Artificial intelligence system and program of artificial intelligence system

Abstract

To provide an artificial intelligence system which allows a user to select pre-trained artificial intelligence means or a database.SOLUTION: An artificial intelligence system 100 includes a user terminal 110 and servers 120, 130. A browser 111 of the user terminal displays a talk screen (112) that allows a user to communicate with pre-trained artificial intelligence means 131 of the server 130 and that includes a button 115a to select the pre-trained artificial intelligence means and a button 115b to select a database 121. In response to a selection operation on the button to select the pre-trained artificial intelligence means, and on receipt of a text input as question information, text data generation is executed using the pre-trained artificial intelligence means, and the browser displays generated text data obtained by the execution on the talk screen. In response to a selection operation on the button to select the database, the browser displays the question information and answer information based on the database.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to an artificial intelligence system that includes a user terminal, a server, and an administrator terminal, in which the user terminal displays or outputs audio based on the server, and a program for the system. [Background technology]

[0002] Conventionally, there is known an information processing device that includes a language data receiving unit that receives language data in response to a request from a user terminal, a language data processing unit that processes the language data to generate basic summary information that forms the basis for generating a summary, a transmitting unit that transmits presentation information including the basic summary information to the user terminal, an editing information receiving unit that receives editing information related to the presentation information from the user terminal, and a summary generation unit that provides the editing information to a large-scale language model that receives input of the basic summary information and generates a summary sentence that reflects the editing information (for example, Patent Document 1). [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Patent Publication No. 2023-73095 Summary of the Invention [Problem to be solved by the invention]

[0004] However, the conventional information processing devices described above were limited to generating summaries of basic summary information using large-scale language models, so even if a user asks a question and receives an answer, such as in a so-called FAQ or Q&A, the answer may not be 100% correct, and there was a problem that the devices were not user-friendly enough when a 100% correct answer was desired.

[0005] Therefore, the present invention solves the problems of the prior art as described above, that is, the object of the present invention is to provide an artificial intelligence system and a program thereof that allow a user to select, depending on the situation, either a pre-trained artificial intelligence means that provides flexible answers that may not be 100% correct, or a database that is not flexible and has fixed categories or answers that are 100% correct. [Means for solving the problem]

[0006] The invention according to claim 1 is an artificial intelligence system comprising a user terminal, a server and an administrator terminal, the user terminal displaying or outputting audio based on the server, the reference range of the pre-trained artificial intelligence means being preset based on at least one of one or more pieces of URL information, site map information, information for distribution, document information and folder designation information that are preset by the administrator terminal, category information, question information and answer information in the server's database being preset by the administrator terminal, the browser of the user terminal being able to freely communicate with the pre-trained artificial intelligence means of the server, displaying a talk screen having a button for selecting the pre-trained artificial intelligence means and a button for selecting the database, and the talk screen In this system, when a button for selecting a pre-trained artificial intelligence means is selected and text is input, the user terminal transmits the text data to a server as question information, the server uses the pre-trained artificial intelligence means to generate text data based on the question information within the reference range, and transmits the generated text data obtained by this execution to the user terminal, the user terminal displays the received generated text data on the talk screen of the browser, and when a button for selecting a database is selected on the talk screen, the browser of the user terminal displays category information, or question information and answer information, which have been preset based on the database, thereby solving the above-mentioned problem.

[0007] The invention of claim 2 further solves the above-mentioned problem by, in addition to the configuration of the artificial intelligence system described in claim 1, the talk screen further has a button for chatting with an operator, and when the button for chatting with an operator is selected on the talk screen, the talk screen of the user terminal displays message data between the user terminal and the operator terminal.

[0008] The invention of claim 3 further solves the above-mentioned problem by configuring the artificial intelligence system as described in claim 2 such that, in addition to the configuration of the artificial intelligence system, the talk screen further has a button for speaking with an operator, and when the button for speaking with an operator is selected on the talk screen, the microphone of the user terminal becomes active and the user terminal makes a call to an operator terminal, and when a call operation is performed on the operator terminal, the user terminal and the operator terminal enter a call state.

[0009] The invention of claim 4 further solves the above-mentioned problem by, in addition to the configuration of the artificial intelligence system described in any one of claims 1 to 3, the talk screen has a talk display area that displays question information and generated text data which are an exchange between a user and a pre-trained artificial intelligence means, and an item display area that displays items of the exchange, the reference range of the pre-trained artificial intelligence means is set in advance for each item, and when an operation to select or change an item in the item display area is performed, the browser displays the exchange corresponding to the item in the talk display area.

[0010] The invention of claim 5 further solves the above-mentioned problem by configuring the artificial intelligence system as described in claim 4, in which the items are the names of departments in an organization, and the reference range set for each of the items is handling information by the department of the organization corresponding to the item name set based on at least one of URL information, site map information, distribution information, document information, and folder designation information.

[0011] The invention of claim 6 further solves the above-mentioned problem by configuring the artificial intelligence system as described in claim 5, and further configuring the server to use the pre-trained artificial intelligence means to generate text data based on question information within a reference range, and when the server transmits the generated text data obtained by the execution to a user terminal, the server registers the question information and the generated text data as a set in a database as question information and answer information.

