Information processing apparatus, content generation system, control method, and control program

Abstract

To optimize use of two kinds of generation models subjected to machine learning.SOLUTION: An information processing apparatus according to the present invention has a reception unit for accepting input of a query, a determination unit for determining a generative model for executing generation of content corresponding to the query among first generation models equipped by a self apparatus and second generation models equipped by another apparatus, and presentation means for presenting, to a user, the content generated in accordance with determination by the determination unit.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present disclosure relates to an information processing device, a content generation system, a control method, and a control program. [Background technology]

[0002] There is known a technology for generating content using a machine-learned generative model (also called artificial intelligence (AI)). For example, the following Patent Document 1 discloses a terminal device that generates a response to a user's comment using an automatic response AI and notifies the user of the response. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2019-91387 A Summary of the Invention [Problem to be solved by the invention]

[0004] In the technology for generating content using a generative model, including the automatic response AI described in Patent Document 1, content generation is generally performed by a server rather than an information processing device used by a user. Using a server has the advantage of making it possible to generate highly accurate content using abundant computer resources.

[0005] However, when the server generates the content, there is a disadvantage that a waiting time occurs between when the user inputs a query to command the information processing device to generate the content and when the generated content is presented to the user. Another disadvantage is that the content cannot be generated in a state where communication with the server is not possible.

[0006] The above-mentioned disadvantages can be eliminated by providing a generative model in an information processing device, but the above-mentioned advantages cannot be enjoyed when using a generative model provided in an information processing device. An exemplary objective of the present disclosure is to provide a technology that enables optimization of the use of two types of generative models, a generative model provided in an information processing device and a generative model provided in another device. [Means for solving the problem]

[0007] An information processing device according to an exemplary aspect of the present disclosure includes a receiving means for receiving a query input, a determination means for determining which of a first generation model provided in the device itself and a second generation model provided in another device will be used to generate content in response to the query, and a presentation means for presenting to a user the content generated in accordance with the determination of the determination means.

[0008] A content generation system according to an exemplary aspect of the present disclosure includes a first device having a first generative model and a second device having a second generative model, where the first device determines which of the first and second generative models is to generate content in response to a query input to the first device, and presents the content generated in accordance with the determination to a user.

[0009] In a control method according to an exemplary aspect of the present disclosure, one or more processors execute a reception process for receiving a query input, a determination process for determining which of a first and a second generation model will be used to generate content in response to the query, and a presentation process for presenting to a user the content generated in accordance with the determination made in the determination process.

[0010] A control program according to an exemplary aspect of the present disclosure causes a computer to function as a receiving means for receiving a query input, a determination means for determining which of a first and a second generative model will be used to generate content in response to the query, and a presentation means for presenting to a user the content generated in accordance with the determination of the determination means. Effect of the Invention

[0011] According to an exemplary aspect of the present disclosure, an exemplary effect is achieved in that it becomes possible to optimize the use of two types of generative models: a generative model provided by an information processing device and a generative model provided by another device. [Brief description of the drawings]

[0012] [Figure 1] 1 is a block diagram showing a configuration of an information processing device according to the present disclosure. [Diagram 2] FIG. 4 is a flow diagram showing the flow of a control method according to the present disclosure. [Diagram 3] 1 is a block diagram showing a configuration of a content generation system according to the present disclosure. [Figure 4] FIG. 1 is a diagram showing an overview of another content generation system according to the present disclosure. [Diagram 5] FIG. 2 is a block diagram showing a configuration of a terminal device according to the present disclosure. [Figure 6] FIG. 2 is a block diagram showing a configuration of a generating device according to the present disclosure. [Figure 7] 5 is a diagram showing an example of content generation in the content generation system shown in FIG. 4. [Figure 8] 5 is a diagram showing another example of content generation in the content generation system shown in FIG. 4. [Figure 9] 7 is a diagram showing yet another example of content generation in the content generation system shown in FIG. [Figure 10] 6 is a diagram showing an example of a display screen displayed by the terminal device shown in FIG. 5. [Figure 11]6 is a flowchart showing the flow of processing executed by the terminal device shown in FIG. 5. [Figure 12] FIG. 1 is a block diagram showing a configuration of a computer that functions as an information processing device according to the present disclosure. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0013] The following are examples of embodiments of the present invention. However, the present invention is not limited to the exemplary embodiments shown below, and various modifications are possible within the scope of the claims. For example, an embodiment obtained by appropriately combining the technical means employed in each of the exemplary embodiments shown below may also be included in the scope of the present invention. In addition, an embodiment obtained by appropriately omitting a part of the technical means employed in each of the exemplary embodiments shown below may also be included in the scope of the present invention. In addition, the effects mentioned in each of the exemplary embodiments shown below are examples of effects expected in the exemplary embodiments, and do not define the scope of the present invention. In other words, an embodiment that does not exhibit the effects mentioned in each of the exemplary embodiments shown below may also be included in the scope of the present invention.

[0014] First Exemplary Embodiment A first exemplary embodiment, which is an example of an embodiment of the present invention, will be described in detail with reference to the drawings. This exemplary embodiment is the basic form of each exemplary embodiment described later. The scope of application of each technical means adopted in this exemplary embodiment is not limited to this exemplary embodiment. That is, each technical means adopted in this exemplary embodiment can be adopted in other exemplary embodiments included in this disclosure to the extent that no particular technical obstacle occurs. In addition, each technical means shown in the drawings referred to for explaining this exemplary embodiment can also be adopted in other exemplary embodiments included in this disclosure to the extent that no particular technical obstacle occurs.

[0015] (Configuration of information processing device 1) The configuration of the information processing device 1 will be described with reference to Fig. 1. Fig. 1 is a block diagram showing the configuration of the information processing device 1. As shown in Fig. 1, the information processing device 1 includes a receiving unit 101, a determining unit 102, and a presenting unit 103.

[0016] The reception unit 101 receives a query input. Here, the term "query" refers to a content generation instruction or generation command. Therefore, the term "query" in the following description can be replaced with a "generation instruction," a "generation command," or a "prompt." The format of the query is not particularly limited. For example, the reception unit 101 may receive a query input in text format, or may receive a query input in another format such as voice.

[0017] The determination unit 102 determines a generation model for executing generation of content according to a query received by the reception unit 101, from among a first generation model provided in the information processing device 1 and a second generation model provided in another device. That is, although not shown in Fig. 1, the information processing device 1 is provided with a first generation model. In addition, the information processing device 1 is configured to be able to communicate with another device that is provided with a second generation model.

[0018] The determination unit 102 may determine one of the first and second generation models as the generation model for executing content generation, or may determine both of them as the generation model for executing content generation. In addition, the criteria by which the determination unit 102 determines the generation model for executing content generation may be determined in advance.

[0019] The first and second generative models are both generative models that are machine-learned to generate content in response to an input query. As the first and second generative models, models corresponding to the content to be generated may be used.

