Information processing device, information processing method, and program

Abstract

To improve convenience of providing information on publicly-managed races.SOLUTION: An information processing device 10 acquires a question sentence 1. The information processing device 10 extracts an extracted keyword from a reply sentence 3 output by a generative model 2, which generates a response sentence for an input sentence, in response to the question sentence 1. The information processing device 10 identifies corresponding image data for the extracted keyword by referring to image information 11a, which shows correspondence between a publicly-managed race keyword and image data. The information processing device 10 outputs the corresponding image data and the reply sentence 3.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to an information processing device, an information processing method, and a program. [Background technology]

[0002] Services using AI (Artificial Intelligence) chatbots are sometimes provided in publicly managed sports events such as horse racing, bicycle racing, boat racing, and auto racing. For example, the AI ​​chatbot receives questions from users and generates responses to those questions.

[0003] As a technology related to publicly managed racing, for example, a prediction device has been proposed that can predict the actual development of a race and notify the user in an easy-to-understand manner. Also, for example, a semantic content assignment device has been proposed that, when there is a definition sentence or similar text indicating a headword and its meaning, can assign content such as appropriate images or videos to each meaning, and can widely cover multiple meanings even for minor meanings. Also, for example, an information processing device has been proposed that can change avatars depending on the user. And, for example, a pseudo-conversation device has been proposed that can adjust the content of a pseudo-conversation generated by an artificial intelligence to suit the user's preferences. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Application Publication No. 2024-10336 [Patent Document 2] Japanese Patent Application Laid-Open No. 2011-180824 [Patent Document 3] Japanese Patent Publication No. 2023-72109 [Patent Document 4] Japanese Patent Application Laid-Open No. 2012-141713 Summary of the Invention [Problem to be solved by the invention]

[0005] When users ask questions to an AI chatbot, the information provided by the AI ​​chatbot is text-based, which can be difficult for users to understand. This reduces the convenience of providing information on publicly managed racing.

[0006] On the one hand, this case aims to improve the convenience of providing information on publicly managed lotteries. [Means for solving the problem]

[0007] One proposal provides an information processing device having a memory unit and a processing unit. The memory unit stores image information indicating the correspondence between keywords for publicly managed racing games and image data. The processing unit acquires a question sentence, extracts extracted keywords from a response sentence output by a generative model that generates a response sentence to an input sentence, refers to the image information to identify corresponding image data that corresponds to the extracted keywords, and outputs the corresponding image data and the response sentence. [Effects of the Invention]

[0008] According to one aspect, it is possible to improve the convenience of providing information on publicly managed gaming. [Brief explanation of the drawings]

[0009] [Figure 1] FIG. 1 illustrates an example of an information processing device according to a first embodiment. [Figure 2] FIG. 10 illustrates an example of an information processing system according to a second embodiment. [Figure 3] FIG. 2 illustrates an example of a hardware configuration of a server. [Figure 4] FIG. 2 is a block diagram illustrating an example of the functions of a server. [Figure 5] FIG. 10 is a diagram illustrating an example of image information. [Figure 6] FIG. 10 is a diagram illustrating an example of avatar information. [Figure 7] FIG. 10 is a diagram illustrating an example of response information. [Figure 8] FIG. 10 is a diagram illustrating an example of a response process to a question about a term. [Figure 9] FIG. 10 is a diagram illustrating an example of an avatar selection process. [Figure 10] FIG. 10 is a diagram illustrating an example of a response process to a question about event information. [Figure 11] 10 is a flowchart illustrating an example of a procedure for a user response process. [Figure 12] 10 is a flowchart illustrating an example of a procedure for an avatar selection process. DETAILED DESCRIPTION OF THE INVENTION

[0010] The present embodiment will be described below with reference to the drawings. Note that each embodiment can be implemented in combination with a plurality of other embodiments within a range that does not contradict each other. [First embodiment] First, the first embodiment will be described.

[0011] FIG. 1 is a diagram showing an example of an information processing device according to a first embodiment. The first embodiment generates answers to questions about publicly managed races posed by users. The information processing device 10 is a computer that generates answers to questions about publicly managed races. The information processing device 10 is, for example, a server computer that acquires questions via a terminal operated by a user, generates answers to the acquired questions, and outputs them to the terminal. The information processing device 10 has a memory unit 11 and a processing unit 12. The memory unit 11 stores data used in processing performed by the information processing device 10. The memory unit 11 is, for example, a memory or storage device that the information processing device 10 has.

[0012] The storage unit 11 stores image information 11a. The image information 11a indicates the correspondence between keywords for publicly managed racing and image data. For example, the image information 11a stores "keirin" and "grand prix" in association with image data indicating that the grand prix is ​​the highest rank of keirin races.