[0012] The invention of claim 7 is a program for an artificial intelligence system in which a user terminal displays or outputs audio based on a server, comprising: a reference range setting step in which a reference range of a pre-trained artificial intelligence means is set based on at least one of one or more URL information, site map information, distribution information, document information, and folder designation information specified by an administrator terminal; a database setting step in which category information, question information, and answer information in the server's database are set by the administrator terminal; a talk screen display step in which the user terminal accesses the server, the browser of the user terminal is capable of interacting with the server's pre-trained artificial intelligence means, and the talk screen has a button for selecting a pre-trained artificial intelligence means and a button for selecting a database; and a button for selecting a pre-trained artificial intelligence means or a button for selecting a database on the talk screen. The above-mentioned problem is solved by having a selection operation presence / absence determination step for determining whether a selection operation has occurred, a prompt transmission step in which, when a button to select the pre-trained artificial intelligence means is selected and text is input, the user terminal transmits text data to a server as question information, a text generation and transmission step in which the server uses the pre-trained artificial intelligence means to generate text data based on question information within a reference range and transmits the generated text data obtained by this execution to the user terminal, a generated text display step in which the user terminal displays the received generated text data on the talk screen of the browser, and a database content display step in which, when a button to select the database is selected on the talk screen, the browser of the user terminal displays category information, or question information and answer information that has been preset based on the database. Effect of the Invention

[0013] The artificial intelligence system of the present invention is equipped with a user terminal and a server, which not only enables the user terminal to display or output audio based on the server, but also provides the following unique effects.

[0014] According to the artificial intelligence system of the invention of claim 1, a button for selecting a pre-trained artificial intelligence means and a button for selecting a database are displayed on the talk screen, so that the user can select, depending on the situation, either the pre-trained artificial intelligence means which provides flexible answers that may not be 100% correct, or the database which is not flexible but has fixed categories and answers that are 100% correct. In other words, the appropriate one can be selected depending on the situation, and an answer appropriate to the situation can be obtained. Additionally, you can switch between the two, compare the answers, and adopt the more appropriate answer.

[0015] According to the artificial intelligence system of the invention of claim 2, in addition to the effects achieved by the invention of claim 1, the option to chat with an operator is increased by the button to chat with an operator, so that when a user cannot obtain the intended answer even by using a pre-trained artificial intelligence means or a database, the user can chat with an operator to obtain the desired answer.

[0016] According to the artificial intelligence system of the invention of claim 3, in addition to the effects achieved by the invention of claim 2, the option of speaking with an operator is increased by the button to speak to an operator, so that when the user cannot obtain the intended answer even by using the pre-trained artificial intelligence means or database, the user can speak to an operator and obtain the desired answer.

[0017] According to the artificial intelligence system of the invention of claim 4, in addition to the effects achieved by the invention of any one of claims 1 to 3, the reference range is changed by selecting an item through the item selection and change operation, so that the user can select the desired reference range and interact with the pre-trained artificial intelligence means.

[0018] According to the artificial intelligence system of the invention of claim 5, in addition to the effect of the invention of claim 4, when an item indicating the name of a department in an organization is selected, the reference range becomes the information handled by that department, so that the user can select an item indicating the name of the department responsible for the question and interact with the pre-trained artificial intelligence means to obtain the desired answer. In other words, by selecting the correct department in charge of the question, the risk of a misguided answer from the pre-trained artificial intelligence means can be avoided and the desired answer with a high accuracy rate can be obtained from the pre-trained artificial intelligence means.

[0019] According to the artificial intelligence system of the invention of claim 6, in addition to the effect of the invention of claim 5, the generated answers are sequentially added to the database in sets with the questions, making it possible to create an FAQ database using pre-trained artificial intelligence means.

[0020] According to the program of the artificial intelligence system of the invention of claim 7, similar to the effect achieved by the invention of claim 1, a button for selecting a pre-trained artificial intelligence means and a button for selecting a database are displayed on the talk screen, so that the user can select either a pre-trained artificial intelligence means that provides flexible answers that may not be 100% correct, or a database that is not flexible but has fixed categories or answers that are 100% correct, depending on the situation. In other words, the appropriate one can be selected depending on the situation, and an answer appropriate to the situation can be obtained. Additionally, you can switch between the two, compare the answers, and adopt the more appropriate answer. [Brief description of the drawings]

