Information processing device

Abstract

The present invention provides an information processing device for acquiring information that is expected by a user from a machine learning device, by inputting instruction information to the machine learning device, the instruction information being obtained by converting utterance content of the user into a format that is readily comprehended by the machine learning device.　An information processing device 100 is an information processing device that mutually communicates with a machine learning device 300 having a machine learning function, inputs instruction information to the machine learning device 300, and acquires response information from the machine learning device 300 in response to the instruction information. The information processing device includes a speech recognition unit 111 that performs recognition processing of uttered speech of a user, an utterance comprehension unit 112 that comprehends meaning of the utterance of the user on the basis of results of the recognition by the speech recognition unit, an instruction information generation unit 113 that generates the instruction information to be input to the machine learning device 300 on the basis of the meaning of the utterance of the user, a response information acquisition unit 114 that acquires the response information from the machine learning device, and a control unit 116 that performs display control of information regarding the instruction information, and information regarding the response information.

Description

Information processing device

[0001] The present invention relates to an information processing device.

[0002] In recent years, smart speakers and the like that understand the meaning of a user's speech and provide the user with information related to the user's instructions have become widespread.

[0003] For example, one such technology proposed is one in which the speech of a user riding in a vehicle is transmitted to a machine learning device connected to a network, and information based on the speech is provided to the user (see, for example, Patent Document 1).

[0004] Japanese Patent Application Laid-Open No. 2021-157298

[0005] In general, in order to obtain a user's expected answer from a machine learning device, it is necessary to input command information, which is obtained by converting the user's utterance content into a format that is easy for the machine learning device to understand, to the machine learning device. However, with the technology described in Patent Literature 1, command information is input to the machine learning device without converting it into a format that is easy for the machine learning device to understand, which poses a problem in that the information that the user expects may not be obtained.

[0006] Therefore, the present invention has been made in consideration of the above-mentioned problems, and aims to provide an information processing device that obtains the information expected by a user from a machine learning device by inputting command information into the machine learning device that converts the content of a user's utterance into a format that is easy for the machine learning device to understand.

[0007] Form 1: One or more embodiments of the present invention propose an information processing device that communicates with a machine learning device having a machine learning function, inputs command information to the machine learning device, and acquires response information to the command information from the machine learning device, the information processing device comprising: a speech recognition unit that performs recognition processing of a user's spoken voice; an utterance understanding unit that understands the meaning of the user's utterance based on the recognition result of the speech recognition unit; a command information generation unit that generates command information to be input to the machine learning device based on the meaning of the utterance; a response information acquisition unit that acquires the response information from the machine learning device; and a control unit that controls the display of information related to the command information and information related to the response information, wherein the command information generation unit generates the command information by converting the meaning of the user's utterance into a format that is easily understandable by the machine learning device.

[0008] Form 7: One or more embodiments of the present invention propose an information processing device that communicates with a machine learning device having a machine learning function, inputs command information to the machine learning device, and acquires response information to the command information from the machine learning device, the information processing device comprising one or more processors and one or more memories communicably connected to the one or more processors, wherein the one or more processors include: a speech recognition unit that performs recognition processing of a user's spoken voice; an utterance understanding unit that understands the meaning of the user's utterance based on the recognition result of the speech recognition unit; a command information generation unit that generates command information to be input to the machine learning device based on the meaning of the utterance; a response information acquisition unit that acquires the response information from the machine learning device; and a control unit that controls the display of information related to the command information and information related to the response information, wherein the command information generation unit generates the command information by converting the meaning of the user's utterance into a format that is easily understandable by the machine learning device.

[0009] According to one or more embodiments of the present invention, by inputting command information into a machine learning device that converts the content of a user's speech into a format that is easy for the machine learning device to understand, it is possible to obtain the information that the user expects from the machine learning device.

[0010] FIG. 1 is a diagram showing a communication configuration of an information processing device according to a first embodiment of the present invention. FIG. 2 is a diagram showing a configuration of an information processing device according to a first embodiment of the present invention. FIG. 3 is a diagram showing a configuration of a processor of an information processing device according to a first embodiment of the present invention. FIG. 4 is a diagram showing a processing flow of an information processing device according to a first embodiment of the present invention. FIG. 5 is a diagram showing a communication configuration of an information processing device according to a second embodiment of the present invention. FIG. 6 is a diagram showing a configuration of an information processing device according to a second embodiment of the present invention. FIG. 7 is a diagram showing a configuration of a processor of an information processing device according to a second embodiment of the present invention. FIG. 8 is a diagram showing a processing flow of an information processing device according to a second embodiment of the present invention.