[0013] The processing unit 12 controls the information processing device 10 and is capable of executing required processing. The processing unit 12 is, for example, a processor or an arithmetic circuit included in the information processing device 10. First, the processing unit 12 acquires a question sentence 1. For example, the processing unit 12 acquires, from a terminal operated by a user and installed at a public racetrack, question sentence 1 obtained by converting speech input into text from the terminal. The processing unit 12 inputs question sentence 1 to a generative model 2. The generative model 2 generates a response sentence to the input sentence. The generative model 2 is, for example, a generative AI that generates a response to the input sentence. Note that the generative model 2 may be executed by the information processing device 10 or by another computer. The generative model 2 outputs an answer sentence 3 as a response to the question sentence 1.

[0014] The processing unit 12 extracts extracted keywords from the answer sentence 3 output by the generative model 2 in response to the question sentence 1. The processing unit 12 extracts keywords registered in the image information 11a from the answer sentence 3. Here, it is assumed that the processing unit 12 has extracted "keirin" and "grand prix." The processing unit 12 identifies corresponding image data that corresponds to the extracted keywords by referring to the image information 11a. Here, the processing unit 12 identifies the image data registered in association with "keirin" and "grand prix" in the image information 11a as the corresponding image data.

[0015] The processing unit 12 outputs the corresponding image data and the answer sentence 3. For example, the processing unit 12 causes a terminal operated by the user to display the corresponding image data and the answer sentence 3. Also, for example, the processing unit 12 causes a terminal operated by the user to display the corresponding image data and output a sound indicating the answer sentence 3.

[0016] According to the first embodiment, storage unit 11 of information processing device 10 stores image information 11a indicating the correspondence between keywords for publicly managed racing games and image data. Processing unit 12 of information processing device 10 acquires question sentence 1 and extracts extracted keywords from answer sentence 3 output by generation model 2, which generates a response sentence to an input sentence, in response to question sentence 1. Processing unit 12 refers to image information 11a to identify corresponding image data that corresponds to the extracted keywords, and outputs the corresponding image data and answer sentence 3.

[0017] This allows the information processing device 10 to use image data to provide easy-to-understand information to users who do not understand the meaning of terms related to publicly managed racing. As a result, the information processing device 10 can improve the convenience of providing information about publicly managed racing.

[0018] Processing unit 12 may identify the user's characteristics from a user image captured of the user. Processing unit 12 may then refer to avatar information indicating the frequency of use of avatars of players in publicly managed racing games for each person's characteristics and the publicly managed racing event information, and determine the avatar with the highest frequency of use for the user's characteristics among the avatars of players participating in the race on that day as the avatar to be displayed when outputting the answer. This allows information processing device 10 to display an avatar that suits the user.

[0019] The processing unit 12 may also be configured to generate a response sentence by referring to data that combines the numerical data included in the publicly managed racing event information with natural language, thereby enabling the information processing device 10 to improve the accuracy of the response sentence.

[0020] Second Embodiment Next, we will explain the second embodiment, which uses generative AI to provide information to users of publicly managed racing games.

[0021] FIG. 2 is a diagram showing an example of an information processing system according to a second embodiment. The information processing system according to the second embodiment is installed in a public racecourse 20. The public racecourse 20 is a venue where public races are held. The information processing system according to the second embodiment has a terminal 21 and a server 100. The terminal 21 and the server 100 are connected to a network 30. The network 30 is, for example, a local area network (LAN) of the public racecourse 20.

[0022] Terminal 21 is a terminal that can be operated by users of public racing stadium 20. Terminal 21 has a display and a touch panel. Terminal 21 displays an operation screen on the display and accepts touch operations by the user on the operation screen via the touch panel. Terminal 21 also has a camera. Terminal 21 takes an image of the user operating it and transmits the image to server 100. Terminal 21 then displays an avatar on the operation screen that is determined by server 100 based on the image of the user.

[0023] The terminal 21 also has a microphone and a speaker. The terminal 21 acquires a question from a user through the microphone. The terminal 21 converts the acquired question into text and transmits it to the server 100. The terminal 21 then acquires a response to the question generated by the server 100, converts the acquired response into voice, and outputs it from the speaker.

[0024] Server 100 is a server computer that provides a service that answers questions using generative AI. When server 100 acquires an image of a user from terminal 21, it identifies the user's age and gender from the acquired image. Server 100 then displays on terminal 21 the avatar that is most frequently used by users of the identified age and gender from among the avatars of athletes participating in the competition that day.

[0025] When server 100 receives a question from terminal 21, it uses generative AI to generate a response to the question. Server 100 extracts keywords for the publicly managed racing game from the generated response and identifies image data corresponding to the extracted keywords. Server 100 then transmits the generated response and the identified image data to terminal 21, causing it to output the response and display the image data.