[0021] [Figure 1] FIG. 1 is a diagram showing the concept of an artificial intelligence system according to an embodiment of the present invention. [Diagram 2] FIG. 2 is a chart showing an example of the operation of the artificial intelligence system according to the embodiment of the present invention. [Diagram 3] FIG. 13 is a diagram showing how the reference range is set on the administrator screen of the administrator terminal of the artificial intelligence system according to an embodiment of the present invention. [Figure 4] FIG. 13 is a diagram showing the database settings on the administrator screen of the administrator terminal of the artificial intelligence system according to an embodiment of the present invention. [Diagram 5] FIG. 2 is a diagram showing an example of a FAQ-AI screen as a talk screen of a user terminal of an artificial intelligence system according to an embodiment of the present invention. [Figure 6] A figure showing an example of interaction with a pre-trained artificial intelligence means when a button to select a pre-trained artificial intelligence means is selected on the FAQ-AI screen of a user terminal of an artificial intelligence system which is an embodiment of the present invention. [Figure 7] A figure showing the appearance of a browser displaying items from an FAQ database when a button for selecting a database is selected on the FAQ-AI screen of a user terminal of an artificial intelligence system which is an embodiment of the present invention. [Figure 8] A figure showing an example of an interaction with an operator on the FAQ-AI screen when a button to chat with an operator is selected on the FAQ-AI screen of a user terminal of an artificial intelligence system which is an embodiment of the present invention. [Figure 9] A figure showing the appearance of a conversation with an operator in which an operator conversation window is displayed when the button to talk to an operator is selected on the FAQ-AI screen of a user terminal of an artificial intelligence system which is an embodiment of the present invention. [Figure 10] A figure showing an example in which a server of an artificial intelligence system according to an embodiment of the present invention additionally registers a set of question information from a user and answer information from a pre-trained artificial intelligence means in an FAQ database. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0022] The artificial intelligence system of the present invention comprises a user terminal, a server, and an administrator terminal, and the reference range of the pre-trained artificial intelligence means is set in advance based on at least one of one or more URL information, site map information, distribution information, document information, and folder designation information that are specified in advance by the administrator terminal, category information, question information, and answer information in the server's database are set in advance by the administrator terminal, the browser of the user terminal is capable of freely interacting with the pre-trained artificial intelligence means of the server, and displays a talk screen having a button for selecting the pre-trained artificial intelligence means and a button for selecting the database, and when the button for selecting the pre-trained artificial intelligence means is selected on the talk screen and text is input, the user terminal transmits the text data to the server as question information. The specific implementation may be any as long as the server uses a pre-trained artificial intelligence means to generate text data based on question information within a reference range, and transmits the generated text data obtained by the execution to a user terminal, and the user terminal displays the received generated text data on a talk screen of a browser, and when a button to select a database is selected on the talk screen, the browser of the user terminal displays category information, or question information and answer information that has been preset based on the database, so that the user can select, depending on the situation, either a pre-trained artificial intelligence means that gives flexible answers that may not be 100% correct, or a database that is 100% correct with fixed categories or answers that are not flexible. In addition, the program of the artificial intelligence system of the present invention includes a reference range setting step in which the reference range of the pre-trained artificial intelligence means is set based on at least one of one or more pieces of URL information, sitemap information, distribution information, document information, and folder designation information specified by the administrator terminal; a database setting step in which category information, question information, and answer information in the server's database are set by the administrator terminal; a talk screen display step in which the user terminal accesses the server, and the browser of the user terminal can communicate with the pre-trained artificial intelligence means of the server, and a talk screen having buttons for selecting the pre-trained artificial intelligence means and a button for selecting the database is displayed; a selection operation presence / absence determination step for determining the presence or absence of a selection operation of the button for selecting the pre-trained artificial intelligence means or the button for selecting the database on the talk screen; a prompt transmission step in which when there is a selection operation of the button for selecting the pre-trained artificial intelligence means and there is text input, the user terminal transmits text data to the server as question information; a text generation and transmission step in which the server executes text data generation based on the question information within the reference range using the pre-trained artificial intelligence means and transmits the generated text data obtained by the execution to the user terminal; a generated text display step in which the user terminal displays the received generated text data on the talk screen of the browser; and a database content display step in which when there is a selection operation of the button for selecting the database on the talk screen, the browser of the user terminal displays the category information preset based on the database, or the question information and the answer information. Thus, the user can select either the pre-trained artificial intelligence means that gives a flexible answer, although it may not be 100% correct, or the database that has a fixed category or answer and is 100% correct according to the situation. Any specific implementation manner may be adopted as long as it can be selected according to the situation.

[0023] For example, the user terminal may be a desktop personal computer terminal, a notebook personal computer terminal, a smartphone terminal, a tablet terminal, or the like that sends and receives information, and may be any terminal that can be freely connected to a server via a communication network including a wide area network, such as the Internet, a local network, or a telephone line. Also, the server may be a single server or multiple servers on a cloud. The pre-trained artificial intelligence means may be composed of an interactive model, also known as a large-scale language model, such as ChatGPT (Generative Pre-trained Transformer) (hereinafter referred to as ChatGPT), and may be configured to be installed on one server or multiple servers on the cloud. EXAMPLES

[0024] An artificial intelligence system 100 according to an embodiment of the present invention will now be described with reference to FIGS. Here, FIG. 1 is a diagram showing the concept of an artificial intelligence system 100 which is an embodiment of the present invention, FIG. 2 is a chart diagram showing an example of the operation of the artificial intelligence system 100 which is an embodiment of the present invention, FIG. 3 is a diagram showing the setting of the reference range RG on the administrator screen 142 of the administrator terminal 140 of the artificial intelligence system 100 which is an embodiment of the present invention, FIG. 4 is a diagram showing the setting of the database 121 on the administrator screen 142 of the administrator terminal 140 of the artificial intelligence system 100 which is an embodiment of the present invention, FIG. 5 is a diagram showing an example of a FAQ-AI screen 112 as a talk screen of the user terminal 110 of the artificial intelligence system 100 which is an embodiment of the present invention, FIG. 6 is a diagram showing an example of an interaction with chat GPT which is an example of the pre-trained artificial intelligence means 131 when a button 115a for selecting chat GPT which is an example of the pre-trained artificial intelligence means 131 is selected on the FAQ-AI screen 112 of the user terminal 110 of the artificial intelligence system 100 which is an embodiment of the present invention, and FIG. 7 is a diagram showing an example of an interaction with chat GPT which is an example of the pre-trained artificial intelligence means 131 when a button 115a for selecting chat GPT which is an example of the pre-trained artificial intelligence means 131 is selected on the FAQ-AI screen 112 of the user terminal 110 of the artificial intelligence system 100 which is an embodiment of the present invention, and FIG. FIG. 8 is a diagram showing an example of an interaction with an operator on the FAQ-AI screen 112 when button 115b for selecting database 121 is selected on the FAQ-AI screen 112 of the user terminal 110 of the artificial intelligence system 100 which is an embodiment of the present invention is selected, and FIG. 9 is a diagram showing an operator call window that is displayed and a conversation with an operator that is being held when button 115d for talking to an operator is selected on the FAQ-AI screen 112 of the user terminal 110 of the artificial intelligence system 100 which is an embodiment of the present invention is selected, and FIG. 10 is a diagram showing an example in which the first server 120 of the artificial intelligence system 100 which is an embodiment of the present invention additionally registers a set of question information QT from a user and answer information AN of a chat GPT, which is an example of a pre-trained artificial intelligence means 131, in the FAQ database.