[0011] <Embodiment> An information processing apparatus according to this embodiment will be described with reference to Figs. 1 to 8 .

[0012] First Embodiment An information processing apparatus 100 according to this embodiment will be described with reference to FIGS. 1 to 4. FIG.

[0013] <Communication configuration of information processing device 100> As shown in Fig. 1, the information processing device 100 according to this embodiment is connected to at least one or more machine learning devices 300_1 to 300_N via a network 200 so that they can communicate with each other. The network 200 may be, for example, the Internet. Hereinafter, the machine learning devices 300_1 to 300_N connected to the network 200 will be collectively referred to as "machine learning devices 300." The information processing device 100 transmits command information to the machine learning devices 300 via the network 200.

[0014] The machine learning device 300 is configured to include at least a trained machine learning model and a machine learning function. In this embodiment, the machine learning device 300 outputs response information in response to command information received from the information processing device 100. The machine learning device 300 transmits the response information to the information processing device 100 via the network 200. The command information and response information will be described later.

[0015] <Configuration of Information Processing Apparatus 100> As shown in FIG. 2, the information processing apparatus 100 according to this embodiment includes a processor 110, a memory 120, a communication unit 130, and a display unit 140.

[0016] The processor 110 transmits command information to the machine learning device 300 via the communication unit 130, which will be described later. The processor 110 receives response information from the machine learning device 300 via the communication unit 130, which will be described later. The configuration of the processor 110 will be described later.

[0017] The memory 120 includes a read-only memory (ROM) and a random access memory (RAM), both of which are not shown. The memory 120 stores, for example, a control program and various data received from the processor 110. In this embodiment, the memory 120 pre-stores, for example, information about the machine learning devices 300. Examples of the information about the machine learning devices 300 include information about the languages ​​supported by each machine learning device 300 and information about the areas of expertise of each machine learning device 300. In this embodiment, the memory 120 pre-stores, for example, an acoustic model and a pattern matching dictionary for speech recognition. In this embodiment, the memory 120 pre-stores, for example, a list, created by the user, of machine learning devices 300 from which reliable response information can be obtained.

[0018] The communication unit 130 is, for example, a wireless communication module, and communicates with the machine learning device 300 via the network 200. In this embodiment, the communication unit 130 transmits, for example, command information received from the processor 110 to the machine learning device 300 connected to the network 200. In this embodiment, the communication unit 130 transmits, for example, response information received from the machine learning device 300 via the network 200 to the processor 110.

[0019] The display unit 140 is, for example, a display panel such as a liquid crystal display, and displays information received from the processor 110. In this embodiment, the display unit 140 displays, for example, information related to command information and information related to response information, which will be described later. For example, when the information processing device 100 is installed in a vehicle, the display unit 140 may be a display panel provided in a navigation device or a meter display device. The information displayed by the display unit 140 will be described later.

[0020] 3, the processor 110 includes a speech recognition unit 111, an utterance understanding unit 112, a command information generation unit 113, a response information acquisition unit 114, a conversion information generation unit 115, and a control unit 116. The individual units of the processor 110 and the memory 120 transmit and receive various types of information via a bus line BL.

[0021] The voice recognition unit 111 performs a recognition process for the user's speech. The voice recognition unit 111 converts the user's speech, which is picked up by a microphone (not shown), into text data. The voice recognition unit 111 converts the user's speech into text data using, for example, an acoustic model such as a hidden Markov model and a pattern matching dictionary. The voice recognition unit 111 transmits the text data to the control unit 116 (described later) via the bus line BL.

[0022] The utterance understanding unit 112 understands the meaning of the user's utterance based on the recognition result of the speech recognition unit 111. The utterance understanding unit 112 understands the meaning of the utterance by, for example, performing natural language processing (NLP) such as morphological analysis on text data received from the control unit 116 (described later). The utterance understanding unit 112 transmits information regarding the meaning of the user's utterance to the control unit 116 (described later) via the bus line BL.