[0026] Server 100 may be installed outside public racetrack 20 and connected to terminal 21 via a network such as the Internet. Alternatively, server 100 may acquire questions from users inside public racetrack 20 via mobile terminals such as smartphones or tablet devices owned by users, instead of terminal 21, and transmit responses to the mobile terminals.

[0027] 3 is a diagram showing an example of a hardware configuration of a server. The entire server 100 is controlled by a processor 101. A memory 102 and multiple peripheral devices are connected to the processor 101 via a bus 109. The processor 101 may be a multiprocessor. The processor 101 is, for example, a central processing unit (CPU), a micro processing unit (MPU), or a digital signal processor (DSP). At least some of the functions realized by the processor 101 executing a program may be realized by an electronic circuit such as an application specific integrated circuit (ASIC) or a programmable logic device (PLD).

[0028] The memory 102 is used as a main storage device of the server 100. The memory 102 temporarily stores at least a part of the OS (Operating System) programs and application programs to be executed by the processor 101. The memory 102 also stores various data used in processing by the processor 101. As the memory 102, for example, a volatile semiconductor storage device such as a RAM (Random Access Memory) is used.

[0029] The peripheral devices connected to the bus 109 include a storage device 103, a GPU (Graphics Processing Unit) 104, an input interface 105, an optical drive device 106, a device connection interface 107, and a network interface 108.

[0030] The storage device 103 electrically or magnetically writes and reads data to and from a built-in recording medium. The storage device 103 is used as an auxiliary storage device for a computer. The storage device 103 stores an OS program, application programs, and various data. Note that the storage device 103 may be, for example, a hard disk drive (HDD) or a solid state drive (SSD).

[0031] The GPU 104 is connected to a monitor 31. The GPU 104 displays an image on the screen of the monitor 31 in accordance with an instruction from the processor 101. The monitor 31 may be a display device using an organic EL (Electro Luminescence) display device, a liquid crystal display device, or the like.

[0032] The input interface 105 is connected to a keyboard 32 and a mouse 33. The input interface 105 transmits signals sent from the keyboard 32 and the mouse 33 to the processor 101. The mouse 33 is an example of a pointing device, and other pointing devices can also be used. Examples of other pointing devices include a touch panel, a tablet, a touch pad, and a trackball.

[0033] The optical drive device 106 uses a laser beam or the like to read data recorded on the optical disc 34. The optical disc 34 is a portable recording medium on which data is recorded so that it can be read by reflected light. The optical disc 34 includes a DVD (Digital Versatile Disc), a DVD-RAM, a CD-ROM (Compact Disc Read Only Memory), a CD-R (Recordable) / RW (Rewritable), etc.

[0034] The device connection interface 107 is a communication interface for connecting peripheral devices to the server 100. For example, a memory device 35 or a memory reader / writer 36 can be connected to the device connection interface 107. The memory device 35 is a recording medium equipped with a function for communicating with the device connection interface 107. The memory reader / writer 36 is a device for writing data to the memory card 37 or reading data from the memory card 37. The memory card 37 is a card-type recording medium.

[0035] The network interface 108 is connected to the network 30. The network interface 108 transmits and receives data to and from other computers or communication devices via the network 30.

[0036] The server 100 can realize the processing functions of the second embodiment with the hardware configuration described above. The information processing device 10 shown in the first embodiment can also be realized with the same hardware as the server 100 shown in FIG. 3. The processor 101 is an example of the processing unit 12 shown in the first embodiment. The memory 102 or the storage device 103 is an example of the storage unit 11 shown in the first embodiment.

[0037] The server 100 realizes the processing functions of the second embodiment by executing a program recorded on, for example, a computer-readable recording medium. The program describing the processing to be executed by the server 100 can be recorded on various recording media. For example, the program to be executed by the server 100 can be stored in a storage device 103. The processor 101 loads at least a portion of the program in the storage device 103 into the memory 102 and executes the program. The program to be executed by the server 100 can also be recorded on a portable recording medium such as an optical disk 34, a memory device 35, or a memory card 37. The program stored on the portable recording medium becomes executable after being installed on the storage device 103 under the control of, for example, the processor 101. The processor 101 can also read and execute the program directly from the portable recording medium. Next, the functions of the server 100 will be described in detail.

[0038] 4 is a block diagram showing an example of the functions of the server. The server 100 has a storage unit 120, a terminal control unit 130, an avatar setting unit 140, an output control unit 150, and a generation unit 160. The storage unit 120 is realized using a storage area of ​​the memory 102 or the storage device 103. The terminal control unit 130, the avatar setting unit 140, the output control unit 150, and the generation unit 160 are realized by the processor 101 executing a program stored in the memory 102.