[0020] For example, when a query written in a natural language is input and a sentence in the natural language is generated as content, a language model may be used as the first and second generation models. A language model is a model that uses machine learning to learn the arrangement of components (such as words) in a sentence or the arrangement of sentences in a piece of text. The first generation model and the second generation model may generate different types of content. For example, a language model that generates a sentence in a natural language may be used as the first generation model, while an image generation model that generates an image may be used as the second generation model.

[0021] The presentation unit 103 presents the content generated according to the determination by the determination unit 102 to the user of the information processing device 1. Specifically, the presentation unit 103 presents to the user the content generated by the first generation model, the content generated by the second generation model, or both of them.

[0022] The presentation mode of the content is not particularly limited. For example, the presentation unit 103 may present the generated content by displaying it on a display device, by outputting it as sound on an audio output device, or by printing it on a printer. The device that presents the content may be included in the information processing device 1, or may be an external device to the information processing device 1.

[0023] As described above, the information processing device 1 includes a receiving unit 101 that receives a query input, a determination unit 102 that determines which of a first generation model provided in the device itself and a second generation model provided in another device will be used to generate content corresponding to the received input query, and a presentation unit 103 that presents the content generated in accordance with the determination to a user.

[0024] According to the above configuration, when the determination unit 102 determines to have the second generation model generate content, the content is generated by another device equipped with the second generation model, and the generated content is presented to the user by the presentation unit 103. In this case, it is possible to realize presentation of content that takes advantage of the merit of having another device generate content. The merit in this case is, for example, that highly accurate content can be presented by having the content generated by another device capable of generating highly accurate content by utilizing abundant computer resources.

[0025] Furthermore, according to the above configuration, when the determination unit 102 determines to have the first generation model generate content, the content is generated by the first generation model included in the information processing device 1, and the generated content is presented to the user by the presentation unit 103. In this case, it is possible to realize presentation of content that takes advantage of the merit of generating content by the information processing device 1. Advantages in this case include, for example, that the content can be presented even in a situation where communication with other devices is not possible, or that the content can be presented in a shorter time than when the content is generated by other devices, etc.

[0026] In this way, the information processing device 1 has the effect of optimizing the use of two types of generative models, a first generative model provided in the information processing device 1 and a second generative model provided in another device. Note that "optimizing the use of two types of generative models" does not necessarily mean that the two types of generative models are used in an optimal manner. For example, using two types of generative models in a manner that is considered preferable in consideration of various conditions such as the content of an input query and the communication status with other devices can be said to "optimize the use of two types of generative models" even if that manner has some disadvantages.

[0027] (Control Program) The above-mentioned functions of the information processing device 1 can also be realized by a program. The control program according to this exemplary embodiment causes a computer to function as a receiving means for receiving a query input, a determining means for determining which of the first and second generating models is to be used to generate content corresponding to the query, and a presenting means for presenting the content generated according to the determination by the determining means to a user. Therefore, the control program according to this exemplary embodiment has the effect of optimizing the use of two types of generating models, the first and second generating models.

[0028] (Control method flow) The flow of the control method will be described with reference to Fig. 2. Fig. 2 is a flow diagram showing the flow of the control method. Note that the execution subject of each step in this control method may be a processor included in the information processing device 1, or may be a processor included in another device, or each step may be executed by a processor provided in a different device.

[0029] In S1 (reception process), at least one processor receives a query input.

[0030] In S2 (determination process), at least one processor determines, from the first and second generative models, which generative model will be used to generate content in response to the query received as an input in S1.

[0031] In S3 (presentation process), at least one processor presents the content generated according to the decision in S2 to the user.

[0032] As described above, in the control method according to the present exemplary embodiment, at least one processor executes a reception process for receiving a query input, a determination process for determining which of the first and second generation models will be used to generate content corresponding to the query, and a presentation process for presenting the content generated according to the determination in the determination process to a user. This provides an advantage in that it is possible to optimize the use of two types of generation models, the first and second generation models.

[0033] (Content Generation System Configuration) The configuration of the content generation system 5 will be described with reference to Fig. 3. Fig. 3 is a block diagram showing the configuration of the content generation system 5. As shown in the figure, the content generation system 5 includes a first device 2 having a first generation model 111 and a second device 3 having a second generation model 112.

[0034] The first device 2 determines which of the first and second generative models 111 and 112 is to be used to generate content in response to a query input to the first device 2, and presents the content generated according to the determination to the user. The first device 2 being equipped with the first generative model 111 means that the first generative model 111 is stored in a storage device accessible by the first device 2. The storage device may be one that the first device 2 is equipped with, or may be an external storage device to the first device 2.

[0035] More specifically, when the first device 2 decides to have the first generative model 111 generate content, it inputs the input query to the first generative model 111 to generate content, and presents the generated content to the user. On the other hand, when the first device 2 decides to have the second generative model 112 generate content, it transmits the input query to the second device 3. Then, the first device 2 acquires the content generated by the second generative model 112 from the second device 3, and presents the acquired content to the user.

[0036] The second device 3 generates content using a second generative model 112 included in the second device 3. The second device 3 being provided with the second generative model 112 means that the second generative model 112 is stored in a storage device accessible by the second device 3. The storage device may be one included in the second device 3, or may be an external storage device of the second device 3.

[0037] More specifically, when the second device 3 receives a query from the first device 2, the second device 3 inputs the query into the second generative model 112 to generate content, and transmits the generated content to the first device 2. The content transmitted by the second device 3 is presented to the user by the first device 2.

[0038] As described above, the content generation system 5 according to this exemplary embodiment includes the first device 2 having the first generation model 111 and the second device 3 having the second generation model 112, and the first device 2 determines which of the first generation model 111 and the second generation model 112 is to generate content in response to a query input to the first device 2, and presents the content generated according to the determination to the user. Therefore, the content generation system 5 has the effect of optimizing the use of two types of generation models, the first generation model 111 and the second generation model 112.

[0039] Second Exemplary Embodiment A second exemplary embodiment, which is an example of an embodiment of the present invention, will be described in detail with reference to the drawings. Components having the same functions as those described in the above exemplary embodiment will be given the same reference numerals, and their description will be omitted as appropriate. The scope of application of each technical means adopted in this exemplary embodiment is not limited to this exemplary embodiment. That is, each technical means adopted in this exemplary embodiment can be adopted in other exemplary embodiments included in this disclosure, as long as no particular technical hindrance occurs. In addition, each technical means shown in each drawing referred to for explaining this exemplary embodiment can be adopted in other exemplary embodiments included in this disclosure, as long as no particular technical hindrance occurs.

[0040] (Configuration of content generation system 5A) The configuration of the content generation system 5A will be described with reference to Fig. 4. Fig. 4 is a diagram showing an overview of the content generation system 5A. The content generation system 5A is a system having a function of presenting content generated by a generation model to a user of the content generation system 5A.

[0041] As shown in the figure, the content generation system 5A includes a terminal device 2A, a first generation model 111A, a generation device 3A, and a second generation model 112A. In the content generation system 5A, the terminal device 2A can communicate with the generation device 3A via any network.

