Ai personality system, program for ai personality system, radio broadcasting platform, and program for radio broadcasting platform

Abstract

To provide a AI personality system, a program for the AI personality system, a radio broadcasting platform, and a program for the radio broadcasting platform capable of providing a musical piece corresponding to an atmosphere.SOLUTION: AI personality system 100 includes condition setting unit 300 to which a predetermined condition determined by a possessor in a predetermined space is AI, control unit 200 that creates an instruction in accordance with a predetermined music request and / or posting and the predetermined condition of condition setting unit 300, AI box unit 400 to which the instruction created by control unit 200 is input, audio unit 500 that synthesizes text output from input box unit 400 as narration in a predetermined voice, and playback unit 600 that plays back the narration synthesized by audio unit 500 and the music of the music request.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to an AI personality system and a program for the AI ​​personality system that automatically plays songs when a specific song request and / or post is made, a radio broadcasting platform, and a program for the radio broadcasting platform. [Background technology]

[0002] For example, Patent Document 1 (JP 2021-5114 A) discloses an information output device and an information output method for outputting music-related information to enhance the performance effect of the music being played.

[0003] The information output device described in Patent Document 1 comprises: a generation unit that selects one template from a plurality of template phrases based on the aspect of the song identified from at least one of the song's attribute information and feature information, and generates song-related information that includes the content of the attribute information in the template phrase; and a control unit that controls the output timing of the song-related information of the subsequent song based on the output duration of the song-related information of the subsequent song.

[0004] Furthermore, Patent Document 2 (Japanese Patent Laid-Open Publication No. 2011-209570) discloses a karaoke device that can introduce each song without creating a sense of incongruity.

[0005] The karaoke device described in Patent Document 2 includes an audio database in which audio fragments are registered, a sentence extraction unit that extracts sentences for each karaoke song, a melody database in which melodies are registered, and a singing processing unit that reads out the audio database based on the sentences extracted by the sentence extraction unit and the melodies registered in the melody database, and sings the sentences for each karaoke song.

[0006] Furthermore, Patent Document 3 (JP 2018-112667 A) discloses an information output device that outputs music-related information to enhance the presentation effect of the music being played.

[0007] The information output device described in Patent Document 3 includes a generation unit that selects one template from a plurality of templates based on the aspect of the song identified from at least one of the song's attribute information and feature information, and generates song-related information that includes the content of the attribute information in the template; and a control unit that controls the output timing of the song-related information of the subsequent song based on the relationship between the output duration of the song-related information of the subsequent song, the duration of each non-vocal section of the preceding song, and the duration of each non-vocal section of the subsequent song, and characteristic information related to listening to each non-vocal section of the subsequent song.

[0008] Furthermore, Patent Document 4 (Japanese Patent Laid-Open Publication No. 2004-62769) discloses a content output device that is capable of outputting content with an appropriate length of time depending on the situation.

[0009] The content output device described in Patent Document 4 selects desired content data from a content database in which a plurality of content data are stored using a selection means, and outputs the selected content data using an output means.The content output device further includes a designation means for designating a target time as a selection criterion for the content data to be output.The content database stores each content data along with the output time length required for output.The selection means searches the content database, compares the output time length of each content data with the target time designated by the designation means, and selects content data whose comparison results satisfy predetermined conditions.

[0010] Furthermore, Patent Document 5 (JP Patent Publication No. 2010-79091) discloses an audio output device that can output an audio message more effectively when an audio message to be played back occurs during content playback, without interfering with the playback of the content as much as possible.

[0011] The audio output device described in Patent Document 5 is an audio output device for audibly outputting audio messages generated during playback of content, and includes a content acquisition unit that acquires content of a predetermined section from the content being played back, a generation unit that classifies the acquired content into multiple sections and generates attribute information for the content of each section of the multiple sections, an output method determination unit that determines a control method for audio output of the message based on at least the attribute information for each section of the content, and an output audio playback unit that audio outputs the message in accordance with the control method. [Prior art documents] [Patent documents]

[0012] [Patent Document 1] Patent Publication No. 2021-5114 [Patent Document 2] Japanese Patent Application Laid-Open No. 2011-209570 [Patent Document 3] Japanese Patent Application Publication No. 2018-112667 [Patent Document 4] Japanese Patent Application Laid-Open No. 2004-62769 [Patent Document 5] Japanese Patent Application Laid-Open No. 2010-79091 Summary of the Invention [Problem to be solved by the invention]

[0013] As described in Patent Documents 1 to 5, inventions related to AI personality systems have been made. However, there is a problem in that music that does not match the atmosphere of the venue is played. Specifically, fast-paced music such as hard rock may be played in a coffee shop where quiet background music would be appropriate, or calm music may be played at a festival venue where excitement is expected.

[0014] In addition, in recent years, many generative AI natural language processing models such as ChatGPT have been developed, and it is hoped that they will be put to effective use. However, even with natural language processing models, results can vary significantly depending on how instructions are expressed, and it is not easy for owners who are unfamiliar with using generative AI, such as the manager of an izakaya restaurant, to give instructions to generative AI in appropriate terms.