[0039] The storage unit 120 stores image information 121, avatar information 122, event information 123, and answer information 124. The image information 121 is information indicating the correspondence between keywords for publicly managed racing games and image data. The avatar information 122 is information indicating the frequency of avatar use by age and gender. The avatar information 122 registers avatars of players in publicly managed racing games. The event information 123 is information indicating the results of publicly managed racing games held at the publicly managed racetrack 20 and the schedule of publicly managed racing games to be held at the publicly managed racetrack 20. For example, the event information 123 includes the finishing order of public racing games held at the publicly managed racetrack 20 and the payout amounts for each betting type. The event information 123 also includes, for example, the players participating in the races held at the publicly managed racetrack 20 on that day. The answer information 124 registers data that combines the numerical data included in the event information 123 with natural language.

[0040] The terminal control unit 130 controls the terminal 21 and transmits and receives data to and from the terminal 21. The terminal control unit 130 acquires a question sentence from the terminal 21. For example, the terminal control unit 130 acquires a question sentence from the terminal 21, which is obtained by converting a voice input accepted by the terminal 21 into text data. The terminal control unit 130 causes the terminal 21 to display an answer sentence generated by the generation unit 160 in response to the question sentence and image data identified by the output control unit 150.

[0041] The avatar setting unit 140 selects an avatar to be displayed on the terminal 21. The avatar setting unit 140 acquires from the terminal 21 an image of a user operating the terminal 21. The avatar setting unit 140 identifies the age and gender of the user from the acquired image. For example, the avatar setting unit 140 inputs the acquired image into a trained model that outputs the age and gender of people appearing in the input image to identify the age and gender of the user. The avatar setting unit 140 refers to the avatar information 122 and the event information 123 and selects the avatar of an athlete who is most frequently used for the identified age and gender and who will be participating in the competition on the current date (today). The avatar setting unit 140 displays the selected avatar on the terminal 21.

[0042] After the terminal 21 displays the avatar, the avatar setting unit 140 acquires from the terminal 21 an image of the user operating the terminal 21. The avatar setting unit 140 determines whether the user has a friendly expression from the acquired image. For example, the avatar setting unit 140 inputs the acquired image into a trained model that outputs whether a person in the input image has a friendly expression, and determines whether the user has a friendly expression. If the avatar setting unit 140 determines that the user does not have a friendly expression, it selects the avatar of an athlete who is participating in the competition on that day and who has the next highest usage frequency, and causes the selected avatar to be displayed on the terminal 21.

[0043] The output control unit 150 specifies image data to be displayed along with the answer when a question about publicly managed racing terminology is received, and configures the generation unit 160 when a question about event information 123 is received. When a question about publicly managed racing terminology is received from a user, the output control unit 150 extracts keywords registered in the image information 121 from the answer text generated by the generation unit 160. The output control unit 150 then specifies image data registered in the image information 121 in association with the keywords in the answer text. When a question about event information 123 is received from a user, the output control unit 150 configures the generation unit 160 to refer to the answer information 124 when generating a sentence.

[0044] The generation unit 160 is a generative AI that generates a response sentence to an input sentence. For example, the generation unit 160 is a trained model that has been trained using conversational sentences on the Internet, etc., and further trained using question sentences about terminology and event information for publicly managed racing games, and answers to those questions. The generation unit 160 generates a response sentence to a question sentence acquired by the terminal control unit 130 from the terminal 21. Note that, when the output control unit 150 is set to refer to the answer information 124, the generation unit 160 searches the answer to the question sentence from the answer information 124 and generates the response sentence.

[0045] 4 indicate part of the communication paths, and communication paths other than those shown in the figure can also be set. Next, the information stored in storage unit 120 will be described in detail.

[0046] FIG. 5 is a diagram showing an example of image information. Image information 121 is information that shows the correspondence between keywords for publicly managed racing games and image data. Image information 121 has fields for keywords and image data. The keyword field contains a keyword for the publicly managed racing game. Note that multiple keywords may be set in the keyword field. The image data field contains image data that corresponds to the keyword.

[0047] 6 is a diagram showing an example of avatar information. Avatar information 122 is information indicating the frequency of use of avatars for each age and gender. Avatar information 122 includes the following items: age, gender, number of favorable comments, avatar, and speech speed.

[0048] The age field is set to the age of the user. The age field is set to an age range, such as the 60s (60 to 69 years old). The gender field is set to the gender of the user. The number of favorable comments field is set to the number of times the user made a favorable expression when the avatar was displayed (i.e., the number of times the user was used as an avatar). The avatar field is set to the avatar of a player in a public race held at the public race track 20. The voice speed field is set to the voice speed to be output when the avatar is displayed. The records of the avatar information 122 are sorted in descending order of the number of favorable comments field for each record with the same age and gender fields. The avatar information 122 may also have a volume field for setting the volume of the voice to be output when the avatar is displayed.