[0025] An artificial intelligence system 100 according to an embodiment of the present invention includes a user terminal 110, a first server 120 and a second server 130 which are examples of servers, and an administrator terminal 140, as shown in FIG. Of these, the first server 120 has a database 121 including an FAQ database, and a bot 122 which is also a display program. On the other hand, the second server 130 has a chat GPT, which is an example of a pre-trained artificial intelligence means 131. As a technical concept, the first server 120 and the second server 130 may have a common physical configuration, or may have physically different configurations. And, the user terminal 110 is provided to display or output audio based on the first server 120 .

[0026] Specifically, first, the reference range RG of the chat GPT, which is an example of the pre-trained artificial intelligence means 131, is pre-set based on at least one of one or more URL information, site map information, distribution information, document information, and folder specification information that are pre-specified by the administrator terminal 140. Moreover, the category information CT, the question information QT, and the answer information AN in the database 121 of the first server 120 are preset by the administrator terminal 140. The browser 111 of the user terminal 110 can freely communicate with the chat GPT, which is an example of the pre-trained artificial intelligence means 131 of the second server 130, via the first server 120.

[0027] Furthermore, when the user terminal 110 accesses the first server 120, the browser 111 displays a FAQ-AI screen 112 as a talk screen having a button 115a for selecting chat GPT, which is an example of a pre-trained artificial intelligence means 131, and a button 115b for selecting the database 121. On the FAQ-AI screen 112, when a button 115a is selected to select chat GPT, which is an example of a pre-trained artificial intelligence means 131, and text is input, the user terminal 110 transmits the text data to the first server 120 as question information PT, which is an example of instruction command information also known as a prompt. Then, the first server 120 uses the chat GPT, which is an example of the pre-trained artificial intelligence means 131 of the second server 130, to generate text data based on the question information PT within the reference range RG.

[0028] Then, the first server 120 transmits the generated text data GT obtained through the execution to the user terminal 110. The user terminal 110 then displays the received generated text data GT on the FAQ-AI screen 112 of the browser 111. On the other hand, when button 115b for selecting database 121 is selected on FAQ-AI screen 112, browser 111 of user terminal 110 is configured to display category information CT, or question information QT and answer information AN, which are preset based on database 121 of first server 120.

[0029] As a result, a button 115a for selecting chat GPT, which is an example of the pre-trained artificial intelligence means 131, and a button 115b for selecting the database 121 are displayed on the FAQ-AI screen 112. As a result, the user can choose between Chat GPT, an example of a pre-trained artificial intelligence means 131 that provides flexible answers that may not be 100% correct, or the database 121, which is not flexible but has fixed categories and answers that are 100% correct, depending on the situation. In other words, the appropriate one can be selected depending on the situation, and an answer appropriate to the situation can be obtained. Additionally, you can switch between the two, compare the answers, and adopt the more appropriate answer.

[0030] Next, an example of the operation of the artificial intelligence system 100 will be described in detail. As shown in FIG. 2, in step S1, the user terminal 110 accesses the server as a talk screen display step. More specifically, when the user terminal 110 accesses the first server 120 and makes an HTTP request, the first server 120 sends data of the FAQ-AI screen 112 (see Figure 5) as a talk screen to the user terminal 110 as an HTTP response. Then, the browser 111 of the user terminal 110 displays the FAQ-AI screen 112.

[0031] Here, the FAQ-AI screen 112 can freely interact with chat GPT, which is an example of pre-trained artificial intelligence means 131 of the second server 130, via the first server 120, and has an input field 114a, a button 115a for selecting chat GPT, which is an example of pre-trained artificial intelligence means 131, a button 115b for selecting database 121, a button 115c for chatting with an operator, and a button 115d for talking to an operator. As a prerequisite, prior to step S1 (pre-setting), there are a reference range setting step and a database setting step. In the reference range setting step, the reference range RG of the chat GPT, which is an example of the pre-trained artificial intelligence means 131, is set based on at least one of one or more URL information, site map information, distribution information, document information, and folder specification information specified by the administrator terminal 140. More specifically, as shown in FIG. 3, when an administrator logs in to the artificial intelligence system 100 as the administrator using an administrator terminal 140, a browser 141 of the administrator terminal 140 displays an administrator screen 142.

[0032] The administrator screen 142 allows the items 113a of the FAQ-AI screen 112 to be freely set, and also allows the reference range RG of the chat GPT, which is an example of the pre-trained artificial intelligence means 131 corresponding to each item 113a, to be freely set. As for the item 113a, the item name is freely settable. Examples of item names include "expenses related," "vacation application," "company regulations," "telework related," "video conferencing related," "general affairs department," "human resources department," "sales department," "public relations department," and "legal department."

[0033] In addition, the reference range RG is set up to specify an arbitrary folder in a file server as an example of at least one of one or more URL information, site map information, distribution information, document information, and folder designation information, thereby identifying the reference range RG. In other words, the chat GPT, which is an example of the pre-trained artificial intelligence means 131, reads and references the text of the contents of a document file in a specified folder.