[0042] Although three pairs of the generating device 3A and the second generating model 112A are shown in Fig. 4, the content generating system 5A may include at least one pair of the generating device 3A and the second generating model 112A. The content generating system 5A may include a plurality of terminal devices 2A. Fig. 4 shows an example in which the terminal device 2A is a smartphone. However, the terminal device 2A is not limited to a smartphone as long as it is a device capable of realizing the functions of each block shown in Fig. 5 described later.

[0043] The terminal device 2A includes a first generation model 111A. Moreover, the generation device 3A includes a second generation model 112A. When the terminal device 2A receives an input of a query, it determines a generation model to execute generation of content according to the input query from among the first generation model 111A and the second generation model 112A.

[0044] Here, when the terminal device 2A decides to have the first generative model 111A generate content, it inputs the query to the first generative model 111A to generate content, and presents the generated content to the user.

[0045] The advantage of having the first generation model 111A generate content is that, when the first generation model 111A is used, it is possible to generate content without communicating with an external device. Another advantage is that, when the first generation model 111A is used, the time required from content generation to presentation is relatively short. Another advantage is that, for example, since the first generation model 111A is provided in the terminal device 2A, it is easy to generate content that reflects the user-related information of the terminal device 2A.

[0046] On the other hand, when the terminal device 2A decides to have the second generation model 112A generate content, it transmits the input query to the generation device 3A. As will be described in detail later, the terminal device 2A decides which generation device 3A to transmit the query to. The generation device 3A that receives the query inputs the received query to the second generation model 112A to generate content, and transmits the generated content to the terminal device 2A. The terminal device 2A then presents the received content to the user.

[0047] Advantages of having the second generation model 112A generate content include, for example, the ability to generate content of higher quality than the first generation model 111A, the ability to generate content using information obtained from a data source that the terminal device 2A cannot access, and the ability to generate content in a data format that cannot be generated by the first generation model 111A.

[0048] Generally, the more explanatory variables a language model has, the higher the quality of the answer (content) it can generate, but the greater the computational power required to generate the answer. For this reason, when the first generation model 111A and the second generation model 112A are language models, it is preferable to set the number of explanatory variables of the first generation model 111A to be less than that of the second generation model 112A. This allows the terminal device 2A to quickly generate content using the first generation model 111A even if the computational power of the terminal device 2A is inferior to that of the generation device 3A. The terminal device 2A can also cause the generation device 3A, which has superior computational power, to generate high-quality content using the second generation model 112A.

[0049] (Configuration of terminal device 2A) The configuration of the terminal device 2A will be described with reference to FIG. 5. FIG. 5 is a block diagram showing the configuration of the terminal device 2A. As shown in the figure, the terminal device 2A includes a control unit 20A that controls each unit of the terminal device 2A in an integrated manner, and a storage unit 21A that stores various data used by the terminal device 2A. The first generation model 111A may be stored in the storage unit 21A. The terminal device 2A also includes a communication unit 22A for the terminal device 2A to communicate with other devices, an input unit 23A that accepts input of various data to the terminal device 2A, and an output unit 24A for the terminal device 2A to output various data. The functions of the output unit 24A and the input unit 23A may be realized by a touch panel. The output unit 24A may output sound. As shown in the figure, the control unit 20A includes a reception unit 101A, a determination unit 102A, a presentation unit 103A, a data acquisition unit 201A, and a request destination determination unit 202A. The functions of the various units included in the control unit 20A can also be realized by installing a predetermined application software (control program) in the terminal device 2A, for example.

[0050] The reception unit 101A receives an input of a query, similar to the reception unit 101 of the information processing device 1. The reception unit 101A may convert the data format of the input query into a predetermined data format. For example, when the data format of a query that can be input to the first generation model 111A is text format, and the query is input by voice, the reception unit 101A may convert the query, which is voice data, into text format. A processing block different from the reception unit 101A may be provided, and the processing block may be made to convert the data format.

[0051] The determination unit 102A, like the determination unit 102 of the information processing device 1, determines a generation model to be used to generate content corresponding to the query received by the reception unit 101A, from among the first generation model 111A included in the terminal device 2A and the second generation model 112A included in the generation device 3A. Various methods can be applied as a method for determining a generation model. Specific examples of the method for determining a generation model will be described later with reference to Figs. 7 to 9.

[0052] Furthermore, when the determination unit 102A determines that the first generation model 111A should generate the content, the determination unit 102A inputs the input query to the first generation model 111A to generate the content. On the other hand, when the determination unit 102A determines that the second generation model 112A should generate the content, the determination unit 102A transmits the input query to the generation device 3A to cause the generation device 3A to generate the content using the query and the second generation model 112A. Then, the determination unit 102A acquires the generated content from the generation device 3A.

[0053] The presentation unit 103A presents the content generated according to the determination of the determination unit 102A to the user, similar to the presentation unit 103 of the information processing device 1. For example, when the generated content is text data, the presentation unit 103A may present the content to the user by displaying or outputting the content as sound on the output unit 24A. The presentation unit 103A may also present the content to the user by transmitting the content to a printing device via the communication unit 22A and causing the printing device to print out the content.

[0054] The data acquisition unit 201A acquires various data used when the first generation model 111A or the second generation model 112A generates the content. For example, the data acquisition unit 201A may acquire user-related information related to the user to whom the content is to be presented. In this case, the determination unit 102A can generate content adapted to the user by inputting the user-related information to the first generation model 111A in addition to the query. Note that "adapted to the user" here means that at least a part of the user-related information is reflected. In addition, the user-related information may be related to the user itself, may be related to the query input by the user, or may be related to both of them.

[0055] The request destination determination unit 202A determines which of the multiple generating devices 3A is to generate the content, that is, determines the request destination for content generation. The method of determining the request destination may be determined appropriately. For example, the request destination determination unit 202A may determine the request destination according to a query input by a user. For example, when a medical-related keyword is included in the query, the request destination determination unit 202A may determine the request destination to be a generating device 3A that has high accuracy in generating medical-related content. Also, for example, when a mathematical-related keyword is included in the query, the request destination determination unit 202A may determine the request destination to be a generating device 3A that can generate content including a mathematical formula. Furthermore, for example, the request destination determination unit 202A may determine the request destination to be a generating device 3A according to a language in which the query is written. Also, the request destination determination unit 202A may allow the user to select the request destination. In this case, the request destination determination unit 202A determines the request destination according to the user's operation.

[0056] (Configuration of generating device 3A) The configuration of the generating device 3A will be described with reference to Fig. 6. Fig. 6 is a block diagram showing the configuration of the generating device 3A. As shown in the figure, the generating device 3A includes a control unit 30A that controls each unit of the generating device 3A, and a storage unit 31A that stores various data used by the generating device 3A. The generating device 3A also includes a communication unit 32A that allows the generating device 3A to communicate with other devices, an input unit 33A that accepts input of various data to the generating device 3A, and an output unit 34A that allows the generating device 3A to output various data. As shown in the figure, the control unit 30A includes a content generating unit 301A and a content transmitting unit 302A.

[0057] The content generation unit 301A causes the second generative model 112A to generate the content. More specifically, the content generation unit 301A causes the second generative model 112A to generate the content by inputting the query received from the terminal device 2A to the second generative model 112A.