[0015] The main object of the present invention is to provide an AI personality system and a program for the AI ​​personality system, a radio broadcasting platform, and a program for the radio broadcasting platform that can provide music and narration according to the atmosphere. Another object of the present invention is to provide an AI personality system and a program for the AI ​​personality system that allows an izakaya manager who is unfamiliar with using generative AI to easily use generative AI to provide music and narration that suits the atmosphere. [Means for solving the problem]

[0016] (1) An AI personality system according to one aspect includes a condition setting unit into which predetermined conditions set by an owner of a predetermined space are input, a control unit that creates instructions in accordance with predetermined song requests and / or posts and the predetermined conditions of the condition setting unit, an AI box unit into which the instructions created by the control unit are input, a voice synthesis unit that synthesizes text output from the AI ​​box unit as narration in a predetermined voice, and a playback unit that plays the narration synthesized by the voice synthesis unit and the song of the song request.

[0017] In this case, the control unit creates instructions based on the specified song requests and / or posts and the conditions entered in the condition setting unit, so that song requests and / or posts that do not meet the conditions entered in the condition setting unit can be excluded. Furthermore, since instructions can be given to the AI ​​box section according to the conditions in the condition setting section, the text output from the AI ​​box section can be made to match the conditions. However, in order to ensure that the text output from the AI ​​box is as expected, the instructions in natural language input to the AI ​​box are important.

[0018] (2) In the AI ​​personality system according to one aspect of the second invention, the specified space may include at least one of a store, a radio station, or a venue, and the specified conditions may include at least one of the atmosphere of the store, radio station, or venue, average customer spending, interior design, service, customer demographics, age, date and time, music genre, and the personality of the AI ​​personality.

[0019] In this case, by setting predetermined conditions in a predetermined space, it is possible to play back text that matches the atmosphere or situation of the place, and also to provide music. In addition, by setting the personality of the AI ​​personality, it is possible to easily play music or text that matches the atmosphere of a specified space.

[0020] (3) In a third aspect of the present invention, an AI personality system is an AI personality system in which a predetermined condition is selected by the owner from a plurality of options provided by a condition setting unit, and the control unit has a database in which instructions to the AI ​​box unit corresponding to each of the options from the condition setting unit are recorded, and the instructions to the AI ​​box unit may be created based on song requests and / or posts, the options selected by the owner, and the database.

[0021] The AI ​​box uses generative AI such as ChatGPT. Generative AI can be used to give instructions in natural language, but for izakaya managers and others who are unfamiliar with generative AI, it is often difficult to know how to express their instructions so that the AI ​​box will output the desired narration text that matches the atmosphere of the restaurant. The AI ​​personality system of the third invention provides the owner with options for the type of store, customer demographic, average customer spending, store interior atmosphere, service format, narrator voice, and preferred narrator characteristics, and by having the owner make a selection, even an owner who is unfamiliar with generative AI can clearly indicate the conditions that should be instructed to the AI ​​box. The control unit also has a database that records instructions in natural language to cause the AI ​​box to output the expected text for each option in the condition setting unit. The control unit can then refer to this database to create instructions, allowing the AI ​​box to output text as the owner expects.

[0022] (4) In the AI ​​personality system according to a fourth aspect of the present invention, the predetermined song requests and / or posts may be posted by listeners of the AI ​​personality system on a social networking site or website associated with at least one of the store, radio station, or venue, along with a pen name, handle name, real name, radio name, nickname, username, hash name, avatar name, or a desire to remain anonymous.

[0023] In this case, the AI ​​personality system can create text based on information posted by the user on social media or the web, and can play the text and provide music. As a result, the user or listener can enjoy the experience while feeling like a participant. Furthermore, other users may show their reactions through social media or the web. Specifically, for example, a user may read a QR code (registered trademark) set for each store with their smartphone, connect to the social media service LINE, and post a message. Furthermore, other users or listeners may express their opinions by leaving a "Like" or comment on the LINE message.

[0024] (5) In the AI ​​personality system according to a fifth aspect of the present invention, when the same specified song request is made multiple times within a specified time period, the control unit may issue instructions as a single specified song request.

[0025] In this case, if a specific song is requested multiple times within a specific time period, the control unit can play them as a single specific song request. In other words, if the same song is requested multiple times and it is not desirable to play the same song multiple times, the control unit can combine them into a single playback. As a result, playback can be tailored to the atmosphere of the venue.

[0026] (6) A program of an AI personality system according to another aspect includes a condition setting process in which specified conditions set by the owner of a specified space are input, a control process in which instructions are created in accordance with the specified song request and / or post and the conditions of the condition setting process, an AI box process in which the instructions created by the control process are input, a voice synthesis process in which the text output from the AI ​​box process is synthesized as narration in a specified voice, and a playback process in which the narration synthesized by the voice synthesis process and the requested song are played back.

[0027] In this case, the control unit creates instructions based on the specified song requests and / or posts and the conditions entered in the condition setting unit, so that song requests and / or posts that do not meet the conditions set by the condition setting unit can be excluded. Furthermore, since instructions can be given to the AI ​​box section according to the conditions in the condition setting section, the text output from the AI ​​box section can be made to match the conditions.

[0028] (7) According to yet another aspect, a radio broadcasting platform includes an AI personality system described in any one of the first to fifth aspects of the present invention, a reward provision unit that provides rewards based on the number of listener responses, a branding setting function that can set at least one of program corner and / or timetable setting functions, an advertising distribution function that can deliver advertisements along with at least one of time signals, weather forecasts, news, and event announcements, and a broadcast content control function that can set at least one of the timing of advertising broadcasts, the order in which songs are played, and the consistency of the same songs.