[0023] The command information generation unit 113 generates command information to be input to the machine learning device 300 based on the meaning of the user's utterance. The command information generation unit 113 generates command information for acquiring information that the user wants to acquire from the machine learning device 300. In this embodiment, the command information generation unit 113 extracts what information the user wants to acquire from the machine learning device 300 based on information about the meaning of the user's utterance received from the control unit 116 (described later), and generates command information. The command information generation unit 113 generates command information converted into a format that is easy for the machine learning device 300 to understand. The command information generation unit 113 generates command information by converting the meaning of the user's utterance into a sentence that conforms to a correct sentence pattern. The command information generation unit 113 converts the meaning of the user's utterance into, for example, a sentence that correctly combines a subject, a predicate, an object, and a complement. The command information generation unit 113 generates command information that takes into account the configuration of a trained machine learning model possessed by the machine learning device 300, for example. For example, when the trained machine learning model of the machine learning device 300 is configured in English, the command information generation unit 113 translates the meaning of the user's utterance into English to generate command information. For example, when the user utters multiple sentences in succession, the command information generation unit 113 performs conversion to summarize the content of the multiple sentences, and generates command information. The command information generation unit 113 only needs to generate command information converted into a format that is easy for the machine learning device 300 to understand, and therefore the method for generating command information is not limited to the above-described generation method. The command information generation unit 113 transmits the generated command information to the control unit 116 (described later) via the bus line BL.

[0024] The response information acquisition unit 114 acquires response information from the machine learning device 300 via the communication unit 130. In this embodiment, the response information acquisition unit 114 acquires response information from a reliable machine learning device 300. The response information acquisition unit 114 acquires a list of machine learning devices 300 from which reliable response information can be acquired from the memory 120 via the bus line BL. The response information acquisition unit 114 acquires response information received from the machine learning devices 300 included in the acquired list. The response information acquisition unit 114 transmits the acquired response information to the control unit 116, which will be described later, via the bus line BL.

[0025] The conversion information generation unit 115 generates conversion information by converting the response information into a format that is easy for the user to understand. The conversion information generation unit 115 generates the conversion information by converting the response information into a sentence that conforms to a correct sentence pattern. For example, the conversion information generation unit 115 converts response information received from the control unit 116 (described later) into a sentence that correctly combines a subject, a predicate, an object, and a complement. For example, when the conversion information generation unit 115 receives multiple pieces of response information in response to the same command information from the control unit 116 (described later), the conversion information generation unit 115 performs conversion to summarize the acquired response information. For example, the conversion information generation unit 115 performs natural language processing (NLP) on the response information received from the control unit 116 (described later) to generate conversion information. For example, when the response information received from the control unit 116 (described later) has a number of characters greater than a predetermined number, the conversion information generation unit 115 performs conversion to summarize the response information. For example, the conversion information generation unit 115 converts the response information received from the control unit 116 (described later) into a language spoken by the user. For example, when the response information received from the control unit 116 (described later) is data, the conversion information generation unit 115 performs conversion to visualize the response information so that it can be understood by the user. For example, when the conversion information generation unit 115 receives data related to daily weather forecasts for the next week from the machine learning device 300, the conversion information generation unit 115 converts the daily weather, temperature, etc. into a tabular format. The conversion information generation unit 115 only needs to convert the data into a format that is easy for the user to understand, so the method for generating the conversion information is not limited to the above-described generation method. The conversion information generation unit 115 transmits the conversion information to the control unit 116 (described later) via the bus line BL.

[0026] The control unit 116 controls the entire information processing device 100 in accordance with a control program stored in the memory 120. In this embodiment, the control unit 116 transmits an instruction to the speech recognition unit 111 to convert the user's utterance into text data, for example, via the bus line BL. In this embodiment, the control unit 116 transmits the text data received from the speech recognition unit 111 to the utterance understanding unit 112, for example, via the bus line BL, and causes the utterance understanding unit 112 to execute a process of understanding the meaning of the user's utterance. In this embodiment, the control unit 116 transmits information regarding the meaning of the user's utterance received from the utterance understanding unit 112, for example, via the bus line BL, to the command information generation unit 113, and causes the utterance understanding unit 112 to execute a process of generating command information. In this embodiment, the control unit 116 transmits command information to the machine learning device 300 that can acquire information expected by the user. In this embodiment, the control unit 116 transmits command information to the machine learning device 300 that is highly relevant to the content of the command information and the specialty field of the machine learning device 300 acquired from the memory 120, for example. In the present embodiment, the control unit 116 transmits command information to the machine learning device 300 via the communication unit 130. In the present embodiment, the control unit 116 transmits response information received from the response information acquisition unit 114 to the conversion information generation unit 115, causing the conversion information generation unit 115 to execute processing for generating conversion information. In the present embodiment, the control unit 116 controls the display unit 140 to display information related to the command information and information related to the response information on the display unit 140. In the present embodiment, the control unit 116 transmits command information received from the command information generation unit 113 to the display unit 140. In the present embodiment, when displaying information related to the response information, the control unit 116 causes the display unit 140 to display the conversion information received from the conversion information generation unit 115. In the present embodiment, the control unit 116 transmits the conversion information received from the conversion information generation unit 115 to the display unit 140.