[0058] The content transmission unit 302A transmits the content generated by the content generation unit 301A to the terminal device 2A via the communication unit 32A.

[0059] (Content generation example 1: Determining the generation model to be used depending on the communication situation) An example of content generation in the content generation system 5A will be described with reference to Fig. 7. Fig. 7 is a diagram showing an example of content generation in the content generation system 5A. Fig. 7 shows a content generation example A11 when the terminal device 2A and the generation device 3A cannot communicate with each other, and a content generation example A12 when the terminal device 2A and the generation device 3A can communicate with each other. In both generation examples A11 and A12, a query of "I would like to hear your recommended dishes in Italian" is input by voice to the terminal device 2A, but the answer to this query, i.e., the content, is different between generation examples A11 and A12.

[0060] Specifically, in the generation example A11, the terminal device 2A is in a state where it cannot receive radio waves from a base station of a mobile communication service, that is, it is outside the communication range. Therefore, the terminal device 2A cannot communicate with the generation device 3A, and cannot cause the generation device 3A to generate content. Therefore, in the generation example A11, the determination unit 102A of the terminal device 2A causes the first generation model 111A to generate content. Then, the presentation unit 103A presents the generated content (specifically, the answer to the input query) to the user by outputting the generated content by voice.

[0061] In this way, even if the terminal device 2A is in a state where it cannot communicate with the generation device 3A, it can generate content for the input query and present it to the user. The determination unit 102A may determine to have the first generation model 111A generate content not only when communication with the generation device 3A is not possible. The determination unit 102A may determine to have the first generation model 111A generate content when the communication situation is such that a series of processes from having the generation device 3A generate content to presenting the generated content to the user at the terminal device 2A is hindered or is expected to be hindered. For example, the determination unit 102A may determine to have the first generation model 111A generate content when communication with the generation device 3A is unstable or when the average communication speed with the generation device 3A is equal to or lower than a predetermined threshold.

[0062] On the other hand, in the generation example A12, the terminal device 2A is in a state in which it is possible to communicate with the generation device 3A. Therefore, in the generation example A12, the determination unit 102A of the terminal device 2A transmits the input query to the generation device 3A via the network, and causes the second generation model 112A to generate the content.

[0063] At this time, the determination unit 102A transmits the user's favorite dishes and the user's restaurant usage history to the generation device 3A as user-related information related to the user of the terminal device 2A. In this manner, the user-related information may include information indicating the user's attributes or information indicating the user's activity history. By transmitting the user-related information to the generation device 3A, the content generation unit 301A of the generation device 3A can use the received user-related information to cause the second generation model 112A to generate content adapted to the user.

[0064] For example, the content generation unit 301A may generate a query by inputting user-related information into the "xxxx" part of a template saying "Please generate an answer to the query zzzz for user xxxx," and inputting the received query into the "zzzz" part, and then inputting the generated query into the second generative model 112A to generate content.

[0065] The content generating unit 301A may use the second generation model 112A that has been trained in advance to generate content adapted to a user indicated in the user-related information by inputting a query and user-related information. In this case, the content generating unit 301A does not need to generate a query using a template or the like, and may input the query and the user-related information directly to the second generation model 112A.

[0066] In the generation example A12, the presentation unit 103A presents the generated content (specifically, the answer to the input query) to the user by outputting the generated content by voice. The opening part of the answer presented in the generation example A12 is the same as that of the generation example A11, but it differs from the answer in the generation example A11 in that it includes a sentence generated based on user-related information.

[0067] The content generating unit 301A may also generate content by inputting information other than the user-related information to the second generation model 112A. For example, the content generating unit 301A may input information detected by performing a search using the user-related information together with a query to the second generation model 112A. For example, if the user-related information includes location information indicating the user's location and a predetermined word such as "restaurant," the content generating unit 301A may perform a website search using the word and location information. Then, a sentence extracted from the detected website may be input to the second generation model 112A together with the query received from the terminal device 2A to generate content. This makes it possible to generate sentences related to restaurants accessible from the user's location as content to be presented to the user.

[0068] The user-related information may be input in advance by the user to the terminal device 2A. The data acquisition unit 201A may generate the user-related information from emails, schedules, posting history to a social networking service (SNS), and location information of the terminal device 2A stored in the storage unit 21A. When the user-related information is generated automatically, it is possible to save the user the trouble of inputting information. Furthermore, by reflecting past events that the user has forgotten or matters that the user has not noticed in the user-related information, it is possible to generate content that is meaningful to the user.

[0069] Furthermore, the determination unit 102A may use user-related information when causing the first generation model 111A to generate content. In this case, the determination unit 102A may generate a query reflecting the user-related information by using the above-mentioned template or the like, and input the query to the first generation model 111A to generate content. Furthermore, the determination unit 102A may generate the first content by using the first generation model 111A that has been trained in advance so as to be able to generate content adapted to the user. For example, when a language model is used as the first generation model 111A, the language model may be retrained using teacher data that associates content that the user judges to be preferable with various queries. The determination unit 102A is able to generate content adapted to the user by using such a language model as the first generation model 111A.

[0070] As described above, the determination unit 102A may determine a generation model for executing content generation according to the communication status with the generation device 3A. In this way, in addition to the effects of the information processing device 1, it is possible to obtain an effect that the content can be presented even when there is a problem with the communication status, and that the content generated by the generation device 3A can be presented when there is no problem with the communication status.

[0071] Here, the communication situation may improve after the first generation model 111A generates the content. In such a case, the determination unit 102A may determine to cause the second generation model 112A to generate the content as well. In addition to the effect of the information processing device 1, this provides an effect that the content generated by the generation device 3A can be presented when the communication situation improves.

[0072] As described above, the information processing device 1A includes a data acquisition unit 201A that acquires user-related information related to the user. When the determination unit 102A determines to cause the second generation model 112A to generate content, the determination unit 102A transmits the user-related information and a query to the generation device 3A to generate content according to the user-related information. This provides the effect of causing the generation device 3A to generate content adapted to the user and present the content, in addition to the effect provided by the information processing device 1.

[0073] (Content generation example 2: Determining which generative model to use based on the query) The determination unit 102A may determine a generation model that generates content according to the contents of the query. This configuration provides an effect that appropriate content can be generated according to the contents of the query, in addition to the effects of the information processing device 1. Hereinafter, an example of determining a generation model that generates content according to the contents of the query will be described with reference to FIG. 8.

[0074] Fig. 8 is a diagram showing another example of content generation in the content generation system 5A. Fig. 8 shows a content generation example A21 for a query "What is the meaning of the word 'applicant'?" and a content generation example A22 for a query "Create a study plan for the final exam." In this way, the content generation system 5A can also be used as an education support system.

[0075] In the generation example A21, a query 211A "Please tell me the meaning of the word 'applicant'" is displayed on the display unit of the terminal device 2A. This query is input by a user. In the generation example A21, the determination unit 102A of the terminal device 2A determines to cause the first generative model 111A to generate content, inputs the query 211A to the first generative model 111A, and causes the content to be generated. Then, the presentation unit 103A presents the generated content, i.e., text A212 of an answer to the query 211A, to the user by displaying the answer on the terminal device 2A.

