For example, the business support server 20 includes a data storage unit 200 and a business support unit 201. The data storage unit 200 is realized by the storage unit 22. The business support unit 201 is realized by the control unit 21.

[0062] [Data storage section] The data storage unit 200 stores data necessary for supporting a user's business. For example, the data storage unit 200 stores a user database DB2.

[0063] 6 is a diagram showing an example of the user database DB2. The user database DB2 is a database in which various data of users who use the business support system 2 is stored. For example, the user database DB2 stores the organization ID of the organization to which the user belongs, the user ID of the user, and activity data. Any data may be stored in the user database DB2. For example, the user database DB2 may store other data such as a user's login password.

[0064] The organization ID is the ID of an organization that has signed a contract with the business support system 2. The user ID is the ID of a user that belongs to the organization. For example, the user ID is used as a login account for a user to log in to the business support system 2. At least one user ID is associated with one organization ID. The activity data is data that indicates a user's activity. For example, the activity data indicates the date and time when the activity was performed and the specific content of the activity.

[0065] In this embodiment, posting a message to a thread corresponds to an activity, so the activity data indicates text, which is the specific content of the message. The activity data may include the date and time when the user posted the message to the thread, a reaction to the message, or other information. Every time a certain user of an organization posts a message to a thread, activity data is created by the business support unit 201, and the activity data is stored in the user database DB2 in association with the organization ID of the organization and the user ID of the user.

[0066] The data stored in the data storage unit 200 is not limited to the user database DB2. The data storage unit 200 may store any data. For example, the data storage unit 200 may store a re-learned business support model M2. The data storage unit 200 may store various data such as the contents of threads registered in the business support system 2, in addition to the activity data.

[0067] [Business Support Department] The business support unit 201 executes various processes for supporting the business of the user. For example, the business support unit 201 creates activity data based on operation data indicating the operation content of the user acquired from the user terminal 30. The business support unit 201 updates the user database DB2 so that the user ID of the user and the created activity data are associated with each other. The business support unit 201 transmits the activity data stored in the user database DB2 to the relearning terminal 10. Training data is created based on the activity data.

[0068] For example, the business support unit 201 supports a user's business based on the re-trained business support model M2. In this embodiment, since a case where a summary creation of a thread corresponds to a business support task is taken as an example, the business support unit 201 inputs thread data of a thread for which a user wishes to create a summary to the business support model M2. The re-trained business support model M2 calculates an embedded expression of the thread indicated by the thread data. The re-trained business support model M2 creates summary data indicating a summary according to the calculated embedded expression. The business support unit 201 transmits the summary data created by the re-trained business support model M2 to the user terminal 30. The business support unit 201 similarly inputs data indicating some text to the business support model M2 for other business support tasks, and supports the user's business based on the data output from the business support model M2.

[0069] [3-3. Functions realized on user terminals] For example, the user terminal 30 includes a data storage unit 300, a display control unit 301, and an operation reception unit 302. The data storage unit 300 is realized by the storage unit 32. The display control unit 301 and the operation reception unit 302 are realized by the control unit 31.

[0070] [Data storage section] The data storage unit 300 stores data necessary for a user to use the task support system 2. For example, the data storage unit 300 stores a browser. For example, the data storage unit 300 stores a program dedicated to the task support system 2.

[0071] [Display control section] The display control unit 301 causes various screens for business support to be displayed on the display unit 35. For example, the display control unit 301 causes a screen in the business support system 2 to be displayed on a browser. The display control unit 301 causes a screen in the business support system 2 to be displayed on a program dedicated to the business support system 2. After the business support model M2 is created, the user can use the business support model M2 from the business support screen SC. In the example of FIG. 2, when the user selects "Create summary", summary data of the displayed thread is created by the business support model M2.

[0072] [Operation reception section] The operation acceptance unit 302 accepts operations for each screen that the display control unit 301 causes to be displayed on the display unit 35. The operation acceptance unit 302 transmits, to the business support server 20, operation data indicating the content of the operation performed by the user.

[0073] [4. Processing performed by the re-learning system] Fig. 7 is a diagram showing an example of processing executed by the re-learning system 1. The control units 11, 21, and 31 execute programs stored in the storage units 12, 22, and 32, respectively, to thereby execute the processing in Fig. 7.

[0074] 7, the business support server 20 creates activity data based on operation data received from the user terminal 30 and stores the activity data in the user database DB2 (S1). The business support server 20 transmits the activity data stored in the user database DB2 to the re-learning terminal 10 (S2). The re-learning terminal 10 receives the activity data from the business support server 20 (S3). The re-learning terminal 10 stores training data created based on the activity data in the training database DB1 (S4).

[0075] The re-learning terminal 10 acquires the pre-trained large-scale language model M1 stored in the storage unit 12 (S5). The re-learning terminal 10 acquires an arbitrary number of training data from the training database DB1 (S6). The re-learning terminal 10 creates a business support model M2 by re-learning the pre-trained large-scale language model M1 based on the training data acquired in S6 (S7). The re-learning terminal 10 transmits the business support model M2 created in S7 to the business support server 20 (S8). The business support server 20 receives the business support model M2 from the re-learning terminal 10 (S9). The business support server 20 executes a process between the user terminal 30 and the user terminal 30 to allow the user to use the business support model M2 (S10), and the process ends.