[0034] For example, in the “Expenses-related” item 113a, the chat GPT, which is an example of a pre-trained artificial intelligence means 131, reads and references the text of the contents of a document file in the “Accounting-related folder” folder in the “Accounting department” folder (lower hierarchy). Furthermore, in the database setting step, the category information CT, the question information QT, and the answer information AN in the database 121 of the first server 120 are set by the administrator terminal 140. More specifically, as shown in FIG. 4, the administrator screen 142 is provided so that question information QT, answer information AN, and category information CT of the FAQ database 121 can be freely set.

[0035] On the premise of these settings, in step S1, as shown in FIG. 5, the browser 111 of the user terminal 110 displays the FAQ-AI screen 112 as a talk screen based on the HTTP response. Here, the FAQ-AI screen 112 has, as an example, a talk display area 114 that displays question information PT and generated text data GT, which are an exchange between the user and a chat GPT, which is an example of a pre-trained artificial intelligence means 131, and an item display area 113 that displays items 113a of the exchange. Here, based on the information of item 113a set on administrator screen 142, examples of items 113a include "expenses related," "vacation application," "company regulations," "telework related," "video conferencing related," "general affairs department," "personnel department," "sales department," "public relations department," and "legal department."

[0036] In step S2, an item selection determination step, the browser 111 determines whether any of the items 113a is selected. For example, assume that the user selects the “expense-related” item 113a in the item display area 113 on the FAQ-AI screen 112. In this way, if it is determined that any of the items 113a such as "expense related" has been selected, the process proceeds to step S3, whereas if it is determined that no such item has been selected yet, step S2 is repeated. It should be noted that the item 113a displayed at the top may be selected by default.

[0037] In step S3, as a selection operation determination step, the browser 111 determines whether or not a selection operation has been performed on the button 115a on the FAQ-AI screen 112 to select chat GPT, which is an example of the pre-trained artificial intelligence means 131. If it is determined that an operation has been performed, the process proceeds to step S4. On the other hand, if it is determined that an operation has not been performed, the process proceeds to step S8.

[0038] In step S4, as a text input / transmission operation presence / absence determination step, the browser 111 determines whether or not text such as a question has been input in the input field 114a and whether or not a transmission operation has been performed. For example, as shown in FIG. 6, the user inputs "How do I settle my travel expenses?" as an example of a question as question information PT, and operates the send button. In this way, if it is determined that text has been input and a transmission operation has been performed, the process proceeds to step S5, whereas if it is determined that no text has been input and a transmission operation has been performed, step S4 is repeated. As for the method of inputting text data such as question information PT on FAQ-AI screen 112 of browser 111 of user terminal 110, the text data may be input using a keyboard, or the microphone of user terminal 110 may be activated by operating a microphone icon provided in input field 114a of FAQ-AI screen 112, and the voice data input by voice may be converted into text data by browser 111 and input.

[0039] In addition, when item 113a is selected and button 115a for selecting chat GPT, which is an example of pre-trained artificial intelligence means 131, is selected, as an example, chat GPT, which is an example of pre-trained artificial intelligence means 131, generates multiple pieces of question information QT based on the reference range RG, in which information regarding the content information of the reference range RG becomes answer information AN. Then, the first server 120 transmits the generated questions GQ to the user terminal 110. Then, as shown in FIG. 5, a plurality of pieces of generated question information GQ are displayed in the talk display area 114 of the FAQ-AI screen 112 of the user terminal 110. As shown in FIG. 6, when the user selects one of the multiple generated question information GQ by clicking or tapping, the text of the selected generated question information GQ is entered into the input field 114a and it is determined that a send operation has been performed.

[0040] In step S5, which is a prompt transmission step, user terminal 110 transmits text data to first server 120 as question information PT. For example, text data to the effect of "How do I settle the travel expenses?" is transmitted as question information PT.

[0041] In step S6, as the text generation and transmission step, the first server 120 uses ChatGPT, which is an example of the pre-trained artificial intelligence means 131 of the second server 130, to execute the generation of text data based on the question information PT within the reference range RG. For example, when the item 113a of "expense-related" is selected and there is question information PT to the effect of "What is the method for settling business trip expenses?", ChatGPT, which is an example of the pre-trained artificial intelligence means 131, refers to (pre-learns) the information specified within the reference range RG by documents in the "expense-related folder" of the preset file server as an example, and generates text data to the effect of "Perform in the following steps. 1. Log in to the expense settlement flow system...." as an answer to the question. Then, the first server 120 transmits the generated text data GT to the effect of "Perform in the following steps. 1. Log in to the expense settlement flow system...." obtained by execution to the user terminal 110.

[0042] In step S7, as the generated text display step, the user terminal 110 displays the received generated text data GT on the FAQ-AI screen 112 of the browser 111. For example, the generated text data GT to the effect of "Perform in the following steps. 1. Log in to the expense settlement flow system...." is displayed on the FAQ-AI screen 112.

[0043] In step S8, as the selection operation presence / absence determination step, in the FAQ-AI screen 112, the browser 111 determines the presence / absence of a selection operation of the button 115b for selecting the database 121. If it is determined that there is an operation, it proceeds to step S9. On the other hand, if it is determined that there is not yet an operation, it returns to step S3.

[0044] In step S9, as the database content display step, the browser 111 of the user terminal 110 displays the preset category information CT based on the database 121, or the question information QT and the answer information AN. More specifically, as shown in FIG. 7, the browser 111 of the user terminal 110 displays the FAQ database screen 116 based on the database 121 of the first server 120 (the settings of the category information CT, question information QT, and answer information AN on the administrator screen 142). The FAQ database screen 116 is configured to display the category information CT, question information QT, and answer information AN based on the settings of the category information CT, question information QT, and answer information AN on the administrator screen 142.