[0076] On the other hand, in the generation example A22, a query A221 such as "Create a study plan until the final exam" is input. The determination unit 102A of the terminal device 2A determines to cause the second generation model 112A to generate content, transmits the query A221 and user-related information to the generation device 3A, and causes the second generation model 112A to generate content. Then, the presentation unit 103A displays the generated content, i.e., text A222 of an answer to the query A221, on the terminal device 2A, thereby presenting the answer to the user.

[0077] In addition, in the generation example A22, a link A223 for DL ​​(download) the teaching material is displayed on the display unit of the terminal device 2A. The teaching material is generated using the third generation model 113A. In this way, the generation device 3A may generate content using multiple generation models. The third generation model 113A may be included in the generation device 3A, or may be included in another device. In this way, the generation device 3A may generate content by using other systems (e.g., AI systems) in addition to the second generation model 112A. This makes it possible to present even more substantial content to the user.

[0078] As a method for realizing the above-mentioned distinction between the first generation model 111A and the second generation model 112A, for example, a method of determining a generation model for executing content generation according to a keyword included in a query can be applied. For example, a keyword such as "study plan", "study plan", "schedule", or a keyword similar in meaning to these may be set in advance. In this way, when a query such as A221 including any of the set keywords is input, the determination unit 102A can determine to use the second generation model 112A. On the other hand, when a query such as A211 including none of the set keywords is input, the determination unit 102A can determine to use the first generation model 111A.

[0079] The second generation model 112A used when a query containing a keyword related to the generation of a learning plan as described above is input is preferably a model suitable for generating a learning plan. For example, the second generation model 112A may be trained using teacher data that associates user information related to the learning of various users and a query requesting the generation of a learning plan for the user with a learning plan suitable for the user. This makes it possible to present a suitable learning plan.

[0080] In addition, for example, the determination unit 102A may determine to have the second generative model 112A generate content when the query contains keywords such as "details," "video," and "image," or keywords with similar meanings. This makes it possible to have the second generative model 112A generate content that is difficult to generate with the first generative model 111A (for example, detailed answers to technical questions and advanced mathematical problems, detailed explanations of complex concepts and theories such as physics and chemistry, content including images, etc.) and present it to the user.

[0081] Furthermore, when there are multiple available second generation models 112A, the determination unit 102A may determine which of the multiple second generation models 112A to generate content from, depending on a keyword included in the query. For example, when a query including keywords such as "mathematical formula" and "equation" is input, the determination unit 102A may determine that the second generation model 112A capable of generating a mathematical formula, explaining using a mathematical formula, and performing calculations may generate content. For example, when a query including keywords such as "chemical formula" and "chemical reaction" is input, the determination unit 102A may determine that the second generation model 112A capable of generating a chemical formula, explaining using a chemical formula, and performing calculations may generate content.

[0082] In addition, the determination unit 102A may determine to have the first generative model 111A generate content when a predetermined keyword is included in the query. For example, the determination unit 102A may determine to have the first generative model 111A generate content when a keyword such as "quickly" or "simply" or a keyword similar in meaning to these keywords is included in the query.

[0083] Also, for example, if a query contains both the keywords "word" and "meaning," the query is for the meaning of a word, and it is assumed that the difficulty of generating an answer is relatively low. Therefore, when a query contains these keywords or keywords with similar meanings, the determining unit 102A may determine to have the first generative model 111A generate content.

[0084] Furthermore, the determining unit 102A may determine to have the first generative model 111A generate content when the query contains two keywords, "grammar" and "check," or a keyword with a similar meaning to these keywords, and a sentence to be checked. This allows the process of checking for grammatical errors to be smoothly executed using the first generative model 111A.

[0085] Also, for example, the determination unit 102A may determine to have the first generation model 111A generate content when the query includes two keywords, "progress" and "check," or keywords with similar meanings. When the first generation model 111A is used, it is easy to use the user-related information stored in the terminal device 2A, so that content indicating the progress of the user's learning identified from the user-related information can be smoothly generated. In addition, the first generation model 111A can also generate advice according to the progress, messages encouraging appropriate review, and the like.

[0086] (Another example of how to determine a generative model) Furthermore, the determination unit 102A may determine a generation model that generates content according to a user. Here, the user may be a user who inputs a query, a user related to the input query, or a user to which content is to be presented. When determining a generation model that generates content according to a user, the data acquisition unit 201A may acquire user-related information related to the user. In this way, the determination unit 102A can determine a generation model that generates content using the acquired user-related information.

[0087] For example, the data acquisition unit 201A may acquire user-related information indicating the progress of the user's learning. This allows the determination unit 102A to determine whether the content is to be generated by a generation model according to the progress of the user's learning. For example, when the user's learning is not progressing, the first generation model 111A can generate a simple answer to a query regarding the user's learning, and when the user's learning is progressing, the second generation model 112A can generate a highly specialized answer. Similarly, the user's age, educational background, qualifications, academic performance, understanding and proficiency in various academic fields, etc. can be used as user-related information, and a generation model that generates content using this information can be determined.

[0088] The determination unit 102A may also determine a generation model that generates content depending on the input method of the query. For example, the determination unit 102A may determine that the first generation model 111A generates content for a query input by voice, and that the second generation model 112A generates content for a query input by text. This allows for voice input when emphasis is placed on response speed, and text input when emphasis is placed on response content.

[0089] (Business support system) As described above, the determination unit 102A can determine whether the first generation model 111A or the second generation model 112A is to generate content, depending on the content of the query. Therefore, the content generation system 5 can also be used as a business support system that supports the user's business.

[0090] In this case, various information related to the user's business may be used as the user-related information. For example, by using a business manual as the user-related information, the first generative model 111A or the second generative model 112A can generate an answer in accordance with the business manual.

[0091] In this case, the determination unit 102A may determine to use the first generation model 111A for queries that can be handled by the first generation model 111A, and may determine to use the second generation model 112A for queries that are difficult to handle by the first generation model 111A.

[0092] For example, a question about a business can be input as a query, and content showing an answer to the question can be presented to support the business. In this case, the determination unit 102A may determine to use the second generation model 112A for a query including vocabulary that the terminal device 2A cannot understand or a query related to an unparalleled case, and may determine to use the first generation model 111A for other queries. For example, the determination unit 102A may determine to use the second generation model 112A when the input query includes a word that is not included in the user-related information. Also, for example, the determination unit 102A may determine to use the second generation model 112A that is suitable for the technical field when the input query includes technical terms, etc. Target technical terms, etc. may be registered in advance.

[0093] It is also assumed that there is user-related information that is not preferable to be transmitted to an external device such as the generating device 3A for security reasons. In such a case, the determining unit 102A may determine to use the first generating model 111A when generating content using the user-related information, and may determine to use the second generating model 112A when generating content without using the user-related information. This allows content to be generated without transmitting the user-related information to the outside of the terminal device 2A. Note that whether or not to use the user-related information may be determined based on the contents of the query, etc. Also, the user-related information that is not preferable to be transmitted to an external device such as the generating device 3A may be used only for generating content using the first generating model 111A.