[0076] [5. Summary of the embodiment] The re-learning system 1 of this embodiment acquires a pre-trained large-scale language model M1. The re-learning system 1 acquires training data for re-learning the large-scale language model M1, which is created based on activity data. The re-learning system 1 executes re-learning of the large-scale language model M1 based on the training data. For example, it is not realistic for the operating company of the business support system 2 to create the business support model M2 from scratch, but the re-learning system 1 can realize a business support model M2 useful for the business support system 2 by re-learning the pre-trained large-scale language model M1. For example, the re-learning system 1 can effectively support the user's business based on the business support model M2 in which the user's activity in the business support system 2 is learned, compared to the general-purpose large-scale language model M1. Since the large-scale language model M1 has basic language analysis capabilities, the re-learning system 1 can create a highly accurate business support model M2 even with a small amount of training data.

[0077] [6. Modifications] The present disclosure is not limited to the examples of the embodiments, and may be modified without departing from the spirit and scope of the present disclosure.

[0078] [6-1. Variation 1] In the embodiment, the training data including thread data for relearning and summary data that is the correct answer at the time of relearning is described as an example. The relearning system 1 can provide more effective business support by creating multiple business support models M2 based on training data created from various perspectives rather than performing relearning with training data created from one such perspective. Therefore, the relearning system 1 of the first modification performs multi-stage relearning based on training data created from various perspectives.

[0079] The training data acquisition unit 102 of the first modification acquires a plurality of training data created from different viewpoints. A viewpoint is a criterion when creating the training data. In other words, a viewpoint is a type of activity data used in creating the training data. In the first modification, as an example of a plurality of training data created from different viewpoints, training data created for each business support task and training data created for each organization will be described. Based on the training data created from these viewpoints, multi-stage re-learning is performed.

[0080] FIG. 8 is a diagram showing an example of multi-stage re-learning. The re-learning unit 103 of the first modification creates multiple task support models M2 by executing multi-stage re-learning of the large-scale language model M1 based on multiple training data acquired by the training data acquisition unit 102 of the first modification. For example, the re-learning unit 103 creates common task support models M2A1 and M2A2, which are the first-stage task support models M2, by executing the first-stage re-learning based on training data created from a certain viewpoint. Hereinafter, when the common task support models M2A1 and M2A2 are not distinguished from each other, they are simply referred to as the common task support model M2A. There may be only one common task support model M2A, or there may be three or more common task support models M2A.

[0081] For example, the re-learning unit 103 executes second-stage re-learning based on training data created from another perspective, thereby creating specific task support models M2B1, M2B2, which are the second-stage task support model M2. Hereinafter, when the specific task support models M2B1, M2B2 are not distinguished from each other, they are simply referred to as the common task support model M2B. There may be only one common task support model M2B, or there may be three or more common task support models M2B. The re-learning unit 103 may execute three or more stages of re-learning.

[0082] In the first modification, as in the embodiment, it is assumed that users belonging to each of the multiple organizations use the business support system 2. Furthermore, it is assumed that activity data of each user of the multiple organizations is stored in the user database DB2. For example, the training data acquisition unit 102 acquires common training data, which is training data common to the multiple organizations and is created based on the activity data of each of the multiple organizations, for each business support task performed by the business support model M2.

[0083] For example, suppose there are two task support tasks: thread summary creation and machine translation. In this case, the training data acquisition unit 102 acquires common training data created based on thread data (an example of activity data) of each of multiple organizations to create a common task support model M2A1 that performs thread summary creation. For example, the training data acquisition unit 102 acquires common training data created based on activity data indicating some text input by a user belonging to each of multiple organizations to create a common task support model M2A2 that performs machine translation.

[0084] For example, the re-learning unit 103 creates a common task support model M2A, which is a task support model M2 common to multiple organizations, by executing re-learning of the large-scale language model M1 based on common training data for each task support task. The common task support model M2A of the first modified example is a model specific to a certain task support task, but is not a model specific to a specific organization. The common task support model M2A can generically support the tasks of each of multiple organizations. The common task support model M2A is an intermediate model for creating a specific task support model M2B described later. The method of re-learning based on the common training data is similar to the re-learning based on the training data described in the embodiment.

[0085] For example, suppose there are two task support tasks: thread summary creation and machine translation. In this case, the re-learning unit 103 creates a common task support model M2A1 that creates thread summaries by re-learning the large-scale language model M1 based on common training data created from thread data (an example of activity data) of each of a plurality of organizations. The common task support model M2A1 that creates thread summaries has higher accuracy in creating summaries than the large-scale language model M1, but cannot create summaries specific to a particular organization.

[0086] For example, the re-learning unit 103 creates a common task support model M2A2 that performs machine translation by re-learning the large-scale language model M1 based on common training data created from activity data indicating some text input by users belonging to each of multiple organizations. The common task support model M2A2 that performs machine translation has a higher accuracy of machine translation than the large-scale language model M1, but cannot perform machine translation specific to a specific organization.