[0045] Here, on the FAQ database screen 116, in order to make it easier for the user to find the answer he or she desires, first, only the category information CT may be displayed, or only the question information QT may be displayed, or the category information CT, question information QT, and answer information AN may be displayed as a set. In other words, it is sufficient that at least one of the category information CT and the question information QT is displayed, and the answer information AN may be foldable so that it can be freely switched between displayed and hidden.

[0046] As an example, as shown in FIG. 7, the FAQ database screen 116 displays information on the item 113a and the corresponding category information CT in a list form. When one of the category information CT is selected, the browser 111 pops up an FAQ details display window 117 . In the FAQ details display window 117, the question information QT and answer information AN in the category information CT selected on the FAQ database screen 116 are displayed in a list form. As an example, the FAQ details display window 117 displays question information QT in a list, and when an arrow on the right side of the question information QT is operated, answer information AN to that question is displayed below the question information QT.

[0047] As a result, as described above, a button 115a for selecting chat GPT, which is an example of a pre-trained artificial intelligence means 131, and a button 115b for selecting database 121 are displayed on the FAQ-AI screen 112. As a result, the user can choose between Chat GPT, an example of a pre-trained artificial intelligence means 131 that provides flexible answers that may not be 100% correct, or the database 121, which is not flexible but has fixed categories and answers that are 100% correct, depending on the situation. In other words, the appropriate one can be selected depending on the situation, and an answer appropriate to the situation can be obtained. Additionally, you can switch between the two, compare the answers, and adopt the more appropriate answer.

[0048] In addition, when an item 113a is selected in the item display area 113 of the FAQ-AI screen 112 and a button 115b for selecting database 121 is operated, the FAQ details display window 117 for the category information CT corresponding to that item 113a may be configured to open. This eliminates the need for the user to repeatedly select the item 113a on the FAQ-AI screen 112 and the category information CT on the FAQ database screen 116. As a result, the user can immediately find the desired question information QT in the FAQ details display window 117 and obtain the answer information AN to that question.

[0049] Furthermore, in this embodiment, as shown in FIGS. 5, 6 and 8, the FAQ-AI screen 112 further includes a button 115c for chatting with an operator. As shown in FIG. 8, when button 115c indicating that you want to chat with an operator is selected on the FAQ-AI screen 112, the FAQ-AI screen 112 of the user terminal 110 is configured to display message data MG between the user terminal 110 and the operator terminal.

[0050] Specifically, when button 115c indicating that the user wants to chat with an operator is selected, the interaction between the user and chat GPT, an example of a pre-trained artificial intelligence means 131, in the talk display area 114 is switched to an interaction between the user and the operator using text message data MG. In other words, the icon representing the person with whom you are interacting is switched from bot icon 114b to operator icon 114f.

[0051] This increases the option of chatting with an operator by using the button 115c for chatting with an operator. As a result, if a user is unable to obtain the intended answer even when using chat GPT, an example of the pre-trained artificial intelligence means 131, or the database 121, the user can chat with an operator to obtain the desired answer.

[0052] In this embodiment, as shown in FIGS. 5, 6, 8, and 9, the FAQ-AI screen 112 further includes a button 115d for speaking with an operator. Then, as shown in FIG. 9, when button 115d for speaking with an operator is selected on FAQ-AI screen 112, browser 111 activates the microphone of user terminal 110, and user terminal 110 makes a call to the operator terminal.

[0053] At this time, the browser 111 displays an operator call display window 118 in a pop-up form. When a call operation is performed at the operator terminal, the user terminal 110 and the operator terminal are configured to be in a call state. During the call, the operator call display window 118 displays "Connecting" or "Ringing", and when the call is in progress, the display changes to "Connected to operator. Please speak to an operator!".

[0054] This increases the option of speaking with an operator by using button 115d for speaking with an operator. As a result, if a user is unable to obtain the intended answer even when using chat GPT, an example of the pre-trained artificial intelligence means 131, or the database 121, the user can speak to an operator to obtain the desired answer.

[0055] Furthermore, in this embodiment, as shown in Figures 5, 6, 8, and 9, the FAQ-AI screen 112 has a talk display area 114 that displays question information PT and generated text data GT, which are an exchange between the user and the chat GPT, which is an example of a pre-trained artificial intelligence means 131, and an item display area 113 that displays items 113a of the exchange. Furthermore, a reference range RG of the chat GPT, which is an example of the pre-trained artificial intelligence means 131, is set in advance for each item 113a.

[0056] When an operation for selecting and changing an item 113a in the item display area 113 is performed, the browser 111 is configured to display, in the talk display area 114, an exchange corresponding to the item 113a. Thereby, the reference range RG is changed by selecting the item 113a through a selection change operation of the item 113a. As a result, the user can select the desired reference range RG and interact with chat GPT, which is an example of a pre-trained artificial intelligence means 131.

[0057] In this embodiment, the item 113a is the name of a department in an organization. For example, as shown in FIG. 5, FIG. 6, FIG. 8, and FIG. 9, the items 113a are department names such as "General Affairs Department", "Personnel Department", "Sales Department", "Public Relations Department", and "Legal Affairs Department", as examples. Furthermore, the reference range RG set for each item 113a is handling information by a department of an organization corresponding to the item name set based on at least one of URL information, site map information, distribution information, document information, and folder designation information. For example, as shown in FIG. 3, the reference range RG set for the "General Affairs Department" in the item 113a is information on document files in the "General Affairs Department Information Folder" folder.