[0029] In this case, depending on the listeners' reactions, it is possible to smoothly provide special benefits, create various segments as part of the program, automatically play program segments, announce the time, weather forecasts, news, and events, broadcast advertisements, change the order in which songs are played, and standardize the use of the same songs.

[0030] (8) A radio broadcasting platform program according to another aspect further includes a program of an AI personality system according to another aspect, a reward provision process that provides rewards according to the number of listener responses, a branding setting process that can set at least one of program corner and / or timetable setting functions, an advertising distribution process that can distribute advertisements along with at least one of time signals, weather forecasts, news, and event announcements, and a broadcast content control process that can set at least one of the timing of advertising broadcasts, the order in which songs are played, and the consistency of the same songs.

[0031] In this case, depending on the listeners' reactions, it is possible to smoothly provide special benefits, create various segments as part of the program, automatically play program segments, announce the time, weather forecasts, news, and events, broadcast advertisements, change the order in which songs are played, and standardize the use of the same songs. [Brief explanation of the drawings]

[0032] [Figure 1]FIG. 1 is a schematic diagram illustrating an example of an AI personality system. [Figure 2] FIG. 10 is a schematic diagram illustrating an example of a processing operation between a control unit and a smartphone of a manager or the like via a condition setting unit. [Figure 3] FIG. 10 is a schematic diagram showing an example of condition settings on a smartphone. [Figure 4] FIG. 10 is a schematic diagram illustrating an example of a processing operation between a control unit and a user's smartphone. [Figure 5] FIG. 10 is a schematic diagram showing an example of a request made by a user's smartphone after reading a QR code (registered trademark). [Figure 6] FIG. 2 is a schematic diagram showing an example of processing operations of a control unit and an AI box unit. [Figure 7] FIG. 2 is a schematic diagram showing an example of processing operations of a control unit and a playback unit. [Figure 8] FIG. 1 is a schematic diagram illustrating an example of a radio broadcasting platform. [Figure 9] FIG. 10 is a schematic diagram illustrating an example of a branding setting function. [Figure 10] FIG. 10 is a schematic diagram showing an example of selecting a branding setting function on a user's smartphone. [Figure 11] 7 is a schematic diagram showing another example of the processing operation of the control unit and the AI ​​box unit shown in FIG. 6. FIG. [Figure 12] FIG. 10 is a schematic diagram illustrating an example of an operation process of a radio broadcasting platform. [Figure 13] FIG. 7 is a schematic diagram showing another example of the operation process of the song request corner shown in FIG. 6. [Figure 14] FIG. 10 is a schematic diagram showing an example of the operation process of a regular letter request or a "ignorance is bliss" corner request. [Figure 15] FIG. 10 is a schematic diagram illustrating an example in which a control unit performs processing according to the number of listener responses to a request. DETAILED DESCRIPTION OF THE INVENTION

[0033] Hereinafter, an embodiment of the present invention will be described with reference to the drawings. In the following description, the same components are designated by the same reference numerals. Furthermore, when the reference numerals are the same, the names and functions of the components are also the same. Therefore, detailed descriptions thereof will not be repeated.

[0034] [Embodiment] (Overview of AI personality system 100) 1 is a schematic diagram showing an example of an AI personality system 100. In the following, an example in which the AI ​​personality system 100 is applied to a store will be described, but the system may also be applied to any other venue, such as a radio station or venue, a concert venue, a wedding venue, a food festival venue, an exhibition venue, or any other venue.

[0035] As shown in FIG. 1, the AI ​​personality system 100 mainly includes a control unit 200, a condition setting unit 300, an AI box unit 400, an audio unit 500, and a playback unit 600. In addition, the AI ​​personality system 100 is connected to a smartphone 700 and a music database server 800 via the Internet 900. In this embodiment, the control unit 200 , the condition setting unit 300 , the AI ​​box unit 400 , the audio unit 500 and the playback unit 600 may also be connected via the Internet 900 . The smartphones 700 include at least a smartphone 710 of a manager or store manager, etc., and a smartphone 720 of a user or a listener of music.

[0036] (control unit 200) The control unit 200 is composed of a personal computer, tablet terminal, mobile terminal, etc. that includes a CPU (Central Processing Unit) and a recording unit. The control unit 200 also stores a program for the AI ​​personality system 100, which will be described later.

[0037] (Condition setting unit 300) The condition setting unit 300 can set the conditions for the atmosphere of the place at least for each store or venue. In other words, by setting the conditions for the atmosphere of the place, it is possible to set the personality of the AI ​​personality system 100 from various options.

[0038] (AI box part 400) The AI ​​box unit 400 includes a generative AI chat service system, such as ChatGPT, Perplexity AI, MICROSOFT Copilot, ChatSonic, YouChat, Gemini, Notion AI, etc.

[0039] (Audio section 500) The audio unit 500 stores the voices used in the AI ​​personality system 100, and records and stores the synthesized voices used when playing back the text described below, as well as the basic voices of store managers, shop owners, part-time workers, entertainers, actors, voice actors, etc. The voice unit 500 uses the stored synthesized voice or basic voice to synthesize the text output from the AI ​​box unit.

[0040] (Playback section 600) The playback unit 600 is an audio output device such as a speaker that can output the synthesized voice and music. The playback unit 600 may be connected by wire or wirelessly via Bluetooth (registered trademark) or the like.