[0094] (Content Generation Example 3: Generate content corresponding to each part of the query) Furthermore, the determination unit 102A may determine to have the first generation model 111A generate content corresponding to a part of the query, and may determine to have the second generation model 112A generate content corresponding to another part of the query. This configuration provides an effect that it is possible to generate appropriate content according to the content of each part of the query, in addition to the effect of the information processing device 1. Hereinafter, an example of having the first generation model 111A and the second generation model 112A generate content corresponding to each part of the query will be described with reference to FIG. 9.

[0095] FIG. 9 is a diagram showing yet another example of content generation in the content generation system 5A. In the content generation example A3 shown in FIG. 9, content A32 and a link A33 are presented in response to a query A31 input by a user to the terminal device 2A. Specifically, the query A31 is "I want to start exercising for my health. Please show me a video of an exercise that I can easily start indoors." The content A32 is text saying, "That's great. Why not try strength training in addition to your daily stretching?" and A33 is a link to access a video that explains how to train the upper body. In this way, the content generation system 5A can also be used for health care.

[0096] In the generation example A3, the determination unit 102A inputs the first half of the query A31, "I want to start exercising for my health," and user-related information indicating the user's exercise habits, muscle mass, etc., into the first generation model 111A. As a result, a message for the user that takes into account the user-related information is generated and presented as content A32.

[0097] Furthermore, the determination unit 102A transmits the latter part of the query A31, "I would like a video showing exercises that can be easily started indoors," to the generation device 3A. As a result, the generation device 3A inputs the part into the second generative model 112A to generate a video, and transmits the generated video to the terminal device 2A. As a result, a link A33 to the video is presented on the terminal device 2A. Note that the determination unit 102A may transmit user-related information to the generation device 3A in addition to the latter part of the query A31.

[0098] When the content corresponding to each part of the query is generated in a shared manner, the determination unit 102A divides the input query into a plurality of parts. The division method is arbitrary, and for example, the determination unit 102A may divide the input query by dividing it at the position of a predetermined delimiter (such as a period, an exclamation mark, a question mark, or a space). Then, the determination unit 102A determines a generation model that executes content generation for each part obtained by division. The generation model may be determined based on the contents of each part. For example, the determination unit 102A may cause the second generation model 112A to generate content for a part that includes a predetermined keyword such as "video," and cause the first generation model 111A to generate content for a part that does not include such a keyword.

[0099] (About the waiting time until the content generated by the second generative model is presented) In the content generation system 5A, the content generated by the generation device 3A can be presented to the terminal device 2A. At this time, due to factors such as the communication environment, the data size of the generated content, and the generation method of the content, a waiting time may occur between when the user inputs a query to the terminal device 2A and when the content generated by the generation device 3A is presented to the user. Therefore, the terminal device 2A may present the content generated by the first generation model 111A to the user during the period until the content generated by the generation device 3A can be presented.

[0100] This will be described with reference to Fig. 10. Fig. 10 is a diagram showing an example of a display screen displayed by the terminal device 2A. More specifically, Fig. 10 shows a screen A41 that is displayed during the period from when the user inputs a query until the content generated by the generation device 3A becomes available for presentation, and a screen A42 on which the content generated by the generation device 3A is presented.

[0101] In the screens A41 and A42, three display areas A411 to A413 are defined. Among them, A411 is a display area for displaying a query input by the user. In the display area A411 shown in Fig. 10, a query such as "I want to go on a trip somewhere next week" is displayed.

[0102] 10, the determination unit 102A determines to cause the second generation model 112A to generate content. Therefore, the determination unit 102A transmits the above query (and user-related information) to the generation device 3A to instruct the generation of content. The determination unit 102A also inputs the above query (and user-related information) to the first generation model 111A to generate content.

[0103] The display area A412 is a display area for displaying the content generated by the first generation model 111A. Since the first generation model 111A is included in the terminal device 2A, the content can be quickly displayed in the display area A412 as shown on the screen A41. The determination unit 102A may input the query input by the user directly to the first generation model 111A, or may process the query (for example, add a note that the generation device 3A is being requested to generate the content) and then input the query to the first generation model 111A.

[0104] On the other hand, the display area A413 is a display area for displaying the content generated by the second generation model 112A. At the stage when the screen A41 is displayed, the content instructed to be generated by the generation device 3A is not in a state where it can be presented, so a character string indicating that an answer is waiting is displayed in the display area A413. Note that, when it is not necessary for the user to recognize that an answer is waiting, such a display may be omitted.

[0105] While displaying the screen A41, the generating device 3A causes the second generation model 112A to generate content, and the determining unit 102A acquires the generated content. Then, the presenting unit 103A causes the acquired content to be displayed in the display area A413. This switches the screen A41 to the screen A42. The content generated by the second generation model 112A is displayed in the display area A413 of the screen A42. Note that the presenting unit 103A may end the presentation of the content generated by the first generation model 111A when starting the presentation of the content generated by the second generation model 112A. In this case, it is not necessary to separate the display areas A412 and A413.

[0106] As described above, when the determination unit 102A determines that both the first generation model 111A and the second generation model 112A execute content generation, the presentation unit 103A may present the content generated by the first generation model 111A during the period until the content generated by the second generation model 112A becomes available for presentation. In this way, in addition to the effects of the information processing device 1, an effect is obtained that the user is less likely to feel stressed during the waiting time until the content generated by the second generation model 112A is presented.

[0107] (Processing flow) The flow of the process executed by the terminal device 2A will be described with reference to Fig. 11. Fig. 11 is a flow diagram showing the flow of the process executed by the terminal device 2A. The flow of Fig. 11 includes each step of the control method according to this exemplary embodiment.

[0108] In S11 (reception process), the reception unit 101A receives an input of a query. For example, the reception unit 101A may receive, as a query, text input by the user of the terminal device 2A via the input unit 23A.

[0109] In S12, the data acquisition unit 201A acquires user-related information. The user-related information may be input by the user, or may be stored in advance in the storage unit 21A, etc. Furthermore, the data acquisition unit 201A may acquire, as the user-related information, information related to the input query from among various pieces of information related to the user that are stored in advance in the storage unit 21A, etc.

[0110] In S13 (determination process), the determination unit 102A determines a generation model that executes the generation of the content according to the query received in S11. More specifically, in S13, the determination unit 102A determines whether or not to use the second generation model 112A for the generation of the content. For example, the determination unit 102A may determine whether or not to use the second generation model 112A according to the communication status with the generation device 3A, the query content, or the user-related information acquired in S12. If it is determined that the second generation model 112A is to generate the content in S13 as YES, the process proceeds to S16, and if it is determined that the first generation model 111A is to generate the content in S13 as NO, the process proceeds to S14.

[0111] In S14, the determination unit 102A inputs the query input in S11 to the first generation model 111A to generate content. At this time, the determination unit 102A may input the user-related information acquired in S12 to the first generation model 111A together with the query input in S11. Next, in S15 (presentation process), the presentation unit 103A presents the content generated in S14 to the user. This ends the illustrated process.