[0058] Similarly, the reference range RG set for the "Personnel Department" in item 113a is information on document files in the "Personnel Department Information Folder" folder. Moreover, the reference range RG set for the "Sales department" in the item 113a is information on document files in the "Sales department information folder" folder. Furthermore, the reference range RG set for the "Public Relations Department" in item 113a is information on document files in the "Public Relations Department Information Folder" folder. Also, the reference range RG set for the "Legal department" in item 113a is information on document files in the "Legal department information folder" folder.

[0059] As a result, when the item 113a of the department name of the organization is selected, the reference range RG becomes the information handled by that department. As a result, the user can select the item 113a of the name of the department in charge of the question and interact with Chat GPT, an example of the pre-trained artificial intelligence means 131, to obtain the desired answer. In other words, by selecting the correct department in charge of the question, the risk of receiving a misguided answer from Chat GPT, an example of pre-trained artificial intelligence means 131, can be avoided and the desired answer with a high accuracy rate can be obtained from Chat GPT, an example of pre-trained artificial intelligence means 131.

[0060] Furthermore, in this embodiment, as described above, the first server 120 generates text data based on the question information PT within the reference range RG using the chat GPT, which is an example of the pre-trained artificial intelligence means 131 of the second server 130. Then, as shown in FIG. 10, when the first server 120 transmits the generated text data GT obtained by execution to the user terminal 110, the first server 120 is configured to register the question information PT and the generated text data GT as a set in the database 121 as question information QT and answer information AN. For example, as shown in Figure 5, one of the multiple generated question information GQ generated by chat GPT, which is an example of a pre-trained artificial intelligence means 131, ``How do I settle travel expenses?'', is entered as question information QT in the input field 114a of the FAQ-AI screen 112, as shown in Figure 6.

[0061] Then, as shown in Figure 6, the chat GPT, which is an example of the pre-trained artificial intelligence means 131, sends generated text data GT to the user terminal 110 as answer information AN to the question, stating, "Please follow the steps below: 1. Log in to the expense reimbursement flow system..." At this time, the first server 120 is configured to additionally register the question information QT and answer information AN as a set in the database 121, as shown in FIG. As a result, the generated answer information AN is sequentially added to the database 121 together with the question information QT. As a result, a database 121 of FAQs based on chat GPT, which is an example of a pre-trained artificial intelligence means 131, can be created.

[0062] At this time, the information on the selected item 113a may be added as category information CT to the set of question information QT and answer information AN and additionally registered in the FAQ database 121. As a result, the category information CT is included in the lower layer of the list display on the FAQ database screen. As a result, when the FAQ details display window corresponding to that category information CT is popped up, the set of additionally registered question information QT and answer information AN can be displayed.

[0063] The talk display area 114 of the FAQ-AI screen 112 is configured to display a speaker icon 114c, a good evaluation icon 114d, and a bad evaluation icon 114e corresponding to each exchange. Of these, when the speaker icon 114c is operated, the contents of the exchange, such as the question information PT and the generated text data GT, are converted into voice data and output as voice using the speaker of the user terminal 110. In addition, when the good evaluation icon 114d corresponding to the generated text data GT as the answer information AN is operated, the question information QT and the answer information AN are added and registered as a set in the database 121, and on the other hand, when the bad evaluation icon 114e corresponding to the generated text data GT as the answer information AN is operated, the question information QT and the answer information AN are not added and registered as a set in the database 121.

[0064] Additionally, embeddings may be used to improve the accuracy of answers to questions. More specifically, first, the first server 120 divides the language text data of the reference range RG into multiple parts, such as by a specified number of characters or by sentence, and converts each of the divided divided language text data into vector coordinate system divided language data using a vector coordinate conversion means, and associates the converted vector coordinate system divided language data with the divided language text data before conversion and stores them in vector indexes.

[0065] Then, the first server 120 converts the question information PT, such as the received question, into vector coordinate system question data using a vector coordinate conversion means, searches for vector coordinate system divided language data that is most closely related to the vector of the vector coordinate system question data in the vector index, and selects divided language text data that corresponds to the most closely related vector coordinate system divided language data or the top multiple most closely related vector coordinate system divided language data. Next, the first server 120 may be configured to add the selected divided language text data to the question information PT, and perform text data generation using chat GPT, which is an example of the pre-trained artificial intelligence means 131 of the second server 130, and transmit the generated text data GT obtained by the execution to the user terminal 110.

[0066] As a result, language text data regarding content information identified by a predetermined document file, URL information, and folder designation information is divided into multiple pieces, each of which is converted into vector coordinate system divided language data and stored in a vector index, and question information PT entered by the user is also converted into vector coordinate system question data, vector coordinate system divided language data with a similar relationship is searched for, and divided language text data corresponding to the vector coordinate system divided language data is added to the question information PT, and based on this, generated text data GT is generated by a text generation means, which is a chat GPT, an example of a pre-trained artificial intelligence means 131, and displayed as an answer on the browser 111 of the user terminal 110.

[0067] As a result, the accuracy of the answer according to the user's intention can be improved. In other words, the language text data is converted into vector coordinate system data that is easy for a computer to compare (understand), a correlation search is performed on the vector coordinate system data for each divided part of the original document and the question information PT, and the divided part of the original document corresponding to the vector coordinate system data with a high correlation is added to the question information PT, and the text generation means, which is the chat GPT as an example of the pre-trained artificial intelligence means 131, does not need to read the less related parts in the language text data, but reads only the highly related parts and the question information PT to generate the generated text data GT, thereby improving the accuracy of the answer.