[0027] <Processing of Information Processing Device 100> The flow of processing of the information processing device 100 will be described with reference to FIG.

[0028] As shown in FIG. 4, the control unit 116 transmits an instruction to convert the user's utterance into text data to the voice recognition unit 111 (step S110).

[0029] The control unit 116 transmits the text data to the utterance understanding unit 112, which then understands the meaning of the user's utterance (step S120).

[0030] The control unit 116 transmits information relating to the meaning of the user's utterance to the command information generation unit 113, causing the command information generation unit 113 to generate command information based on the meaning of the user's utterance (step S130).

[0031] The control unit 116 transmits the command information generated in step S130 to the machine learning device 300 via the communication unit 130 (step S140).

[0032] The control unit 116 transmits the command information generated in step S130 to the display unit 140, and causes the display unit 140 to display the command information (step S150).

[0033] The control unit 116 acquires the response information from the response information acquisition unit 114 (step S160).

[0034] The control unit 116 transmits the response information acquired in step S160 to the conversion information generation unit 115, causing the conversion information generation unit 115 to generate conversion information (step S170).

[0035] The control unit 116 transmits the conversion information generated in step S170 to the display unit 140, causes the display unit 140 to display the conversion information (step S180), and ends the process.

[0036] <Actions and Effects> As described above, the information processing device 100 according to this embodiment is an information processing device that communicates with a machine learning device 300 having a machine learning function, inputs command information to the machine learning device 300, and acquires response information to the command information from the machine learning device 300. The information processing device 100 includes a speech recognition unit 111 that performs recognition processing of a user's utterance, an utterance understanding unit 112 that understands the meaning of the user's utterance based on the recognition result of the speech recognition unit, a command information generation unit 113 that generates command information to be input to the machine learning device 300 based on the meaning of the user's utterance, a response information acquisition unit 114 that acquires response information from the machine learning device 300, and a control unit 116 that controls the display of information related to the command information and the response information. The command information generation unit 113 generates command information by converting the meaning of the user's utterance into a format that is easy for the machine learning device 300 to understand. In other words, the information processing device 100 generates command information in which the meaning of the user's utterance has been converted into a format that is easy for the machine learning device 300 to understand, so that the response information expected by the user can be acquired from the machine learning device 300. This allows the information processing device 100 to acquire information expected by the user from the machine learning device 300. The information processing device 100 displays command information generated based on the meaning of the user's utterance on the display unit 140. This allows the user to check whether command information that matches the content of the utterance has been input to the machine learning device 300.

[0037] The command information generation unit 113 of the information processing device 100 according to this embodiment generates command information by converting the meaning of a user's utterance into a sentence that conforms to a correct sentence structure. That is, the command information generation unit 113 converts the meaning of a user's utterance into a sentence that conforms to a correct sentence structure so that response information expected by the user can be obtained from the machine learning device 300. This allows the information processing device 100 to obtain information expected by the user from the machine learning device 300 even if the user's utterance does not conform to a correct sentence structure.

[0038] The information processing device 100 according to this embodiment includes a converted information generation unit 115 that generates converted information by converting the response information acquired from the machine learning device 300 into a format that is easy for the user to understand. In other words, the information processing device 100 converts the response information acquired from the machine learning device 300 into language, sentences, etc. that are easy for the user to understand, and displays the converted information on the display unit 140. This allows the user to easily understand the content of the response information.

[0039] The conversion information generation unit 115 of the information processing device 100 according to this embodiment generates conversion information by converting response information into sentences conforming to correct sentence patterns, thereby enabling the user to easily understand the content of the response information.

[0040] The response information acquisition unit 114 of the information processing device 100 according to this embodiment acquires response information from a reliable machine learning device 300. In other words, the information processing device 100 provides the user with response information from a machine learning device 300 that is trusted by the user. This allows the user to trust the information provided by the information processing device 100.