[0049] 7 is a diagram showing an example of the answer information. The answer information 124 is registered with data that combines the numerical data included in the event information 123 with natural language. The answer information 124 is registered with text data that explains the numerical data such as the order of finish and the payout amount, combining the numerical data such as the order of finish and the payout amount with Japanese.

[0050] For example, the answer information 124 has registered therein text data such as "The payout amount for the trifecta is 47,990 yen for 7-5-3." Of the text data, the part "The payout amount for the trifecta is" is set to be in Japanese format. The part "7-5-3" is set to be in numerical data format. The part "and" is set to be in Japanese format. The part "47,990" is set to be in numerical data format. The part "It is yen." is set to be in Japanese format.

[0051] For example, data of a fixed phrase in Japanese format combining "The payout amount for the trifecta is," "in," and "yen" is prepared in advance. Suppose that the results of a publicly managed game indicating that the trifecta combination is 7-5-3 and the payout amount is 47,990 yen is registered in the event information 123. Then, the numerical data "7-5-3" indicating the finishing order and the numerical data "47,990" indicating the payout amount are combined with the fixed phrase, and the text data "The payout amount for the trifecta is 7-5-3 and 47,990 yen" is registered in the answer information 124.

[0052] Next, we will explain the processing executed by the server 100. First, we will explain the response processing when a question about terminology in publicly managed racing is received from a user. 8 is a diagram showing an example of the process of responding to a question about terminology. Server 100 accepts a question about terminology for publicly managed racing from a user operating terminal 21, generates a response to the accepted question, and displays image data determined based on the response and keywords contained in the response on terminal 21.

[0053] First, the terminal 21 receives voice input from the user using a microphone. The terminal 21 analyzes the input voice and generates a question 51 by converting it into text. The question 51 is text data such as "Please tell me about the Keirin Grand Prix." The terminal 21 transmits the question 51 to the server 100. Alternatively, the terminal 21 may transmit the input voice to the server 100, and the server 100 may generate the question 51.

[0054] The generation unit 160 of the server 100 generates an answer sentence 52 that is a response to the question sentence 51. The answer sentence 52 is text data that reads, "The Keirin Grand Prix is ​​the biggest bicycle race of the year, where the best bicycle racers of the year gather to decide the champion of that year. The Grand Prix is ​​held every December,..." The output control unit 150 of the server 100 extracts keywords registered in the image information 121 from the answer sentence 52. In this example, the output control unit 150 extracts "keirin" and "grand prix" from the answer sentence 52.

[0055] The output control unit 150 identifies image data 53 corresponding to the keyword extracted from the answer sentence 52 from the image information 121. The image data 53 is image data indicating that the Grand Prix is ​​the highest rank in a bicycle race. The terminal control unit 130 of the server 100 transmits the answer sentence 52 and the image data 53 to the terminal 21, and causes the terminal 21 to display the answer sentence 52 and the image data 53. The terminal control unit 130 may also cause the terminal 21 to output the answer sentence 52 as sound from a microphone.

[0056] In this way, the output control unit 150 causes the terminal 21 to display the answer sentence 52 generated by the generation unit 160 as a response to the question sentence 51, and the image data 53 corresponding to the keywords extracted from the answer sentence 52. In this way, the output control unit 150 can provide information that is intuitively easy to understand for users who do not understand the meaning of terms related to publicly managed racing. Therefore, the output control unit 150 can improve the convenience when providing information about publicly managed racing. Next, the avatar selection process will be described.

[0057] 9 is a diagram showing an example of the avatar selection process. The server 100 identifies the age and gender of the user operating the terminal 21 from an image of the user, and causes the terminal 21 to display avatars that are frequently used by users of the identified age and gender.

[0058] First, the terminal 21 takes an image of the user operating the terminal 21 with a camera and generates a user image 61. The terminal 21 transmits the user image 61 to the server 100. The avatar setting unit 140 of the server 100 identifies the age and gender of the user from the user image 61. For example, the avatar setting unit 140 inputs the user image 61 into a trained model that outputs the age and gender of a person appearing in an input image, and identifies the age and gender of the user.

[0059] The avatar setting unit 140 refers to the avatar information 122 and the event information 123 and selects the avatar of the player who has the most favorable comments for the specified age and gender from among the avatars of the players participating in the competition that day. The terminal control unit 130 of the server 100 displays the selected avatar on the terminal 21. The terminal 21 displays a screen 62 including the avatar specified by the server 100.