[0087] For example, the training data acquisition unit 102 acquires, for each organization, specific training data that is training data specific to the organization and that is created based on the activity data of the organization. In the first modification, a case will be described in which the activity data used to create the specific training data is the same as the activity data used to create the common training data, but these may be different.

[0088] For example, suppose there are two task support tasks: thread summary creation and machine translation. In this case, the training data acquisition unit 102 acquires specific training data created based on thread data of a first organization to create a specific task support model M2B1 that creates summaries specific to the first organization. The training data acquisition unit 102 acquires specific training data created based on thread data of a second organization to create a specific task support model M2B2 that creates summaries specific to another second organization.

[0089] For example, the training data acquisition unit 102 acquires specific training data created based on activity data indicating some text input by a user belonging to a first organization in order to create a specific task support model M2B3 that performs machine translation specific to the first organization. The training data acquisition unit 102 acquires specific training data created based on activity data indicating some text input by a user belonging to the second organization in order to create a specific task support model M2B4 that performs machine translation specific to another second organization.

[0090] For example, the re-learning unit 103 executes re-learning of the common task support model M2A for each organization based on the specific training data of the organization, thereby creating a specific task support model M2B which is a task support model M2 specific to the organization. The specific task support model M2B of the first modified example is a model specialized for a specific task support task of a specific organization. The specific task support model M2B can support a specific task of a specific organization. The specific task support model M2B is a final model provided to a user belonging to a specific organization. The method of re-learning based on the specific training data is similar to the re-learning based on the training data described in the embodiment.

[0091] For example, suppose there are two task support tasks: thread summary creation and machine translation. In this case, the re-learning unit 103 creates a specific task support model M2B1 that creates a summary specific to the first organization by re-learning the common task support model M2A1 for summary creation based on the specific training data created from the thread data of the first organization. The re-learning unit 103 creates a specific task support model M2B2 that creates a summary specific to the second organization by re-learning the common task support model M2A1 for summary creation based on the specific training data created from the thread data of the second organization.

[0092] For example, the re-learning unit 103 creates a specific task support model M2B3 that performs machine translation specific to the first organization by re-learning the common task support model M2A2 for machine translation based on specific training data created from activity data indicating some text input by a user belonging to the first organization. The re-learning unit 103 creates a specific task support model M2B4 that performs machine translation specific to the second organization by re-learning the common task support model M2A2 for machine translation based on specific training data created from activity data indicating some text input by a user belonging to the second organization.

[0093] In the above description, the case where there are two business support tasks and two organizations has been described as an example, but the number of business support tasks and organizations may be any number. For example, when there are three or more business support tasks, three or more common business support models M2A may be created. When there are three or more organizations, three or more specific business support models M2B may be created for each business support task.

[0094] For example, the re-learning unit 103 records the unique task support model M2B of each of the multiple organizations in the data storage unit 100. The re-learning unit 103 transmits the unique task support model M2B of each of the multiple organizations to the task support server 20. The task support server 20 associates the organization ID of each of the multiple organizations with the unique task support model M2B of the organization in the data storage unit 200. When a user belonging to a certain organization logs in to the task support system 2, the task support unit 201 supports the task of the user based on the unique task support model M2B associated with the organization ID of the organization.

[0095] When a user belonging to an organization for which the specific task support model M2B has not been created logs in to the task support system 2, the task support unit 201 may support the task of the user based on the common task support model M2A. In this case, the task support unit 201 may support the task of the user based on the specific task support model M2B associated with the organization ID of another organization (for example, an organization in the same industry or with the same number of employees) instead of the common task support model M2A.

[0096] The re-learning system 1 of the first modified example acquires multiple training data created from different viewpoints. The re-learning system 1 creates multiple business support models M2 by performing multi-stage re-learning of the large-scale language model M1 based on the multiple training data. This allows the re-learning system 1 to realize a more useful business support model M2. For example, the re-learning system 1 can realize more accurate business support through multi-stage re-learning.

[0097] Furthermore, the re-learning system 1 creates a common task support model M2A for each task based on common training data. The re-learning system 1 creates a specific task support model M2B for each organization based on the specific training data of the organization. This allows the re-learning system 1 to create a specific task support model M2B specialized for the tendencies of each organization after creating the common task support model M2A based on a larger amount of common training data, thereby efficiently creating a task support model M2 specific to each organization. For example, the re-learning system 1 can realize highly accurate task support according to a specific task and a specific organization.

[0098] [6-2. Variation 2] For example, the multi-stage re-learning method described in Modification 1 is not limited to the multi-stage re-learning method according to organization described in Modification 1. In Modification 2, as an example of another method of multi-stage re-learning, multi-stage re-learning according to business support tasks and business support functions is described. Note that when there are multiple business support systems 2, the business support function may be the business support system 2 itself, rather than a function possessed by a single business support system 2. For example, when there are multiple groupware, each groupware may correspond to a business support function.

[0099] For example, the training data acquisition unit 102 acquires common training data, which is training data common to multiple business support functions, created based on activity data of each of the multiple business support functions possessed by the business support system 2, for each business support task performed by the business support model M2.