[0041] The control unit 116 of the information processing device 100 according to this embodiment controls the transmission of command information to the machine learning device 300 that can acquire the information expected by the user. That is, the information processing device 100 transmits command information to the machine learning device 300 whose field of expertise is highly relevant to the content of the command information, and acquires response information. This allows the information processing device 100 to acquire the information expected by the user.

[0042] Second Embodiment An information processing device 100A according to this embodiment will be described with reference to Figures 5 to 8. Note that components with the same reference numerals as those in the first embodiment have similar functions, and therefore detailed descriptions thereof will be omitted.

[0043] 5, the information processing device 100A according to this embodiment is connected to at least one or more machine learning devices 300A_1 to 300A_N via a network 200 so that they can communicate with each other. Hereinafter, the machine learning devices 300A_1 to 300A_N connected to the network 200 will be collectively referred to as the "machine learning device 300A." The information processing device 100A transmits command information to the machine learning device 300A via the network 200.

[0044] The machine learning device 300A is configured to include at least a trained machine learning model and a machine learning function. In this embodiment, the machine learning device 300A outputs response information to command information received from the information processing device 100A. The machine learning device 300A is, for example, a device having a generative artificial intelligence (AI) function, and in this embodiment, outputs the response information to the command information based on information regarding the role to be assigned to the machine learning device 300A and information regarding the output format expected by the user, which are added to the command information received from the information processing device 100A. The information regarding the role to be assigned to the machine learning device 300A and the information regarding the output format expected by the user will be described later. The machine learning device 300A transmits the response information to the information processing device 100A via the network 200.

[0045] <Configuration of Information Processing Apparatus 100A> As shown in FIG. 6, the information processing apparatus 100A according to this embodiment includes a processor 110A, a memory 120, a communication unit 130, and a display unit 140.

[0046] The processor 110A transmits command information to the machine learning device 300A via the communication unit 130. The processor 110A receives response information from the machine learning device 300A via the communication unit 130. The configuration of the processor 110A will be described later.

[0047] 7, the processor 110A is configured to include a speech recognition unit 111, an utterance understanding unit 112, a command information generation unit 113A, a response information acquisition unit 114, a conversion information generation unit 115A, a control unit 116A, a role extraction unit 117, and an output format extraction unit 118. Each unit of the processor 110A and the memory 120 transmits and receives various types of information via a bus line BL.

[0048] The role extraction unit 117 extracts a role to be assigned to the machine learning device 300A from the meaning of the user's utterance. The role extraction unit 117 extracts a role to be assigned to the machine learning device 300A based on information about the meaning of the user's utterance received from the control unit 116A, which will be described later. The role to be assigned to the machine learning device 300A is a condition for acquiring information from the machine learning device 300A. For example, when the machine learning device 300A is assigned the role of CEO, it outputs response information to command information as CEO. Examples of roles to be assigned to the machine learning device 300A include politician, teacher, parent, leader, professional musician, artist, etc. The role extraction unit 117 transmits information about the role to be assigned to the machine learning device 300A to the control unit 116A, which will be described later, via the bus line BL.

[0049] The output format extraction unit 118 extracts information about the information output format expected by the user from the meaning of the user's utterance. The information about the information output format expected by the user is a condition for acquiring information from the machine learning device 300A. For example, when the machine learning device 300A receives information about the user's expected output format from the information processing device 100A, the machine learning device 300A outputs response information according to the user's expected output format. The output format extraction unit 118 extracts, for example, specific examples of information expected by the user and information about the output format from information about the meaning of the user's utterance received from the control unit 116A described below. For example, the output format extraction unit 118 extracts, from the user's utterance, utterances such as "I want specific information including numerical values." For example, the output format extraction unit 118 extracts, from the user's utterance, utterances related to specific output formats, such as "I want information displayed in a news format," "I want information displayed in a table format that allows comparisons," and "I want information summarized in 500 characters or less." In this embodiment, the output format extraction unit 118 transmits information relating to the output format of information expected by the user to a control unit 116A (described later) via the bus line BL.

[0050] The command information generation unit 113A generates command information to be input to the machine learning device 300A based on the meaning of the user's utterance. The command information generation unit 113A converts information about the meaning of the user's utterance received from the control unit 116A (described later) into a format that is easy for the machine learning device 300A to understand, thereby generating command information. The method of converting into a format that is easy for the machine learning device 300A to understand is the same as that of the command information generation unit 113, and therefore a detailed description thereof will be omitted.