[0041] FIG. 2 is a schematic diagram showing an example of processing operation between the control unit 200 and a smartphone 710 of a manager or store manager, etc. via the condition setting unit 300, and FIG. 3 is a schematic diagram showing an example of condition setting on the smartphone 710.

[0042] As shown in FIG. 2, the control unit 200 causes the condition setting unit 300 to provide a condition setting screen to the smartphone 710 of the manager or store manager (step S21). As shown in Fig. 3, the conditions setting screen includes options such as store type, customer demographic, average customer spending, atmosphere, interior design, service format, and audio selection, which can be selected by the store manager using a pull tab. Note that the conditions setting options are not limited to these options, and any other conditions may be added.

[0043] For example, specific store types can be selected from pull tabs, including standing bars, izakayas, family restaurants, casual dining, fine dining, bistros, cafes, bars, food trucks, Japanese restaurants, ethnic restaurants, and international cuisine. In addition, the customer demographic for this store type can be selected from pull tabs, including families, groups of friends, special occasions, business travelers, casual dates, solo customers, event participants, young people, adults, people interested in different cultures, office workers, and children.

[0044] In addition, the average customer spending for that store type can be selected from the pull tab, such as 0 yen to less than 1,000 yen, 1,000 yen to less than 3,000 yen, 3,000 yen to less than 5,000 yen, 5,000 yen to less than 10,000 yen, 10,000 yen to less than 30,000 yen, or more. Furthermore, the atmosphere of the store type can be selected from a pull tab, including relaxed, friendly, luxurious, elegant, comfortable, warm, cool, mature, casual, quiet, calm, exotic, lively, vibrant, bright, etc.

[0045] In addition, the interior of each store type can be selected from a pull tab, including simple, casual, elegant, sophisticated, warm design, cozy design, natural, dim lighting, counter seating, movable, traditional Japanese style, reflecting the country or region, focusing on space efficiency, bright and spacious, etc. In addition, the service format for the store type can be selected from self-service, full-service, a mixture of self-service and full-service, etc. using a pull tab.

[0046] Furthermore, the voice selection for the store type can be male, female, automatic voice reading, voice actor, popular famous DJ, the manager's voice, the store owner's voice, the voice of a part-time worker, etc. In the case of the manager's voice, the store owner's voice, the voice of a part-time worker, etc., it is possible to respond by having the manager read out a predetermined sentence several times and recording and storing it in the voice unit 500.

[0047] Furthermore, a personality of the AI ​​personality system 100 may be created as a condition setting. For example, a natural language setting such as "You are Adam, a popular and famous DJ personality on Buttobi Radio who can talk in a lively and friendly manner and create a lively atmosphere. You are knowledgeable about rock and J-Pop and can liven up the bar with your well-paced music selections" may be separately set.

[0048] The condition setting is input using the smartphone 710 of the manager or store manager, etc. (step S22), and transmitted to the condition setting unit 300. The condition setting unit 300 receives the input condition setting (step S23) and instructs the smartphone 710 of the manager or store manager to input a QR code (registered trademark) (step S24).

[0049] The manager or store manager uses the smartphone 710 to transmit the URL of the SNS as a QR code (registered trademark) to the condition setting unit 300 (step S25). Specifically, the manager provides a website or LINE, etc., as a QR code (registered trademark) where users or people requesting songs can post. In other words, users, listeners, or people requesting songs can read the QR code (registered trademark) using a smartphone 720, connect to a specified site or LINE, and post or request the song or text they want to hear. The QR code (registered trademark) may be displayed on a large display or screen installed in the store, or may be displayed on a medium such as a menu.

[0050] Next, the condition setting unit 300 receives the QR code (registered trademark) (step S26). The condition setting unit 300 transmits all the conditions, that is, the content of the condition settings and the QR code (registered trademark), to the control unit 200 (step S27).

[0051] FIG. 4 is a schematic diagram showing an example of the processing operation between the control unit 200 and the user's smartphone 720, and FIG. 5 is a schematic diagram showing an example of a request from the user's smartphone 720 that reads a QR code (registered trademark).

[0052] 4, the control unit 200 provides a QR code (registered trademark) to the store (step S31). The user reads the QR code (registered trademark) using their smartphone 720 (step S32) and connects to the store's dedicated SNS. Specifically, as shown in FIG. 5, by reading a QR code (registered trademark), a store-specific SNS is displayed on a smartphone 720. 5, the message "Welcome to ●●! Request your favorite song!" is displayed, and the user of smartphone 720 inputs a song request (step S33). Note that the example of FIG. 5 illustrates a case where LINE is used as the SNS.

[0053] Next, the control unit 200 receives the request input by the smartphone 720 (step S34), and transmits a thank you email to the smartphone 720 (step S35). As shown in FIG. 5, the thank you email is displayed on the user's smartphone 720 (step S36). The contents of the request may be displayed on a display device such as a large screen in the store, a display, a large display, or an individual display.

[0054] FIG. 6 is a schematic diagram showing an example of the processing operations of the control unit 200 and the AI ​​box unit 400. As shown in FIG.