[0100] For example, assume that there are two business support tasks: summary creation and machine translation. The summary creation in the second modification is not limited to threads, but is summary creation for the entire business support system 2. In this case, the training data acquisition unit 102 acquires common training data for summary creation created based on the activity data of each of a plurality of business support functions in order to create a common business support model M2A1 that performs summary creation. For example, the training data acquisition unit 102 acquires common training data for machine translation created based on the activity data of each of a plurality of business support functions in order to create a common business support model M2A2 that performs machine translation.

[0101] For example, the re-learning unit 103 creates a common task support model M2A, which is a task support model M2 common to a plurality of task support functions, by executing re-learning of the large-scale language model M1 based on the common training data for each task support task. The common task support model M2A of the second modification is a model specific to a certain task support task, but is not a model specific to a specific task support function. The common task support model M2A can generally support the tasks of a user who uses each of a plurality of task support functions. The common task support model M2A is an intermediate model for creating a specific task support model M2B described later. The method of re-learning based on the common training data is the same as the re-learning based on the training data described in the embodiment.

[0102] For example, suppose there are two business support tasks: summary creation and machine translation. In this case, the re-learning unit 103 creates a common business support model M2A1 that performs summary creation by executing re-learning of the large-scale language model M1 based on common training data for summary creation created from activity data of each of a plurality of business support functions. The common business support model M2A1 that performs summary creation has a higher accuracy of summary creation than the large-scale language model M1, but cannot perform summary creation specific to a specific business support function.

[0103] For example, the re-learning unit 103 creates a common task support model M2A1 that performs machine translation by re-learning the large-scale language model M1 based on common training data for machine translation created from activity data of each of a plurality of task support functions. The common task support model M2A2 that performs machine translation has a higher accuracy of machine translation than the large-scale language model M1, but cannot perform machine translation specific to a specific task support function.

[0104] For example, the training data acquisition unit 102 acquires, for each business support function, specific training data that is training data specific to that business support function and is created based on the activity data of that business support function. In the second modification, as in the first modification, a case will be described in which the activity data used to create the specific training data is the same as the activity data used to create the common training data, but these may be different.

[0105] For example, suppose there are two business support tasks: summary creation and machine translation. Furthermore, suppose there are two business support functions: a thread function and an email management function. In this case, the training data acquisition unit 102 acquires specific training data created based on thread data in order to create a specific business support model M2B1 that creates summaries specific to the thread function. The training data acquisition unit 102 acquires specific training data created based on activity data indicating the body of emails managed by the email management function in order to create a specific business support model M2B2 that creates summaries specific to the email management function.

[0106] For example, the training data acquisition unit 102 acquires specific training data created based on thread data to create a specific task support model M2B3 that performs machine translation specific to the thread function. The training data acquisition unit 102 acquires specific training data created based on activity data that indicates the text of emails managed by the email management function to create a specific task support model M2B4 that performs machine translation specific to the email management function.

[0107] For example, the re-learning unit 103 executes re-learning of the common task support model M2A for each task support function based on the specific training data of the task support function, thereby creating a specific task support model M2B, which is a task support model M2 specific to the task support function. The specific task support model M2B of the second modified example is a model specialized for a specific task support task of a specific task support function. The specific task support model M2B can support the task of a user who uses a specific task support function. The specific task support model M2B is a final model provided to a user who uses a specific task support function. The method of re-learning based on the specific training data is similar to the re-learning based on the training data described in the embodiment.

[0108] For example, suppose there are two task support tasks, summary creation and machine translation. Furthermore, suppose there are two task support functions, a thread function and an email management function. In this case, the re-learning unit 103 re-learns the common task support model M2A1 for summary creation based on the specific training data created from the thread data, thereby creating a specific task support model M2B1 that creates a summary specific to the thread function. The re-learning unit 103 re-learns the common task support model M2A1 for summary creation based on the specific training data created from the activity data of a user who uses the email management function, thereby creating a specific task support model M2B2 that creates a summary specific to the email management function.

[0109] For example, the re-learning unit 103 creates a specific task support model M2B3 that performs machine translation specific to the thread function by executing re-learning of the common task support model M2A2 for machine translation based on specific training data created from thread data. The re-learning unit 103 creates a specific task support model M2B4 that performs machine translation specific to the email management function by executing re-learning of the common task support model M2A2 for machine translation based on specific training data created from activity data of a user who uses the email management function.

[0110] In the above description, an example has been given in which there are two business support tasks and two business support functions, but the number of business support tasks and business support functions may be any number. For example, when there are three or more business support tasks, three or more common business support models M2A may be created. When there are three or more business support functions, three or more specific business support models M2B may be created for each business support task.

[0111] For example, the re-learning unit 103 records the specific task support model M2B for each of the multiple task support functions in the data storage unit 100. The re-learning unit 103 transmits the specific task support model M2B for each of the multiple task support functions to the task support server 20. The task support server 20 associates each of the multiple task support functions with the specific task support model M2B for the task support function and records them in the data storage unit 200. When a user uses a certain task support function, the task support unit 201 supports the task of the user based on the specific task support model M2B associated with the task support function.

[0112] When a user uses a business support function for which the specific business support model M2B has not been created, the business support unit 201 may support the business of the user based on the common business support model M2A. In this case, the business support unit 201 may support the business of the user based on the specific business support model M2B associated with another business support function, rather than the common business support model M2A.