[0051] The command information generation unit 113A generates command information by adding information about the role to be given to the machine learning device 300A and information about the output format of information expected by the user, which are extracted from the meaning of the user's utterance. In this embodiment, the command information generation unit 113A generates command information by converting the command information into a format that is easy for the machine learning device 300A to understand, and adding information about the role to be given to the machine learning device 300A and information about the output format of information expected by the user, which are received from the control unit 116A, which will be described later. The command information generation unit 113A transmits the generated command information to the control unit 116A via the bus line BL.

[0052] The conversion information generation unit 115A generates conversion information by converting the response information into a format that is easy for the user to understand. In this embodiment, the conversion information generation unit 115A converts the response information received from the control unit 116A (described later) according to information regarding the information output format expected by the user, extracted by the output format extraction unit 118, and generates the conversion information. For example, if the information regarding the information output format expected by the user indicates that the information should be displayed in a news format, the conversion information generation unit 115A executes a process of combining all of the response information received from the machine learning device 300A to generate the conversion information. For example, if the information regarding the information output format expected by the user indicates that the information should be displayed in a comparable table format, the conversion information generation unit 115A executes a process of summarizing the response information for each machine learning device 300A into a comparable table format, and generates the conversion information. For example, if the information regarding the information output format expected by the user indicates that the information should be summarized in 500 characters or less, the conversion information generation unit 115A aggregates the response information for each machine learning device 300A and executes a process of summarizing the response information for each machine learning device 300A into 500 characters or less to generate the conversion information. The conversion information generating unit 115A transmits the conversion information to a control unit 116A, which will be described later, via the bus line BL.

[0053] The control unit 116A controls the entire information processing device 100A in accordance with a control program stored in the memory 120. In this embodiment, the control unit 116A transmits an instruction to convert a user's utterance into text data to the speech recognition unit 111, for example, via the bus line BL. In this embodiment, the control unit 116A transmits the text data received from the speech recognition unit 111 to the utterance understanding unit 112, for example, via the bus line BL, and causes the utterance understanding unit 112 to execute a process of understanding the meaning of the user's utterance. In this embodiment, the control unit 116A transmits information regarding the meaning of the user's utterance received from the utterance understanding unit 112 and the extraction results of the role extraction unit 117 and the output format extraction unit 118, for example, via the bus line BL, to the command information generation unit 113A, and causes the utterance understanding unit 112 to execute a process of generating command information. In this embodiment, the control unit 116A transmits the command information received from the command information generation unit 113A to the machine learning device 300A via the communication unit 130. In this embodiment, the control unit 116A controls the transmission of command information to the machine learning device 300A that can acquire information expected by the user. In this embodiment, the control unit 116A transmits the command information to, for example, the machine learning device 300A whose field of expertise, acquired from the memory 120, is highly relevant to the content of the command information. In this embodiment, the control unit 116A transmits the response information received from the response information acquisition unit 114 and the extraction result of the output format extraction unit 118 to the conversion information generation unit 115A, and causes the conversion information generation unit 115A to execute processing to generate conversion information. In this embodiment, the control unit 116A controls the display unit 140 to display information related to the command information and information related to the response information on the display unit 140. In this embodiment, the control unit 116A transmits the command information received from the command information generation unit 113 to the display unit 140. In this embodiment, when displaying information related to the response information, the control unit 116A causes the display unit 140 to display the conversion information received from the conversion information generation unit 115A. In this embodiment, the control unit 116A transmits the conversion information received from the conversion information generation unit 115A to the display unit 140.

[0054] <Processing of Information Processing Apparatus 100A> The processing flow of the information processing apparatus 100A will be described with reference to FIG.

[0055] As shown in FIG. 8, the control unit 116A transmits an instruction to convert the user's utterance into text data to the voice recognition unit 111 (step S210).

[0056] The control unit 116A transmits the text data to the utterance understanding unit 112, causing the utterance understanding unit 112 to understand the meaning of the user's utterance (step S220).

[0057] Control unit 116A transmits information about the meaning of the user's utterance to role extraction unit 117, and causes role extraction unit 117 to extract a role to be assigned to machine learning device 300A (step S230).

[0058] The control unit 116A transmits information relating to the meaning of the user's utterance to the output format extraction unit 118, causing the output format extraction unit 118 to extract information relating to the output format expected by the user (step S240).

[0059] Control unit 116A transmits information relating to the meaning of the user's utterance and the extraction results of steps S230 and S240 to command information generation unit 113A, causing it to generate command information (step S250).