[0055] The control unit 200 transmits the request received from the user to the AI ​​box unit 400 (step S41). The AI ​​box unit 400 receives the request (step S42). Next, the control unit 200 creates instructions based on the condition settings entered by the owner and transmits them to the AI ​​box unit 400 (step S43). The condition settings include at least one of the atmosphere of the store, radio station, or venue, average customer spending, interior design, service, customer demographics, age, date and time, music genre, and the personality of the AI ​​personality. Each of these conditions is provided as a choice, and the store manager or the like can set the conditions by selecting from a pull tab. The control unit 200 has a database that records instructions to the AI ​​box unit 400 corresponding to each option in the condition setting unit 300, and creates instructions to the AI ​​box based on the conditions set by the owner and the database. The AI ​​box unit 400 analyzes the request sentence according to the condition settings and creates a sentence to be used as a response voice (step S44).

[0056] In this case, the AI ​​box unit 400 may make a judgment based on the character or personality of the AI ​​personality in the condition settings described above, and the condition settings may also be set as follows: "Introduce the requested song from the listener in about 45 seconds, full of unique content, including an anecdote or message from the listener, information about the requested song, etc. However, if the listener has a radio name, pen name, or handle name, please use that as the listener's name. Also, if you wish to remain anonymous, please do not use your real name."

[0057] Next, the AI ​​box unit 400 generates a response voice in accordance with the condition settings (step S45). For example, the AI ​​Box 400 may generate a response voice saying, "Good evening, everyone on Crazy Radio! Today's request is from Natsuko, from Osaka Prefecture. Natsuko requested, 'I wish the rainy season would end soon. I want a light-hearted song to blow away the humidity!' So, here's a lively song to blow away the humidity, 'Ah, Summer Vacation' by TUBE! When you think of summer, you think of summer vacation, the beach, summer festivals, ah, summer vacation! Thank you, Natsuko! Come on, everyone, let's have fun together!"

[0058] Furthermore, the AI ​​box unit 400 searches for and saves the music (step S46). In this case, the AI ​​box unit 400 acquires music from a site or the like that has been approved or certified by a certification organization such as JASRAC. As a result, the following is acquired: album NATSU, release year 1990, genre J-POP, artist TUBE, song title "Ah Summer Vacation," lyrics by Nobuteru Maeda, music by Michiya Haruhata & Nobuteru Maeda. Finally, the AI ​​box unit 400 transmits the response voice and the music to the control unit 200, and the control unit 200 stores the response voice and the music (step S47).

[0059] 7 is a schematic diagram showing an example of the processing operations of the control unit 200, the audio unit 500, and the playback unit 600. The playback unit 600 includes an audio output device such as a speaker, a television, or a projector.

[0060] The control unit 200 recognizes that the radio playback WEB has been launched in the store (step S51). Next, the control unit 200 receives audio data to be applied from the audio unit 500 (step S52). Note that instead of receiving audio data at any time, the control unit 200 may receive and record the audio data from the audio unit 500 in advance.

[0061] Next, a response voice instruction is output to the playback unit 600 (step S53). As a result, the playback unit 600 outputs the sentence generated by the AI ​​box unit 400 in a predetermined voice (step S54). Next, the control unit 200 issues a song instruction to the playback unit 600 (step S55), which causes the playback unit 600 to play the song (step S56).

[0062] The following describes a case where a radio broadcasting platform 1000 is constructed using the above-described AI personality system 100. FIG.

[0063] As shown in FIG. 8, the radio broadcasting platform 1000 includes a branding setting function 1100, a broadcast content control function 1200, and an advertisement distribution function 1300 in addition to the functions described above. The control unit 200 of the AI ​​personality system 100 is in a state where it can communicate with the branding setting function 1100, broadcast content control function 1200, and advertisement distribution function 1300, and is able to exchange information.

[0064] The broadcast content control function 1200 can also set the timetable for program segments, the timing of advertising broadcasts (described later), the order in which songs are played, and the consistency of the same songs. Furthermore, the advertisement distribution function 1300 can purchase and set advertisement slots for specific broadcast times. Specifically, these are for time announcements, weather forecasts, news, event announcements, etc. As a result, the owner or store manager can earn advertising revenue. For example, the time can be announced in one- or two-hour increments in a cheerful tone, such as "Claire will announce 12 o'clock," and the weather forecast, if sunny, can say in a cheerful tone, "Today's chance of rain is 0%, with a maximum temperature of 26 degrees and a minimum of 21 degrees, making it a comfortable day," or if it will rain, can say in a calm tone, "Today's chance of rain is 60%, with a maximum temperature of 24 degrees and a minimum of 18 degrees, so be sure to bring an umbrella."In addition, advertisements can be made for a variety of events, not just the store of the manager or store manager, but also events held in exhibition halls, movie theaters, food festivals, kyogen, manzai, rakugo, etc.

[0065] As shown in FIG. 8, the branding setting function 1100 can further set program corner and timetable setting functions in the condition setting section 300 described above.

[0066] FIG. 9 is a schematic diagram showing an example of the branding setting function 1100. As shown in FIG.

[0067] 9, a timetable setting function and program segments can be set in the branding setting function 1100. For example, the timetable setting function and program segments can be set using a smartphone 710 of a manager, a store manager, or the like.

[0068] For example, when setting up a program segment, the segment name can be something like "Ignorance is bliss." It can also be set in natural language as follows: "The 'Ignorance is bliss' segment is a segment where listeners share stories of things they would have been happier without realizing in their daily lives. From funny stories to slightly sad ones, listeners' posts are introduced, and viewers can empathize and be surprised by the stories. The stories posted by listeners are introduced, and the personality engages in a conversation, reacting on the spot."