[0113] The re-learning system 1 of the second modification creates a common task support model M2A by executing re-learning of the large-scale language model M1 based on common training data for each task support task. The re-learning system 1 creates a specific task support model M2B, which is a task support model M2 specific to each task support function, by executing re-learning of the common task support model M2A based on specific training data for each task support function. As a result, the re-learning system 1 can create the common task support model M2A based on more common training data, and then create the specific task support model M2B specialized for the tendency of each task support function, so that the task support model M2 specific to each task support function can be created efficiently. For example, the re-learning system 1 can realize highly accurate task support according to a specific task support task and a specific task support function.

[0114] [6-3. Variation 3] For example, in Modification 3, multi-stage re-learning according to business support functions and organizations will be described as another example of the multi-stage re-learning described in Modification 1. The number of business support tasks in Modification 3 may be one.

[0115] For example, the training data acquisition unit 102 acquires common training data, which is training data common to multiple organizations and is created based on activity data of each of the multiple organizations, for each business support function possessed by the business support system 2.

[0116] For example, assume that there are two task support functions, a thread function and an email management function. In this case, the training data acquisition unit 102 acquires common training data created based on activity data indicating texts input by users belonging to each of multiple organizations in the thread function in order to create a common task support model M2A1 for the thread function. For example, the training data acquisition unit 102 acquires common training data created based on activity data indicating texts input by users belonging to each of multiple organizations in the email management function in order to create a common task support model M2A2 for the email management function.

[0117] For example, the re-learning unit 103 creates a common task support model M2A common to multiple organizations by executing re-learning of the large-scale language model M1 based on common training data for each task support function. The common task support model M2A of variant example 3 is a model specialized for the task support system 2, but is not a model specific to a specific organization. The common task support model M2A can generally support the tasks of each of the multiple organizations. The common task support model M2A is an intermediate model for creating a specific task support model M2B described later. The method of re-learning based on the common training data is similar to the re-learning based on the training data described in the embodiment.

[0118] For example, suppose there are two business support functions, a thread function and an email management function. In this case, the re-learning unit 103 creates a common business support model M2A1 for the thread function by executing re-learning of the large-scale language model M1 based on common training data created from activity data (activity data showing texts entered by users in the thread function) of each of a plurality of organizations. The common business support model M2A1 for the thread function is capable of business support specific to the thread function more than the large-scale language model M1, but is not capable of business support specific to a specific organization.

[0119] For example, the re-learning unit 103 creates a common task support model M2A2 for the email management function by executing re-learning of the large-scale language model M1 based on common training data created from activity data (activity data showing texts entered by users in the email management function) of each of a plurality of organizations. The common task support model M2A2 for the email management function is capable of providing task support specific to the email management function more than the large-scale language model M1, but is not capable of providing task support specific to a specific organization.

[0120] For example, the training data acquisition unit 102 acquires, for each organization, specific training data that is training data specific to the organization and that is created based on the activity of the organization. In the third modification, a case will be described in which the activity data used to create the specific training data is the same as the activity data used to create the common training data, but these may be different.

[0121] For example, suppose there are two business support functions, a thread function and an email management function. In this case, the training data acquisition unit 102 acquires specific training data created based on activity data of a first organization in order to create a specific business support model M2B1 for the thread function of the first organization. The training data acquisition unit 102 acquires specific training data created based on activity data of the second organization in order to create a specific business support model M2B2 for the thread function of another second organization.

[0122] For example, the training data acquisition unit 102 acquires specific training data created based on activity data of a first organization to create a specific task support model M2B3 for an email management function of the first organization. The training data acquisition unit 102 acquires specific training data created based on activity data of the second organization to create a specific task support model M2B4 for an email management function of another second organization.

[0123] For example, the re-learning unit 103 performs re-learning of the common task support model M2A for each organization based on the specific training data of the organization, thereby creating a specific task support model M2B which is a task support model M2 specific to the organization. The specific task support model M2B of the modified example 3 is a model specialized for a specific organization. The specific task support model M2B is a model for a user belonging to a specific organization to use a specific task support function. The specific task support model M2B is a final model provided to a user belonging to a specific organization. The method of re-learning based on the specific training data is similar to the re-learning based on the training data described in the embodiment.

[0124] For example, assume that there are two task support functions, a thread function and an email management function. In this case, the re-learning unit 103 creates a specific task support model M2B1 capable of task support specific to the first organization by executing re-learning of the common task support model M2A1 for the thread function based on the specific training data created from the activity data of the first organization. The re-learning unit 103 creates a specific task support model M2B2 capable of task support specific to the second organization by executing re-learning of the common task support model M2A1 for the thread function based on the specific training data created from the activity data of the second organization.

[0125] For example, the re-learning unit 103 creates a specific task support model M2B3 capable of providing task support specific to the first organization by executing re-learning of the common task support model M2A2 for the e-mail management function based on the specific training data created from the activity data of the first organization. The re-learning unit 103 creates a specific task support model M2B4 capable of providing task support specific to the second organization by executing re-learning of the common task support model M2A2 for the e-mail management function based on the specific training data created from the activity data of the second organization.