[0112] On the other hand, in S16, the request destination determination unit 202A determines the generation device 3A that will generate the content. As described above, various methods can be applied to determine the generation device 3A that will generate the content. Note that, when there is only one generation device 3A that will generate the content, the process of S16 is omitted.

[0113] In S17, the determination unit 102A transmits the query input in S11 to the generation device 3A determined in S16, and instructs the device to generate content. At this time, the determination unit 102A may transmit the user-related information acquired in S12 to the generation device 3A together with the query input in S11. In the generation device 3A that receives the query (and the user-related information), the content generation unit 301A inputs the query (and the user-related information) to the second generation model 112A to generate content. Then, when the generation of the content is completed, the content transmission unit 302A transmits the generated content to the terminal device 2A.

[0114] In S18, the determination unit 102A determines whether or not to use the first generation model 111A. The determination condition for S18 may be determined in advance. For example, the determination unit 102A may determine to use the first generation model 111A when the generation device 3A determined in S16 is a predetermined generation device 3A that takes a relatively long time to generate content. If NO is determined in S18, the process proceeds to S21, and if YES is determined in S18, the process proceeds to S19.

[0115] In S19, the determination unit 102A inputs the query input in S11 to the first generation model 111A to generate content, as in S14. Then, in S20, the content generated in S19 is presented to the user. This allows the content generated by the first generation model 111A to be presented during the period until the content generated by the second generation model 112A is presented. Note that, when it is not necessary to present the content generated by the first generation model 111A during that period, the processes of S18 to S20 may be omitted.

[0116] In S21, the determination unit 102A acquires the content (the content generated by the second generation model 112A) from the generation device 3A that generated the content in response to the processing in S17. As described above, this content is transmitted by the content transmission unit 302A of the generation device 3A. Then, in S22 (presentation processing), the presentation unit 103A presents the content acquired in S21 to the user. This ends the illustrated processing.

[0117] In addition, in S13, when it is determined not to use the second generation model 112A, that is, to use the first generation model 111A, based on the communication status with the generation device 3A, the determination unit 102A performs the process of S13 again after the content generated by the first generation model 111A is presented. Then, if the communication status with the generation device 3A has improved, it is determined to cause the second generation model 112A to generate the content.

[0118] 9, when the contents corresponding to each part of one query are generated by different generation models, the determining unit 102A may divide the input query into a plurality of parts and then perform the process from S13 onward for each part. In this case, the processes from S18 to S20 may be omitted.

[0119] [Modifications] The execution entity of each process described in each of the above exemplary embodiments is arbitrary and is not limited to the above examples. In other words, the functions of the information processing device 1, the first device 2, the second device 3, the terminal device 2A, and the generating device 3A can be realized by a plurality of devices (which can also be called processors) that can communicate with each other. For example, each process described in the flow charts of Figs. 2 and 11 can be shared and executed by a plurality of processors. In other words, the execution entity of each process in each of the above exemplary embodiments may be one processor or multiple processors.

[0120] [Software implementation example] Some or all of the functions of the information processing device 1, the first device 2, the second device 3, the terminal device 2A, and the generating device 3A (hereinafter also referred to as "each of the above devices") may be realized by hardware such as an integrated circuit (IC chip), or by software.

[0121] In the latter case, each of the above devices is realized, for example, by a computer that executes instructions of a program, which is software that realizes each function. An example of such a computer (hereinafter, referred to as computer C) is shown in Fig. 12. Fig. 12 is a block diagram showing the hardware configuration of computer C that functions as each of the above devices.

[0122] The computer C includes at least one processor C1 and at least one memory C2. The memory C2 stores a program (control program) P for causing the computer C to operate as each of the above-mentioned devices. In the computer C, the processor C1 reads and executes the program P from the memory C2, thereby realizing the functions of each of the above-mentioned devices.

[0123] The processor C1 may be, for example, a central processing unit (CPU), a graphic processing unit (GPU), a digital signal processor (DSP), a micro processing unit (MPU), a floating point number processing unit (FPU), a physics processing unit (PPU), a tensor processing unit (TPU), a quantum processor, a microcontroller, or a combination thereof. The memory C2 may be, for example, a flash memory, a hard disk drive (HDD), a solid state drive (SSD), or a combination thereof.

[0124] The computer C may further include a RAM (Random Access Memory) for expanding the program P during execution and for temporarily storing various data. The computer C may further include a communication interface for transmitting and receiving data to and from other devices. The computer C may further include an input / output interface for connecting input / output devices such as a keyboard, a mouse, a display, and a printer.

[0125] Furthermore, the program P can be recorded on a non-transitory tangible recording medium M that can be read by the computer C. Such a recording medium M can be, for example, a tape, a disk, a card, a semiconductor memory, or a programmable logic circuit. The computer C can obtain the program P via such a recording medium M. Furthermore, the program P can be transmitted via a transmission medium. Such a transmission medium can be, for example, a communication network or broadcast waves. The computer C can also obtain the program P via such a transmission medium.

[0126] [Additional Notes] This disclosure includes the techniques described in the following appendices. However, the present invention is not limited to the techniques described in the following appendices, and various modifications are possible within the scope of the claims.

[0127] (Appendix A1) An information processing device comprising: a receiving means for receiving a query input; a determining means for determining a generating model to be used to generate content corresponding to the query from among a first generating model provided in the information processing device itself and a second generating model provided in another device; and a presenting means for presenting the content generated in accordance with the determination of the determining means to a user.

[0128] (Appendix A2) The information processing device according to claim 1, wherein the determining means determines a generation model for executing generation of the content according to a communication state with the other device.

[0129] (Appendix A3) The information processing device described in Appendix A2, wherein the decision means decides to have the second generation model also generate content when communication conditions improve after the first generation model generates the content.

[0130] (Appendix A4) An information processing device as described in any of Appendices A1 to A3, comprising a data acquisition means for acquiring user-related information related to the user, and when the decision means decides to have the second generative model generate content, the decision means transmits the user-related information and the query to the other device to generate content according to the user-related information.

[0131] (Appendix A5) The information processing device according to claim 1, wherein the determining means determines a generation model for executing generation of the content in accordance with a content of the query.

[0132] (Appendix A6) An information processing device described in any of Appendices A1 to A5, wherein the presentation means, when the decision means decides to have both the first generative model and the second generative model generate content, presents the content generated by the first generative model during a period until the content generated by the second generative model becomes available for presentation.

[0133] (Appendix A7) The information processing device described in Appendix A1, wherein the determination means determines to have the first generative model generate content corresponding to a portion of the query, and determines to have the second generative model generate content corresponding to another portion of the query.

[0134] (Appendix B1) A control method including: a reception process in which at least one processor receives input of a query; a determination process in which a generation model among a first generation model and a second generation model is determined to be used to generate content in response to the query; and a presentation process in which the content generated in accordance with the determination in the determination process is presented to a user.

[0135] (Appendix B2) A control method described in Appendix B1, in which, in the determination process, the at least one processor determines a generative model to be used to generate the content depending on communication conditions with another device equipped with the second generative model.