[0069] In addition, the timetable setting function allows you to set the time between 10:00 and 10:55 to be used for regular letter introductions, 10:55 and 11:00 to be used for advertising broadcasts, 11:00 and 11:55 to be used for song introductions, 11:55 and 12:00 to be used for advertising broadcasts, 12:00 and 13:00 to be used for the "Ignorance is Bliss" corner, and 13:00 and 14:00 to be used for random broadcasts.

[0070] FIG. 10 is a schematic diagram showing an example of selecting a branding setting function 1100 on a user's smartphone 720.

[0071] As shown in FIG. 10, after the user reads the QR code (registered trademark) using their smartphone 720 in step S32 described above, a connection is made to the store's dedicated SNS. In this case, a selection area for song requests, regular mail requests, and "Ignorance is bliss" corner requests is displayed in the lower half of the smartphone 720.

[0072] If a song request is selected, the above-mentioned processes from step S33 to step S36 are carried out. Furthermore, if a user selects a Futsuota request, they can scan the QR code and write something like, "Nagasaki Prefecture, Radio Name: Tobidashi Boy. Thank you for always providing us with such a fun program. I look forward to the 'Angry' segment every week without fail. I especially laughed out loud at last week's episode. The stories from the listeners were so funny that they helped me blow away all the fatigue from work." Details of these will be provided later.

[0073] Furthermore, if a request for the "Ignorance is bliss" section is selected, the user scans the QR code on an SNS, such as LINE, and writes something like, "Tokyo, Radio name Mizotaro. One day, in the break room at work, I overheard my colleagues secretly planning a birthday surprise for me. I was anxiously waiting for that day, but it turned out to be my colleague's birthday, not mine. Ignorance really is bliss." Details of these will be provided later.

[0074] Next, FIG. 11 is a schematic diagram showing another example of the processing operation of the control unit 200 and the AI ​​box unit 400 shown in FIG.

[0075] As shown in FIG. 11, when transmitting a request to the AI ​​box unit 400 in the process of step S41 in FIG. 6, the control unit 200 transmits the request received from the user and the type of program corner setting (step S61). Then, the AI ​​box unit 400 receives the request and the type of program corner setting (step S62). A song request is processed in the same way as in Fig. 6, but in the case of a Futsuota corner request or an "Ignorance is Bliss" corner request, after the process of step S45, the AI ​​box unit 400 searches for a song in the AI ​​box unit 400 and stores the response voice without waiting for the process of storing it (step S67).

[0076] In this embodiment, the control unit 200 transmits the type of program corner setting to the AI ​​box unit 400, but this is not limited to this, and the AI ​​box unit 400 may independently grasp the content of the request and determine the type of program corner setting.

[0077] Furthermore, in the sentence analysis and creation process of the AI ​​box unit 400 in step S44, a determination is made as to whether the requested song matches the set conditions, and if it does not match the set conditions, the control unit 200 may display on the user's smartphone 720 a message saying "The song in question is not in the set genre and cannot be played. We apologize for the inconvenience," and the request may be recorded in the control unit 200 as a rejected request.

[0078] Next, Figure 12 is a schematic diagram showing an example of the operational processing of the radio broadcasting platform 1000, Figure 13 is a schematic diagram showing another example of the operational processing of the song request corner shown in Figure 6, and Figure 14 is a schematic diagram showing an example of the operational processing of a Futsuota request or an "Ignorance is Bliss" corner request.

[0079] As shown in FIG. 12, in the radio broadcasting platform 1000, when a radio playback web page is selected from the smartphone 710 of the manager or store manager, the control unit 200 launches the radio playback web page (step S71). Next, the control unit 200 reads the radio program (step S72). In this case, the radio program illustrated in FIG. 9 of the branding setting function 1100 is read.

[0080] The control unit 200 determines whether the radio program at the current time is a song request corner, an unknown is a Buddhist corner, a normal letter corner, or an advertisement broadcast (step S73).

[0081] If the control unit 200 determines in the process of step S73 that it is a song request corner, it performs the song request corner process (step S74) and performs playback processing using the personality and voice of the set personality (step S75). If the control unit 200 determines in the processing of step S73 that the corner is the "I don't know but it's a buddha" corner, it performs the processing of the "I don't know but it's a buddha" corner (step S76) and performs playback processing using the personality and voice of the set personality (step S77).

[0082] If the control unit 200 determines in the process of step S73 that it is a Futsuota corner, it performs the Futsuota corner process (step S78) and performs playback processing using the personality and voice of the set personality (step S79). If the control unit 200 determines in the process of step S73 that the broadcast is an advertisement broadcast, it processes the advertisement broadcast (step S80) and performs playback processing using the personality and voice of the set personality (step S81).

[0083] After the playback processing in steps S75, S77, S79, and S81, the control unit 200 returns to step S72 and repeats the processing.