[0126] In the above description, an example is given in which there are two business support functions and two organizations, but the number of business support functions and organizations may be any number. For example, when there are three or more business support functions, three or more common business support models M2A may be created. When there are three or more organizations, three or more specific business support models M2B may be created for each business support function.

[0127] For example, the re-learning unit 103 records the unique task support model M2B of each of the multiple organizations in the data storage unit 100. The re-learning unit 103 transmits the unique task support model M2B of each of the multiple organizations to the task support server 20. The task support server 20 associates the organization ID of each of the multiple organizations with the unique task support model M2B of the organization in the data storage unit 200. When a user belonging to a certain organization uses a certain task support function, the task support unit 201 supports the task of the user based on the unique task support model M2B for the task support function associated with the organization ID of the organization.

[0128] When a user belonging to an organization for which the specific task support model M2B has not been created logs in to the task support system 2, the task support unit 201 may support the task of the user based on the common task support model M2A. In this case, the task support unit 201 may support the task of the user based on the specific task support model M2B associated with the organization ID of another organization (for example, an organization in the same industry or with the same number of employees) instead of the common task support model M2A.

[0129] The re-learning system 1 of the third modified example creates a common task support model M2A by executing re-learning of the large-scale language model M1 based on common training data. The re-learning system 1 creates a specific task support model M2B for each organization by executing re-learning of the common task support model M2A based on the specific training data of the organization. As a result, the re-learning system 1 can create a specific task support model M2B specialized for the tendency of each organization to which a user who uses a specific task support function belongs after creating the common task support model M2A based on a larger amount of common training data, and therefore can efficiently create a task support model M2 specific to each organization. For example, the re-learning system 1 can realize highly accurate task support according to a specific task support function and a specific organization.

[0130] [6-4. Variation 4] For example, in the first to third modified examples, a plurality of task support models M2, such as a common task support model M2A and a specific task support model M2B, are created. Each of the plurality of task support models M2 may be used in data expansion of the training data.

[0131] FIG. 9 is a diagram showing an example of functions realized in the fourth modification. For example, the re-learning terminal 10 in the fourth modification includes a training data creation unit 104. The training data creation unit 104 is realized by the control unit 11. The training data creation unit 104 creates new training data based on each of the multiple training data and each of the multiple business support models M2. For example, the training data creation unit 104 inputs an input portion of the training data used in the re-learning of the business support model M2 to each of the multiple business support models M2. The business support model M2 performs output according to the input portion.

[0132] For example, the training data creation unit 104 acquires the input portion of the training data input to the business support model M2 as the input portion of the new training data. The training data creation unit 104 acquires the output from the business support model M2 as the output portion of the new training data. The output of the new training data does not have to be the output from the business support model M2 itself. For example, data in which a part of the output from the business support model M2 has been changed may be the output portion of the new training data. The training data creation unit 104 creates a pair of the input portion and the output portion as new training data. The training data creation unit 104 stores the new training data in the training database DB1.

[0133] The re-learning unit 103 of the fourth modification executes further re-learning of at least one of the multiple business support models M2 based on new training data. The re-learning unit 103 may execute re-learning of all business support models M2 based on new training data, or may execute re-learning of only some of the business support models M2. The further re-learning differs from the re-learning of the embodiment and the first to third modifications in that new training data is used, but the method of re-learning may be the same as the embodiment and the first to third modifications.

[0134] The re-learning system 1 of the fourth modified example creates new training data based on each of the multiple training data and each of the multiple business support models M2. The re-learning system 1 executes further re-learning of at least one of the multiple business support models M2 based on the new training data. This allows the re-learning system 1 to realize data expansion using each of the multiple business support models M2, thereby realizing effective re-learning.

[0135] [6-5. Variation 5] For example, even if the re-learning system 1 attempts to create a business support model M2 specific to a certain organization, there may be a case where the organization does not have a sufficient amount of activity data. For this reason, activity data of another organization may be used to create the business support model M2 specific to the organization. In the fifth modification, the organization for which the business support model M2 is created is referred to as a first organization. The other organization whose activity data is used to create the business support model M2 specific to the first organization is referred to as a second organization.

[0136] The training data acquiring unit 102 of the fifth modified example acquires first organization training data, which is training data created based on activity data of a first organization, and second organization training data, which is training data created based on activity data of a second organization related to the first organization. The second organization is another organization related to the first organization. For example, the second organization is another organization having the same industry or employee size as the first organization.

[0137] In the training database DB1 of the fifth modification, each piece of training data is associated with the organization ID of the organization to which the user who performed the activity indicated by the activity data used in creating the training data belongs. Furthermore, the association between the first organization and the second organization is defined in the data storage unit 100. For example, if the second organization is another organization in the same industry as the first organization, the industry of each organization is defined in the data storage unit 100. If the second organization is another organization with the same employee scale as the first organization, the employee scale of each organization is defined in the data storage unit 100.

[0138] The re-learning unit 103 of the fifth modified example creates a task support model M2 specific to the first organization based on the first organization training data and the second organization training data. For example, the re-learning unit 103 creates the task support model M2 specific to the first organization by executing re-learning of the large-scale language model M1 based on the first organization training data and the second organization training data. When the common task support model M2A is created as in the first to third modified examples, the re-learning unit 103 creates a specific task support model M2B of the first organization by executing re-learning of the common task support model M2A based on the first organization training data and the second organization training data.