[0060] When the terminal 21 displays the screen 62, it again takes a picture of the user operating the terminal 21 with the camera and generates a user image 63. The terminal 21 transmits the user image 63 to the server 100. The avatar setting unit 140 determines whether the displayed avatar is appropriate based on the user image 63. The avatar setting unit 140 determines whether the user has a friendly expression from the user image 63. For example, the avatar setting unit 140 inputs the user image 63 into a trained model that outputs whether a person appearing in an input image has a friendly expression, and determines whether the user has a friendly expression.

[0061] If the terminal control unit 130 determines that the user has a favorable expression, it displays the selected avatar on the terminal 21 and has the terminal 21 respond to the user. Then, the avatar setting unit 140 adds 1 to the number of favorable comments for the selected avatar in the avatar information 122. If the avatar setting unit 140 determines that the user does not have a favorable expression, it reselects the avatar of the player who has the next highest number of favorable comments for the user's age and gender from among the avatars of the players participating in the competition that day, and again determines the suitability of the avatar.

[0062] In this way, the avatar setting unit 140 selects the avatar of the player who has the most favorable comments for the user's age and gender from among the avatars of the players participating in the competition on that day, and displays the selected avatar on the terminal 21. In this way, the avatar setting unit 140 can display an avatar that is likely to be perceived as favorable by the user. Furthermore, after displaying the avatar on the terminal 21, the avatar setting unit 140 determines whether the user has a favorable expression, and if the expression is favorable, continues to display the avatar on the terminal 21, but if the expression is not favorable, reselects an avatar. In this way, the avatar setting unit 140 can display an appropriate avatar. Next, a response process when a question regarding the event information 123 is received from a user will be described.

[0063] 10 is a diagram showing an example of a process for responding to a question about event information. The server 100 receives a question about the event information 123 from a user operating the terminal 21, generates a response to the received question, and causes the generated response to be displayed on the terminal 21.

[0064] First, the terminal 21 receives voice input from the user using a microphone. The terminal 21 analyzes the input voice and generates a question 71 in text form. The question 71 is text data such as "Refund amount, tell me the refund amount, refund." The terminal 21 transmits the question 71 to the server 100.

[0065] When the output control unit 150 of the server 100 acquires the question sentence 51, it sets the generation unit 160 of the server 100 to refer to the answer information 124 when generating a sentence. The generation unit 160 then generates an answer sentence 72 that is a response to the question sentence 71. The answer sentence 72 is text data that reads, "Please select the name of the velodrome." The terminal control unit 130 of the server 100 transmits the answer sentence 72 to the terminal 21 and causes the terminal 21 to display the answer sentence 72. The terminal control unit 130 also causes the terminal 21 to display a screen listing the velodrome payouts.

[0066] Terminal 21 receives a question from the user in response to answer sentence 72 using a microphone, analyzes the input voice, and generates question sentence 73 as text. Question sentence 73 is text data such as "Beppu, Beppu Keirin, Beppu payout." Terminal 21 transmits question sentence 73 to server 100.

[0067] When the output control unit 150 acquires the question sentence 73, it sets the generation unit 160 to refer to the answer information 124 when generating a sentence. Then, the generation unit 160 generates an answer sentence 74 that is a response to the question sentence 73. The answer sentence 74 is text data that reads, "Please select a race number." The terminal control unit 130 transmits the answer sentence 74 to the terminal 21 and causes the terminal 21 to display the answer sentence 74. The terminal control unit 130 also causes the terminal 21 to display a payout screen for Beppu.

[0068] The terminal 21 receives a question from the user in response to the answer sentence 74 using a microphone, analyzes the input voice, and generates a text question sentence 75. The question sentence 75 is text data such as "2nd race, 2nd race, 2, 2nd race payout, Beppu 2nd race payout." The terminal 21 transmits the question sentence 75 to the server 100.

[0069] When the output control unit 150 acquires the question 75, it sets the generation unit 160 to refer to the answer information 124 when generating a sentence. Then, the generation unit 160 generates an answer 76 that is a response to the question 75. The answer 76 is text data that reads, "Here are the payouts for the second race of the Beppu Keirin. The trifecta is 12,200 yen for 4-3-6, and the two-horse bet is 6,800 yen for 4-3. The rest is as shown here." The terminal control unit 130 transmits the answer 76 to the terminal 21 and causes the terminal 21 to display the answer 76. The terminal control unit 130 also causes the terminal 21 to display a Beppu second race payout screen.

[0070] In this way, when the output control unit 150 receives a question about the event information 123 from a user, it sets the generation unit 160 to refer to the answer information 124, which is registered text data that combines the numerical data included in the event information 123 with natural language. Then, the generation unit 160 generates an answer to the user's question. This allows the output control unit 150 to improve the accuracy of the answer. Next, a procedure for user response processing by the server 100 will be described.