[0136] (Appendix B3) The control method described in Appendix B2, wherein the at least one processor decides to have the second generative model also generate content if communication conditions improve after the first generative model generates the content.

[0137] (Appendix B4) A control method described in any of Appendices B1 to B3, wherein the at least one processor includes a data acquisition process for acquiring user-related information related to the user, and when the at least one processor decides in the decision process to have the second generative model generate content, the at least one processor transmits the user-related information and the query to the other device to generate content according to the user-related information.

[0138] (Appendix B5) The control method according to Appendix B1, wherein in the determination process, the at least one processor determines a generation model to cause generation of the content in accordance with the content of the query.

[0139] (Appendix B6) A control method described in any of Appendices B1 to B5, in which, when it is decided to have both the first generative model and the second generative model generate content, the at least one processor executes a process of presenting the content generated by the first generative model during a period until the content generated by the second generative model becomes available for presentation.

[0140] (Appendix B7) The control method described in Appendix B1, wherein, in the decision process, the at least one processor decides to have the first generative model generate content corresponding to a portion of the query, and decides to have the second generative model generate content corresponding to another portion of the query.

[0141] (Appendix C1) A control program that causes a computer to function as a receiving means for receiving input of a query, a determination means for determining, from among a first generation model and a second generation model, which generation model will be used to generate content in response to the query, and a presentation means for presenting to a user the content generated in accordance with the determination of the determination means.

[0142] (Appendix C2) The control program according to claim 1, wherein the determining means determines a generation model for executing generation of the content according to a communication status with the other device.

[0143] (Appendix C3) The control program described in Appendix C2, wherein the decision means decides to have the second generation model also generate content when communication conditions improve after the first generation model generates the content.

[0144] (Appendix C4) A control program described in any of Appendices C1 to C3, which causes the computer to function as a data acquisition means for acquiring user-related information related to the user, and when the decision means decides to have the second generative model generate content, sends the user-related information and the query to the other device, causing the other device to generate content according to the user-related information.

[0145] (Appendix C5) The control program according to claim 1, wherein the determining means determines a generation model for executing generation of the content according to a content of the query.

[0146] (Appendix C6) A control program described in any of Appendices C1 to C5, wherein the presentation means, when the decision means decides to have both the first generative model and the second generative model generate content, presents the content generated by the first generative model during a period until the content generated by the second generative model becomes available for presentation.

[0147] (Appendix C7) The control program described in Appendix C1, wherein the determination means determines to have the first generative model generate content corresponding to a part of the query, and determines to have the second generative model generate content corresponding to another part of the query.

[0148] (Appendix D1) An information processing device having at least one processor, the at least one processor executing a reception process for receiving a query input, a determination process for determining a generation model from among a first generation model provided in the information processing device itself and a second generation model provided in another device, which will execute generation of content corresponding to the query, and a presentation process for presenting to a user the content generated in accordance with the determination in the determination process.

[0149] (Appendix D2) The information processing device according to Addition D1, wherein in the determination process, the at least one processor determines a generation model for executing generation of the content depending on a communication status with the other device.

[0150] (Appendix D3) The information processing device described in Appendix D2, wherein the at least one processor decides to have the second generative model also generate content when communication conditions improve after the first generative model generates the content.

[0151] (Appendix D4) An information processing device described in any of Appendices D1 to D3, which executes a data acquisition process to acquire user-related information related to the user, and in the decision process, when the at least one processor decides to have the second generative model generate content, transmits the user-related information and the query to the other device to generate content according to the user-related information.

[0152] (Appendix D5) The information processing device according to claim 1, wherein in the determination process, the at least one processor determines a generation model that causes generation of the content to be executed in accordance with content of the query.

[0153] (Appendix D6) An information processing device described in any of Appendices D1 to D5, wherein, when it is decided to have both the first generative model and the second generative model generate content, the at least one processor executes a process of presenting the content generated by the first generative model during a period until the content generated by the second generative model becomes available for presentation.

[0154] (Appendix D7) The information processing device described in Appendix D1, wherein, in the decision process, the at least one processor decides to have the first generative model generate content corresponding to a part of the query, and decides to have the second generative model generate content corresponding to another part of the query.

[0155] The information processing device may further include a memory, and the memory may store a control program for causing the at least one processor to execute each of the processes.

[0156] (Appendix E1) A non-temporary recording medium having recorded thereon a control program that causes a computer to function as an information processing device, the control program causing the computer to execute a reception process for receiving a query input, a determination process for determining which of a first generation model provided by the information processing device and a second generation model provided by another device will be used to generate content in response to the query, and a presentation process for presenting to a user the content generated in accordance with the determination of the determination process. [Explanation of symbols]

[0157] 1. Information processing device 101 Reception section (reception means) 102 Determination unit (determination means) 103 Presentation unit (presentation means) 111 First Generative Model 112 The second generative model 2. First Device 3. Second Device 5 Content Generation System 2A Terminal device (information processing device, first device) 101A Reception section (reception means) 102A Determination unit (determination means) 103A Presentation unit (presentation means) 201A data acquisition unit (related information acquisition means) 111A First Generative Model 112A Second Generative Model 3A Generator (Second Device) 5A Content Generation System

Claims

1. A receiving means for receiving a query input; a determination means for determining a generation model to be used to generate content in response to the query from among a first generation model included in the own device and a second generation model included in the other device; a presentation unit that presents the content generated in accordance with the decision of the decision unit to a user.

2. The information processing apparatus according to claim 1 , wherein the determining means determines a generation model for executing generation of the content depending on a communication state with the other apparatus.

3. The information processing device according to claim 2 , wherein the determination means determines to cause the second generation model to also generate the content when communication conditions improve after the first generation model has generated the content.

4. a related information acquiring means for acquiring user related information related to the user, 4. The information processing device according to claim 1, wherein when the decision means decides to have the second generative model generate content, the decision means transmits the user-related information and the query to the other device to generate content according to the user-related information.

5. The information processing apparatus according to claim 1 , wherein the determining means determines a generation model for executing generation of the content in accordance with a content of the query.

6. An information processing device as described in any one of claims 1 to 3, wherein when the decision means decides to have both the first generation model and the second generation model generate content, the presentation means presents the content generated by the first generation model during a period until the content generated by the second generation model becomes available for presentation.

7. The information processing device according to claim 1 , wherein the determination means determines to have the first generative model generate content corresponding to a portion of the query, and determines to have the second generative model generate content corresponding to another portion of the query.

8. a first device comprising a first generative model; a second device comprising a second generative model; A content generation system in which the first device determines, from the first generation model and the second generation model, a generation model that will generate content in response to a query input to the first device, and presents the content generated in accordance with the determination to a user.

9. One or more processors A reception process for receiving a query input; a determination process of determining, from among a first generation model and a second generation model, a generation model that will execute generation of content in response to the query; a presentation process for presenting to a user the content generated in accordance with the determination in the determination process; and a control method for executing the present process.

10. Computer, A receiving means for receiving a query input; a determining means for determining which of a first and a second generative model is to be used to generate content in response to the query; and a control program that functions as a presentation means for presenting the content generated in accordance with the decision of the decision means to a user;

Images