[0139] The re-learning system 1 of the fifth modified example creates a business support model M2 specific to the first organization based on the first organization training data and the second organization training data. This allows the re-learning system 1 to efficiently create the business support model M2 specific to the first organization. For example, even if the first organization is a newly registered organization in the business support system 2 and there is not a sufficient amount of activity data, the re-learning system 1 uses the second organization training data based on the activity data of the second organization, so that the accuracy of the business support model M2 specific to the first organization can be improved.

[0140] [6-6. Variation 6] For example, the second organization in the fifth modification may be another organization that has been using the business support system 2 for a longer period of time than the first organization. The data storage unit 100 in the sixth modification stores data indicating the period of use of each organization. The re-learning terminal 10 can identify which organization corresponds to the second organization by referring to the data.

[0141] The training data acquisition unit 102 of the sixth modified example acquires first organization training data of a first organization newly registered in the business support system 2 and second organization training data of a second organization that has been using the business support system 2 for a longer period than the first organization. The training data acquisition unit 102 does not acquire training data for creating a business support model M2 specific to the first organization based on activity data of another organization that has been using the business support system 2 for a shorter period than the first organization.

[0142] The re-learning unit 103 of the sixth modification creates a business support model M2 specific to a first organization newly registered in the business support system 2. The re-learning unit 103 creates the business support model M2 specific to the first organization based on second-organization training data of a second organization that has been using the business support system 2 for a longer period than the first organization, without using training data based on activity data of another organization that has been using the business support system 2 for a shorter period than the first organization. Although the method of acquiring the second-organization training data is different from that of the fifth modification, the method of re-learning based on the first-organization training data and the second-organization training data is the same as that of the fifth modification.

[0143] The re-learning system 1 of the sixth modified example creates a business support model M2 specific to the first organization newly registered in the business support system 2, based on the first organization training data of the first organization newly registered in the business support system 2 and the second organization training data of the second organization that has been using the business support system 2 for a longer period than the first organization. Since the longer the period of use, the more activity data there is, the more efficiently the re-learning system 1 can create the business support model M2 specific to the first organization newly registered in the business support system 2.

[0144] [6-7. Variation 7] For example, data other than the activity data to be processed may be input to the business support model M2. In the seventh modification, as in the embodiment, a case where the business support task is to create a summary of a thread is taken as an example. For example, a thread may have keywords that the user wants to include in the summary. In this case, not only the thread data of the thread for which a summary is to be created, but also keywords specified by the user may be input to the business support model M2. In the seventh modification, a business support model M2 capable of handling such input is created.

[0145] The training data acquisition unit 102 of the seventh modified example acquires training data indicating the entire activity indicated by the activity data and a portion selected from the activity data. The portion selected from the activity data is a portion that particularly requires re-learning. For example, the operating company of the business support system 2 selects a portion from the entire activity and performs annotation. The input portion of the training data is a pair of the entire activity indicated by the activity data and the annotated portion. The output portion of the training data may be the same as that of the embodiment.

[0146] The re-learning unit 103 of the seventh modification creates a business support model M2 by re-learning the large-scale language model M1 based on training data indicating the whole and the part. Although the input part of the training data is different from that of the embodiment, the re-learning method is the same as that of the embodiment. The re-learning unit 103 re-learns the large-scale language model M1 so that when the whole and the part indicated by the input part of the training data are input to the large-scale language model M1, the large-scale language model M1 outputs the output part of the training data.

[0147] For example, when a user uses the business support model M2, the user specifies a portion corresponding to the portion. When the business support task is to create a summary of a thread, the business support unit 201 inputs thread data of the thread for which a summary is to be created and keywords specified by the user from within the thread to the business support model M2. The business support model M2 outputs summary data based on the embedded representation of these texts. The summary indicated by the summary data includes the keywords specified by the user.

[0148] The portion selected from the activity data is not limited to the above examples. The portion may be a portion corresponding to the business support task. For example, if the business support task is creating an answer to a question, the portion may be the content of the question. If the business support task is creating a text, the portion may be a keyword to be included in the text. Alternatively, the portion may have some correlation with the output portion of the training data.

[0149] The re-learning system 1 of the seventh modified example obtains training data representing the entire activity represented by the activity data and a portion selected from the activity data. The re-learning system 1 creates a business support model M2 by re-learning the large-scale language model M1 based on the training data representing the entire activity and the portion. This allows the re-learning system 1 to create a more flexible business support model M2.

[0150] [6-8. Other Modifications] For example, the above modifications may be combined.

[0151] For example, the functions described as being realized by the relearning terminal 10 may be realized by the business support server 20. The functions described as being realized by the relearning terminal 10 may be shared among multiple computers. [Explanation of symbols]

[0152] 1 Re-learning system, 2 Business support system, 10 Re-learning terminal, 11, 21, 31 Control unit, 12, 22, 32 Memory unit, 13, 23, 33 Communication unit, 20 Business support server, 30 User terminal, 14, 34 Operation unit, 15, 35 Display unit, N Network, 100 Data storage unit, 101 Large-scale language model acquisition unit, 102 Training data acquisition unit, 103 Re-learning unit, 104 Training data creation unit, 200 Data storage unit, 201 Business support unit, 300 Data storage unit, 301 Display control unit, 302 Operation reception unit, DB1 Training database, DB2 User database, M1 Large-scale language model, M2 Business support model, M2A, M2A1, M2A2 Common business support model, M2B, M2B1, M2B2, M2B3, M2B4 Specific business support model, SC Business support screen.