[0060] Control unit 116A transmits the command information generated in step S250 to machine learning device 300A via communication unit 130 (step S260).

[0061] Control unit 116A transmits the command information generated in step S250 to display unit 140, and causes display unit 140 to display the command information (step S270).

[0062] Control unit 116A acquires response information from response information acquisition unit 114 (step S280).

[0063] Control unit 116A transmits the response information acquired in step S280 and the information on the output format expected by the user extracted in step S240 to conversion information generation unit 115A, causing conversion information to be generated (step S290).

[0064] Control unit 116A transmits the conversion information generated in step S290 to display unit 140, causes display unit 140 to display the conversion information (step S300), and ends the process.

[0065] <Actions and Effects> As described above, the command information generation unit 113A of the information processing device 100A according to this embodiment generates command information to which information on the role to be assigned to the machine learning device 300A and information on the output format of information expected by the user, which are extracted from the meaning of the user's utterance, is added. The information processing device 100A analyzes the content of the user's utterance and extracts the specific information the user wants to obtain from the machine learning device 300A. The information processing device 100A extracts, from the content of the user's utterance, the role to be assigned to the machine learning device 300A and the output format of information expected by the user. In other words, the information processing device 100A transmits command information to which information on the role to be assigned to the machine learning device 300A and information on the output format of information expected by the user are added, and acquires response information from the machine learning device 300A. This allows the information processing device 100A to acquire the information expected by the user from the machine learning device 300A. The information processing device 100A automatically extracts the role to be assigned to the machine learning device 300A and the information output format expected by the user from the content of the user's utterance, allowing the user to obtain the expected information simply by speaking. The information processing device 100A displays command information to which information regarding the role to be assigned to the machine learning device 300A and information regarding the information output format expected by the user are added on the display unit 140. This allows the user to confirm whether command information in line with the content of the utterance has been input to the machine learning device 300A.

[0066] The converted information generation unit 115A of the information processing device 100A according to this embodiment generates converted information by converting response information based on the output format of information expected by the user. That is, the information processing device 100A displays the converted information converted based on the output format of information expected by the user on the display unit 140. This allows the user to obtain information in the expected output format.

[0067] <Variation 1> The information processing device 100 may be configured by executing the processes of the speech recognition unit 111, the utterance understanding unit 112, the command information generation unit 113, the response information acquisition unit 114, the conversion information generation unit 115, and the control unit 116 of the information processing device 100 according to the above-described embodiment in a server device connected to the network 200. The information processing device 100 collects a user's utterance using a microphone or the like and transmits the collected utterance to the server device. The server device transmits the generated command information and conversion information to the information processing device 100. The information processing device 100 displays the command information and conversion information received from the server device on the display unit 140. This reduces the processing load on the information processing device 100.

[0068] <Variation 2> The information processing device 100 according to the present embodiment described above may be configured to customize the conversion method of the conversion information generation unit 115 according to user preferences. Examples of conversion methods include conversion using simple expressions, conversion not using alphabetic abbreviations, and conversion that includes attaching the conclusion of the response information to the beginning. The user registers their preferred conversion method in advance in the information processing device 100. The conversion information generation unit 115 converts the response information using the conversion method registered by the user. This allows the user to obtain response information converted using their preferred conversion method, making it easier to understand the response information.

[0069] <Modification 3> The information processing devices 100 and 100A according to the present embodiment described above display the conversion information on the display unit 140, but may also transmit the conversion information to a mobile terminal or the like carried by the user. This allows the user to check the conversion information regardless of location or time.

[0070] <Modification 4> The information processing devices 100 and 100A according to the above-described embodiment display the conversion information on the display unit 140, but may also provide the conversion information to the user by voice. In this way, when the information processing devices 100 and 100A are provided in a vehicle, the information is provided by voice, so that the user can obtain the information without gazing at the display unit 140, even while driving.

[0071] <Variation 5> The information processing devices 100 and 100A according to the above-described embodiment may, for example, confirm whether the command information extracted from the user's utterance is correct. That is, before transmitting the command information to the machine learning devices 300 and 300A, the information processing devices 100 and 100A display the command information on the display unit 140, for example, and then confirm whether the command information matches the user's expectations based on the user's utterance. If the information processing devices 100 and 100A confirm that the command information matches the user's expectations, they transmit the command information to the machine learning devices 300 and 300A. This allows the information processing devices 100 and 100A to reliably provide the information expected by the user.