[0084] Also, as shown in FIG. 13, in the song request corner, unlike FIG. 6, it may be determined whether or not the same song exists within a predetermined time period (step S48). Specifically, a popular song may be requested from multiple users' smartphones 720. For example, if the song "Columbus" by artist Mrs. GREEN APPLE is requested by multiple users, including a user with a handle name A and a user with a handle name B, within a predetermined time period, the song may be played while quoting a comment from either the user with the handle name A or the user with the handle name B. Specifically, the process of sentence analysis and creation in step S44 may start with a request from the user with the handle name A, and after the user with the handle name A makes a comment, it may be determined that a similar song request has also been received from the user with the handle name B.

[0085] Furthermore, if a song request is made that does not match the conditions set in accordance with the request reception and condition settings, the AI ​​box unit 400 stops processing the request. In other words, if the conditions set are J-POP, rock, etc., but a classical song such as "farewell song" is requested, the request is discarded.

[0086] In the above description, the AI ​​box unit 400 performs the process of step S48 or the process of discarding the request, but this is not limitative, and the control unit 200 may perform this process.

[0087] 14, in the operation processing of a "Futsuota" corner request, the control unit 200 transmits a request received from a user to the AI ​​box unit 400 (step S91). The AI ​​box unit 400 receives the request (step S92). Next, the control unit 200 transmits a condition setting to the AI ​​box unit 400 (step S93).

[0088] Here, the AI ​​box unit 400 determines whether the request contains any prohibited words (step S94). In addition to prohibited words, words that are voluntarily defined by each broadcasting station, such as words that should be warned not to be broadcast or words that should be refrained from being broadcast, may also be used. If it is determined that the request does not contain any prohibited words, the AI ​​box unit 400 analyzes the request sentence in accordance with the condition settings and creates a sentence to be used as a response voice (step S94).

[0089] In this case, the AI ​​box unit 400 may make a judgment based on the personality of the personality in the condition setting described above, and the condition setting may also be set as follows: "Please introduce the Futsuoto. When reading the Futsuoto, it is important to remember to show respect and gratitude to the listener and to empathize with the content. If there is a nickname, please call it out. Don't forget to react according to the content of the letter."

[0090] Next, the AI ​​box unit 400 generates a response voice in accordance with the condition settings (step S96). For example, the AI ​​Box 400 may generate a response voice saying, "Oh, we'll introduce the Futsuota corner from Tobidashi Boy from Nagasaki Prefecture. 'Thank you for always providing us with such an enjoyable program. I look forward to the 'Angry' corner every week without fail. I especially laughed out loud at last week's episode. The stories from our listeners were so funny that they helped me blow away all my fatigue from work.' Thank you for your passionate message from Nagasaki Prefecture! The stories from our listeners really help me blow away all my fatigue. Thank you."

[0091] Finally, the AI ​​box unit 400 transmits the response voice to the control unit 200, and the control unit 200 stores the response voice (step S97).

[0092] Also, in FIG. 14, in the case of the "Ignorance is bliss" corner, the AI ​​box unit 400 may make a judgment based on the personality of the personality in the condition setting, and the condition setting may be set as follows: "The "Ignorance is bliss" corner is a corner where listeners share stories in their daily lives where they would have been happier if they had not noticed them. We will introduce posts from listeners ranging from funny stories to slightly sad stories, and we ask that you introduce the stories posted by listeners in an easy-to-understand manner without changing the main theme, and that you engage in a conversation that includes reactions, so that listeners can empathize and be surprised through the stories."

[0093] Next, the AI ​​box unit 400 generates a response voice in accordance with the condition settings (step S96). For example, specifically, the AI ​​box unit 400 may generate a response voice saying, "Hello, listeners, this is the crazy personality Adam. We are presenting a unique corner, 'Ignorance is bliss'! This is a request from Mizotaro from Tokyo. It is a sad and joyful story about a man who was planning a birthday surprise in the break room, but it turned out to be his colleague's birthday, not his own. Ignorance is bliss, and it really is true!"

[0094] FIG. 15 is a schematic diagram showing an example of processing performed by the control unit 200 depending on the number of listener responses to a request.

[0095] As shown in FIG. 15, if a user who is a listener feels that a comment or song played from the playback unit 600 is good, the user can leave a like and / or comment from the smartphone 720 (step S101).

[0096] Next, the control unit 200 receives the number of reactions from the listeners from the smartphone 720 (step S102). Here, the number of listener reactions refers to the number of likes and / or comments from listeners in response to requests in the song request, the "Ignorance is a Buddha" corner, and the "Futsuota" request corner.

[0097] The control unit 200 determines whether the number of responses from listeners exceeds 10,000 (step S103). If the number of listener responses exceeds 10,000, the control unit 200 notifies the smartphone 720 that a special benefit has been received, and the user of the smartphone 720 can receive the special benefit (step S104). This allows the user to receive the special benefit, and the manager or store manager can entertain customers by livening up the web radio. For example, if the number of listener responses at an izakaya exceeds 10,000, a coupon for a free drink or a 30% discount on food and drink will be offered. Note that once a special benefit is obtained, the number of listener responses will be reset.

[0098] Furthermore, the control unit 200 preferentially plays back songs or sentences that have received more than 10,000 listener responses during the random broadcast time set by the branding setting function 1100 shown in Fig. 9 (step S105). That is, the process of step S41 described in Fig. 6 may be repeatedly executed.

[0099] Furthermore, if the number of listener responses does not exceed 10,000 but exceeds 1,000 (step S106), the control unit 200 may notify the smartphone 710 and display the result on a screen or the like in the store (step S107). Specifically, for example, "Handle name A, 1,000 listener responses!! Congratulations!!" This allows the user to receive praise, and the manager or store manager can entertain customers by livening up the web radio.