Claims

1. A large-scale language model acquisition unit for acquiring a pre-trained large-scale language model; a training data acquisition unit that acquires common training data, which is activity data indicating the activity of the user performed in the business support system for each business support task performed by a business support model specific to the business support system that supports a user's business, and which is training data common to the multiple organizations and created based on the activity data of each of the multiple organizations; a re-learning unit that re-learns the large-scale language model based on the common training data for each of the business support tasks to create a common business support model, which is the business support model common to the multiple organizations; Including, the training data acquisition unit acquires, for each of the organizations, specific training data that is created based on the activity data of the organization and is specific to the organization; the re-learning unit executes re-learning of the common task support model for each of the organizations based on the specific training data of the organization, thereby creating a specific task support model that is the task support model specific to the organization; Re-learning system.

2. A large-scale language model acquisition unit for acquiring a pre-trained large-scale language model; a training data acquisition unit that acquires common training data, which is activity data indicating the activity of the user performed in the business support system for each business support task performed by a business support model specific to the business support system that supports a user's business, and which is training data common to the multiple business support functions that are created based on the activity data of each of the multiple business support functions that the business support system has; a re-learning unit that re-learns the large-scale language model for each of the task support tasks based on the common training data to create a common task support model, which is the task support model common to the plurality of task support functions; Including, the training data acquisition unit acquires, for each of the business support functions, specific training data that is the training data specific to the business support function and is created based on the activity data of the business support function; the re-learning unit executes re-learning of the common task support model for each of the task support functions based on the specific training data of the task support function, thereby creating a specific task support model which is the task support model specific to the task support function; Re-learning system.

3. A large-scale language model acquisition unit for acquiring a pre-trained large-scale language model; a training data acquisition unit that acquires common training data, which is training data common to a plurality of organizations, created based on activity data indicating the activity of the user in each of the plurality of organizations for each business support function of the business support system that supports a user's business; a re-learning unit that re-learns the large-scale language model for each of the business support functions based on the common training data to create a common business support model that is a business support model specific to the business support system and is common to the multiple organizations; Including, the training data acquisition unit acquires, for each of the organizations, specific training data that is created based on the activity of the organization and is specific to the organization; the re-learning unit executes re-learning of the common task support model for each of the organizations based on the specific training data of the organization, thereby creating a specific task support model that is the task support model specific to the organization; Re-learning system.

4. The re-learning system further includes a training data creation unit that creates new training data based on each of the plurality of training data and each of the plurality of business support models, The re-learning unit performs further re-learning of at least one of the plurality of business support models based on the new training data. The re-learning system according to any one of claims 1 to 3, further comprising:

5. the training data acquisition unit acquires first organization training data, which is the training data created based on the activity data of a first organization newly registered in the business support system, and second organization training data, which is the training data created based on the activity data of a second organization related to the first organization and which has been using the business support system for a longer period than the first organization; the re-learning unit creates the business support model specific to the first organization newly registered in the business support system by having the business support model learn not only the first organization training data but also the second organization training data. The re-learning system according to any one of claims 1 to 3.

6. the training data acquisition unit acquires the training data indicating the entirety of the activity indicated by the activity data and a portion selected from the activity data; the re-learning unit creates the task assistance model by re-learning the large-scale language model based on the training data indicating the whole and the part. The re-learning system according to any one of claims 1 to 3.

7. Obtain a large-scale pre-trained language model, acquire common training data, which is activity data indicating the activity of the user performed in the business support system for each business support task performed by a business support model specific to the business support system that supports a user's business, and which is training data common to the multiple organizations and created based on the activity data of each of the multiple organizations; creating a common task support model, which is the task support model common to the plurality of organizations, by executing re-learning of the large-scale language model based on the common training data for each task; A re-learning method, comprising: acquiring, for each of the organizations, specific training data that is created based on the activity data of the organization and is specific to the organization; for each of the organizations, re-learning the common task support model based on the specific training data of the organization, thereby creating a specific task support model that is the task support model specific to the organization; How to relearn.

8. A large-scale language model acquisition unit for acquiring a pre-trained large-scale language model; a training data acquisition unit that acquires common training data, which is activity data indicating the activity of the user performed in the business support system for each business support task performed by a business support model specific to the business support system that supports the user's business, and which is training data common to the multiple organizations and created based on the activity data of each of the multiple organizations; a re-learning unit that performs re-learning of the large-scale language model based on the common training data for each of the business support tasks, thereby creating a common business support model that is the business support model common to the multiple organizations; The computer functions as the training data acquisition unit acquires, for each of the organizations, specific training data that is created based on the activity data of the organization and is specific to the organization; the re-learning unit executes re-learning of the common task support model for each of the organizations based on the specific training data of the organization, thereby creating a specific task support model that is the task support model specific to the organization; program.

Images