[0072] <Variation 6> The information processing device 100 according to the above-described embodiment may provide the user with response information acquired from the machine learning device 300, for example, in accordance with the content of a chat with the user. That is, the information processing device 100 outputs a voice message such as, for example, "A total of 50 pieces of information have been acquired from five machine learning devices 300. How many would you like to display?", and then provides the user with the information acquired from the machine learning device 300 in accordance with the content of the user's subsequent utterance. This allows the user to acquire information in an output format that they expect.

[0073] The information processing device 100 of the present invention can be realized by recording the processes of the speech recognition unit 111, the utterance understanding unit 112, the command information generation unit 113, the conversion information generation unit 115, and the control unit 116 of the information processing device 100 according to the present embodiment described above on a recording medium readable by a computer system, and loading and executing the program recorded on this recording medium into memory. The above-mentioned computer system includes hardware such as an OS and peripheral devices.

[0074] The above computer system also includes a homepage providing environment (or display environment) if it utilizes a WWW (World Wide Web) system. The above program may be transmitted from a computer system storing the program in a storage device or the like to another computer system via a transmission medium or by transmission waves in the transmission medium. The transmission medium for transmitting the above program refers to a medium having the function of transmitting information, such as a network (communication network) such as the Internet or a communication line (communication line) such as a telephone line.

[0075] The above program may be for realizing some of the above functions, or may be a so-called differential file (differential program) that can be realized in combination with a program already recorded in the computer system.

[0076] The above describes in detail an embodiment of the present invention with reference to the drawings, but the specific configuration is not limited to this embodiment, and includes designs within the scope of the gist of the present invention.

[0077] 100; Information processing device 100A; Information processing device 110; Processor 110A; Processor 111; Speech recognition unit 112; Speech understanding unit 113; Command information generation unit 113A; Command information generation unit 114; Response information acquisition unit 115; Conversion information generation unit 115A; Conversion information generation unit 116; Control unit 116A; Control unit 117; Role extraction unit 118; Output format extraction unit 120; Memory 130; Communication unit 140; Display unit 200; Network 300; Machine learning device 300A; Machine learning device

Claims

1. An information processing device that communicates with a machine learning device having a machine learning function, inputs command information to the machine learning device, and acquires response information to the command information from the machine learning device, comprising: a speech recognition unit that performs recognition processing of a user's spoken voice; an utterance understanding unit that understands the meaning of the user's utterance based on the recognition result of the speech recognition unit; a command information generation unit that generates command information to be input to the machine learning device based on the meaning of the utterance; a response information acquisition unit that acquires the response information from the machine learning device; and a control unit that controls the display of information related to the command information and information related to the response information, wherein the command information generation unit generates the command information by converting the meaning of the user's utterance into a format that is easily understandable by the machine learning device.

2. The information processing device according to claim 1, characterized in that the command information generating unit generates the command information by converting the meaning of the user's utterance into a sentence conforming to a correct sentence pattern.

3. An information processing device as described in claim 1 or 2, further comprising a conversion information generation unit that generates conversion information that converts the response information into a format that is easy for the user to understand, and the control unit displays the conversion information when displaying information related to the response information.

4. An information processing apparatus according to claim 3, characterized in that said conversion information generating section generates said conversion information by converting said response information into a sentence conforming to a correct sentence pattern.

5. The information processing device according to claim 3, characterized in that the response information acquisition unit acquires the response information from a reliable machine learning device.

6. The information processing device of claim 3, wherein the command information generation unit generates the command information by adding information regarding the role to be given to the machine learning device, extracted from the meaning of the user's utterance, and information regarding the output format of information expected by the user, and the conversion information generation unit generates the conversion information by converting the response information based on the output format of information expected by the user.

7. An information processing device that communicates with a machine learning device having a machine learning function, inputs command information to the machine learning device, and acquires response information to the command information from the machine learning device, comprising one or more processors and one or more memories communicatively connected to the one or more processors, wherein the one or more processors include: a speech recognition unit that performs recognition processing on a user's spoken voice; an utterance understanding unit that understands the meaning of the user's utterance based on the recognition result of the speech recognition unit; a command information generation unit that generates command information to be input to the machine learning device based on the meaning of the utterance; a response information acquisition unit that acquires the response information from the machine learning device; and a control unit that controls the display of information regarding the command information and information regarding the response information, wherein the command information generation unit generates the command information by converting the meaning of the user's utterance into a format that is easily understandable by the machine learning device.

Images