[0071] 11 is a flowchart showing an example of the procedure for user support processing. The processing shown in FIG. 11 will be explained below in order of step number. [Step S11] The avatar setting unit 140 executes an avatar selection process, which will be described in detail later (see FIG. 12).

[0072] [Step S12] The terminal control unit 130 determines whether to accept a question about a term. For example, when the terminal control unit 130 is notified by the terminal 21 that a selection operation to ask a question about a term has been accepted, the terminal control unit 130 determines that a question about a term will be accepted. Furthermore, when the terminal control unit 130 is notified by the terminal 21 that a selection operation to ask a question about the event information 123 has been accepted, the terminal control unit 130 determines that a question about a term will not be accepted (a question about the event information 123 will be accepted). If the terminal control unit 130 determines that a question about a term will be accepted, the terminal control unit 130 proceeds to step S13. If the terminal control unit 130 determines that a question about the event information 123 will be accepted, the terminal control unit 130 proceeds to step S20.

[0073] [Step S13] The terminal control unit 130 acquires a question from the terminal 21. For example, the terminal control unit 130 acquires from the terminal 21 a question obtained by converting speech input accepted by the terminal 21 into text data.

[0074] [Step S14] The generation unit 160 generates an answer to the question acquired in step S13. [Step S15] The output control unit 150 extracts keywords registered in the image information 121 from the reply text generated in step S14.

[0075] [Step S16] The output control unit 150 determines whether the answer sentence generated in step S14 contains a keyword registered in the image information 121. If the output control unit 150 determines that the answer sentence contains a keyword, it proceeds to step S18. If the output control unit 150 determines that the answer sentence does not contain a keyword, it proceeds to step S17.

[0076] [Step S17] The terminal control unit 130 causes the reply message generated in step S14 to be displayed on the terminal 21. Then, the process ends. [Step S18] The output control unit 150 identifies image data corresponding to the keywords in the answer sentence generated in step S 14. For example, the output control unit 150 identifies image data registered in the image information 121 in association with the keywords in the answer sentence.

[0077] [Step S19] The terminal control unit 130 displays the reply message generated in step S14 and the image data identified in step S18 on the terminal 21. Then, the process ends.

[0078] [Step S20] The terminal control unit 130 acquires a question from the terminal 21. For example, the terminal control unit 130 acquires from the terminal 21 a question obtained by converting speech input accepted by the terminal 21 into text data.

[0079] [Step S21] The output control unit 150 sets the generation unit 160 to refer to the answer information 124 when generating a sentence. [Step S22] The generation unit 160 generates an answer to the question acquired in step S20. For example, the generation unit 160 searches the answer information 124 for an answer to the question acquired in step S20 and generates an answer.

[0080] [Step S23] The terminal control unit 130 causes the terminal 21 to display the reply message generated in step S22. [Step S24] The terminal control unit 130 determines whether to end the response. For example, when the terminal control unit 130 is notified by the terminal 21 that a selection operation to end use of the terminal 21 has been accepted, the terminal control unit 130 determines to end the response. If the terminal control unit 130 determines to end the response, it ends the processing. On the other hand, if the terminal control unit 130 determines not to end the response, it proceeds to step S20.

[0081] In this way, the output control unit 150 extracts keywords from the answer text output by the generation unit 160 in response to the question text about the terminology, and identifies the image data corresponding to the keywords by referring to the image information 121 that shows the correspondence between publicly managed racing keywords and image data. The terminal control unit 130 then displays the image data and the answer text on the terminal 21. This allows the server 100 to use image data to provide easy-to-understand information to users who do not understand the meaning of publicly managed racing terms. Therefore, the server 100 can improve the convenience of providing information about publicly managed racing.

[0082] Furthermore, when the output control unit 150 receives a question about the event information 123, the output control unit 150 sets the generation unit 160 to generate a response sentence by referencing answer information 124, which is a combination of data included in the publicly managed racing event information 123 and natural language. This enables the server 100 to improve the accuracy of responses.

[0083] 12 is a flowchart showing an example of the procedure for the avatar selection process. The process shown in FIG. 12 will be described below in order of step number. [Step S31] The avatar setting unit 140 acquires from the terminal 21 an image of the user operating the terminal 21.

[0084] [Step S32] The avatar setting unit 140 identifies the age and gender of the user from the image acquired in step S31. For example, the avatar setting unit 140 inputs the image acquired in step S31 into a trained model that outputs the age and gender of the person appearing in the input image, thereby identifying the age and gender of the user.