[0100] In addition, the control unit 200 may preferentially play back the song or sentence that has received more than 1,000 listener responses during the random broadcast time set by the branding setting function 1100 shown in Figure 9 following the processing of step S104 (step S106).

[0101] Furthermore, in the above embodiment, playback is prioritized according to the number of listener responses, but this is not limited to this, and during random broadcast times, playback may be prioritized according to the number of listener responses, or playback may be prioritized for people who have invested in the above advertisements, or people who have invested in the store, or people who have made some kind of voluntary charge to the store, or people who are regular customers of the store, or playback priorities may be changed at the discretion of the management.

[0102] Furthermore, although the control unit 200 has been described above as being different from the AI ​​box unit 400, this is not limiting and the control unit 200 may be formed integrally. In particular, by including a generative AI chat service system in the control unit 200, the control unit 200 can be integrated with the AI ​​box unit 400. Examples of generative AI chat services (including interactive types) include service systems such as ChatGPT, Perplexity AI, MICROSOFT Copilot, ChatSonic, YouChat, Gemini, and Notion AI.

[0103] In the present invention, the condition setting unit 300 corresponds to the "condition setting unit", the control unit 200 corresponds to the "control unit", the AI ​​box unit 400 corresponds to the "AI box unit", the audio unit 500 corresponds to the "audio synthesis unit", the playback unit 600 corresponds to the "playback unit", the AI ​​personality system 100 corresponds to the "AI personality system", the control unit 200 corresponds to the "privilege provision unit", the branding setting function 1100 corresponds to the "branding setting function", the advertising distribution function 1300 corresponds to the "advertising distribution function", the broadcast content control function 1200 corresponds to the "broadcast content control function", and the radio broadcasting platform 1000 corresponds to the "radio broadcasting platform".

[0104] Although the preferred embodiment of the present invention has been described above, the present invention is not limited thereto. It will be understood that various other embodiments can be made without departing from the spirit and scope of the present invention. Furthermore, although the actions and effects of the configuration of the present invention are described in the present embodiment, these actions and effects are merely examples and do not limit the present invention. [Explanation of symbols]

[0105] 100: AI personality system 200: Control section 300: Condition setting section 400: AI box section 500: Audio section 600: Playback section 1000: Radio broadcasting platform 1100: Branding setting function 1200: Broadcast content control function 1300: Advertisement distribution function

Claims

1. a condition setting unit into which predetermined conditions set by an owner of a predetermined space are input; a control unit that generates instructions in accordance with a predetermined song request and / or post and a predetermined condition set by the condition setting unit; an AI box unit to which the instruction created by the control unit is input; a voice synthesis unit that synthesizes the text output from the AI ​​box unit as a narration in a predetermined voice; An AI personality system including: a playback unit that plays back the narration synthesized by the voice synthesis unit and the requested song.

2. the predetermined space includes at least one of a store, a radio station, or a venue; The AI ​​personality system of claim 1, wherein the predetermined conditions include at least one of the atmosphere of the store, radio station, or venue, average customer spending, interior design, service, customer demographics, age, date and time, music genre, and the personality of the AI ​​personality.

3. the predetermined condition is selected by the owner from a plurality of options provided by the condition setting unit, the control unit includes a database in which instructions to the AI ​​box unit corresponding to each of the options in the condition setting unit are recorded, 2. The AI ​​personality system of claim 1, wherein instructions to the AI ​​box unit are generated based on the song request and / or the post, options selected by the owner, and the database.

4. The AI ​​personality system of claim 1, wherein the predetermined song requests and / or posts are posted by listeners of the AI ​​personality system on a social networking site or website assigned to at least one of the store, radio station, or venue, along with a pen name, handle name, real name, radio name, nickname, username, hash name, avatar name, or a desire to remain anonymous.

5. 2. The AI ​​personality system according to claim 1, wherein the control unit, when the same predetermined song request is made multiple times within a predetermined time, issues an instruction as a single predetermined song request.

6. a condition setting process in which predetermined conditions set by an owner of a predetermined space are input; a control process for generating instructions in response to predetermined song requests and / or posts and the conditions of the condition setting process; an AI box process into which the instruction created by the control process is input; a voice synthesis process for synthesizing the text output from the AI ​​box process as a narration in a predetermined voice; A program for an AI personality system, comprising a playback process for playing back the narration synthesized by the voice synthesis process and the requested song.

7. An AI personality system according to any one of claims 1 to 5; a reward providing unit that provides rewards according to the number of listener responses; A branding setting function that can set at least one of a program corner and / or a timetable setting function; an advertisement distribution function capable of distributing advertisements together with at least one of a time signal, a weather forecast, news, and an event announcement; A radio broadcasting platform including a broadcast content control function that can set at least one of timing of advertisement broadcasting, order of music playback, and consistency of the same music.

8. A program for the AI ​​personality system according to claim 6; A reward provision process for providing rewards according to the number of listener responses; A branding setting process capable of setting at least one of a program corner and / or a timetable setting function; an advertisement distribution process capable of distributing advertisements together with at least one of a time signal, a weather forecast, news, and an event announcement; and a broadcast content control process capable of setting at least one of the timing of advertisement broadcast, the order of music playback, and the consistency of the same music.

Images