[0085] [Step S33] The avatar setting unit 140 refers to the avatar information 122 and selects the avatar of the player with the most favorable comments for the age and gender identified in step S32. For example, the avatar setting unit 140 extracts records of avatar information 122 in which the age field is set to a range including the age identified in step S32 and the gender field is set to the gender identified in step S32. Then, the avatar setting unit 140 selects the combination of the avatar and voice speed of the player set in the record with the most favorable comments from the extracted records.

[0086] [Step S34] The avatar setting unit 140 refers to the event information 123 and determines whether the athlete whose avatar was selected in step S33 or step S35 will participate in the competition on the current date (today). If the avatar setting unit 140 determines that the athlete whose avatar was selected will participate in the competition on that day, it proceeds to step S36. If the avatar setting unit 140 determines that the athlete whose avatar was selected will not participate in the competition on that day, it proceeds to step S35.

[0087] [Step S35] The avatar setting unit 140 refers to the avatar information 122 and selects the avatar of the player who has the next highest number of favorable comments for the age and gender identified in step S32, after the previously selected player. For example, the avatar setting unit 140 selects the combination of the avatar and voice speed of the player set in the record extracted in step S33 that has the next highest number of favorable comments after the previously selected record.

[0088] [Step S36] The avatar setting unit 140 displays the avatar last selected in step S33 or step S35 on the terminal 21. The avatar setting unit 140 also sets the speech speed of the terminal 21 to the speech speed last selected in step S33 or step S35.

[0089] [Step S37] The avatar setting unit 140 acquires from the terminal 21 an image of the user operating the terminal 21. [Step S38] The avatar setting unit 140 determines whether the user has a friendly expression from the image acquired in step S37. For example, the avatar setting unit 140 inputs the image acquired in step S37 into a trained model that outputs whether a person in the input image has a friendly expression, and determines whether the user has a friendly expression. If the avatar setting unit 140 determines that the user has a friendly expression, it proceeds to step S39. On the other hand, if the avatar setting unit 140 determines that the user does not have a friendly expression, it proceeds to step S35.

[0090] [Step S39] The avatar setting unit 140 adds 1 to the number of favorable comments corresponding to the displayed avatar. For example, the avatar setting unit 140 adds 1 to the number of favorable comments field of the record last selected in step S33 or step S35.

[0091] In this way, the avatar setting unit 140 identifies the age and gender of the user from an image captured of the user of the terminal 21. Then, the avatar setting unit 140 refers to the avatar information 122, which indicates the order of frequency of use of avatars by age and gender, and the event information 123, to identify the avatar most frequently used for the age and gender of the user among the avatars of the athletes participating in the competition on that day. The avatar setting unit 140 displays the identified avatar on the terminal 21. In this way, the avatar setting unit 140 can display an avatar appropriate for the user.

[0092] Although the embodiments have been described above, the configuration of each part shown in the embodiments can be replaced with other parts having similar functions. Also, any other components or processes may be added. Furthermore, any two or more configurations (features) of the above-described embodiments may be combined. [Explanation of symbols]

[0093] 1. Question 2. Generative Model 3 Answer text 10. Information processing equipment 11 Storage section 11a Image Information 12 Processing section

Claims

1. a storage unit for storing image information indicating the correspondence between keywords for publicly managed gaming events and image data; a processing unit that acquires a question sentence, extracts extracted keywords from a response sentence output by a generation model that generates a response sentence to an input sentence, identifies corresponding image data that corresponds to the extracted keywords by referring to the image information, and outputs the corresponding image data and the response sentence; An information processing device having the above.

2. the processing unit identifies the user's characteristics from a user image taken of the user, and refers to avatar information indicating the order of frequency of use of avatars of players in the publicly managed racing game for each person's characteristics and to information about the running of the publicly managed racing game, and determines the avatar with the highest frequency of use for the user's characteristics from among the avatars of players participating in the race on that day as the avatar to be displayed when the answer sentence is output; 2. The information processing device according to claim 1.

3. the processing unit configures the generative model to generate the response sentence by referencing data that combines numerical data included in the information about the publicly managed racing event and natural language; 2. The information processing device according to claim 1.

4. The computer Get the question text, extracting extracted keywords from a response sentence output by a generative model that generates a response sentence to an input sentence, and Identifying image data corresponding to the extracted keywords by referring to image information indicating the correspondence between keywords and image data for publicly managed racing games; outputting the corresponding image data and the answer sentence; Information processing methods.

5. On the computer, Get the question text, extracting extracted keywords from a response sentence output by a generative model that generates a response sentence to an input sentence, and Identifying image data corresponding to the extracted keywords by referring to image information indicating the correspondence between keywords and image data for publicly managed racing games; outputting the corresponding image data and the answer sentence; A program that executes a process.

Images