[0068] Embedding may also be used for the generated question information GQ. The language text data in the reference range RG is stored in a vector index using embedding, and when generating the question information QT, the question information QT is generated based on the vector coordinate system data and the language text data. When the question information QT and answer information AN are to be added to the database 121 as a set, if it is determined that the correlation values ​​between the vector coordinate system data of the question information QT and answer information AN to be added are similar within a predetermined range in relation to the vector coordinate system data of the question information QT and answer information AN already registered, then the information is not added. On the other hand, if it is determined that the correlation values ​​between the two are not similar and fall outside a predetermined range, they may be added.

[0069] As a result, Q&As similar to those already registered in the database 121 are not added, and only dissimilar Q&As are added. As a result, it is possible to avoid all of the exchanges of question information QT and answer information AN being indiscriminately added to the database 121, and to avoid the registration of duplicate or similar Q&As.

[0070] In addition, when a document file or the like in the reference range RG of the chat GPT is updated, the first server 120 is configured to detect the update and read the text data of the updated document into the chat GPT, which is an example of the pre-trained artificial intelligence means 131. If there is a limit to the number of characters that can be read at one time, the text is of course divided into a number of characters within the limit and then read. In the case of a configuration using embedding, the first server 120 detects that the update has occurred and re-performs the embedding.

[0071] The artificial intelligence system 100, which is an embodiment of the present invention obtained in this manner, comprises a user terminal 110, a first server 120 and a second server 130 as servers, and an administrator terminal 140, in which a reference range RG of the chat GPT, which is an example of a pre-trained artificial intelligence means 131, is set in advance based on at least one of one or more URL information, site map information, distribution information, document information, and folder designation information designated in advance by the administrator terminal 140, category information CT, question information QT, and answer information AN in the database 121 of the first server 120 are set in advance by the administrator terminal 140, the browser 111 of the user terminal 110 is capable of freely interacting with the chat GPT, which is an example of the pre-trained artificial intelligence means 131 of the second server 130, via the first server 120, and displays a FAQ-AI screen 112 as a talk screen having a button 115a for selecting the chat GPT, which is an example of the pre-trained artificial intelligence means 131, and a button 115b for selecting the database 121, and in the FAQ-AI screen 112, the pre-trained artificial intelligence means 131 is displayed. When a button 115a for selecting chat GPT, which is an example of a pre-trained artificial intelligence means 131, is selected and text is input, the user terminal 110 transmits the text data to the first server 120 as question information PT, the first server 120 executes text data generation based on the question information PT for the reference range RG using chat GPT, which is an example of a pre-trained artificial intelligence means 131 of the second server 130, and transmits the generated text data GT obtained by the execution to the user terminal 110, the user terminal 110 displays the received generated text data GT on the FAQ-AI screen 112 of the browser 111, and when a button 115b for selecting database 121 is selected on the FAQ-AI screen 112, the browser 111 of the user terminal 110 is configured to display category information CT, or question information QT and answer information AN that are preset based on the database 121. As a result, the user can select chat GPT, which is an example of a pre-trained artificial intelligence means 131 that provides a flexible answer, although it may not be 100% correct, orDepending on the situation, you can choose between rigid categories and a database of 121 answers that are 100% correct.

[0072] Furthermore, the FAQ-AI screen 112 further includes a button 115c for chatting with an operator, When button 115c indicating that the user wants to chat with an operator is selected on FAQ-AI screen 112, FAQ-AI screen 112 of user terminal 110 is configured to display message data MG between user terminal 110 and operator terminal. This enables a user to chat with an operator to obtain a desired answer when the user is unable to obtain the intended answer even when using chat GPT, an example of pre-trained artificial intelligence means 131, or database 121.

[0073] In addition, the FAQ-AI screen 112 further has a button 115d for speaking with an operator. When the button 115d for speaking with an operator is selected on the FAQ-AI screen 112, the microphone of the user terminal 110 becomes active and the user terminal 110 makes a call to the operator terminal. When a call operation is performed on the operator terminal, the user terminal 110 and the operator terminal are placed in a call state. As a result, the number of options for speaking with an operator increases by the button 115d for speaking with an operator, so that when the user cannot obtain the intended answer even when using chat GPT, an example of the pre-trained artificial intelligence means 131, or the database 121, the user can speak with an operator to obtain the desired answer.

[0074] Furthermore, the FAQ-AI screen 112 has a talk display area 114 that displays question information PT and generated text data GT, which are an interaction between the user and the chat GPT, which is an example of the pre-trained artificial intelligence means 131, and an item display area 113 that displays the interaction items 113a.The reference range RG of the chat GPT, which is an example of the pre-trained artificial intelligence means 131, is set in advance for each item 113a, and when a selection change operation is performed on item 113a in the item display area 113, the browser 111 is configured to display the interaction corresponding to item 113a in the talk display area 114, so that the user can select the desired reference range RG and interact with the chat GPT, which is an example of the pre-trained artificial intelligence means 131.

[0075] In addition, since item 113a is the name of a department in an organization, and the reference range RG set for each item 113a is handling information by the department in the organization corresponding to the item name set based on at least one of URL information, site map information, distribution information, document information, and folder designation information, the user can select item 113a with the name of the department in charge of the question and interact with chat GPT, which is an example of pre-trained artificial intelligence means 131, to obtain the desired answer.

[0076] Furthermore, when the first server 120 generates text data based on question information PT within the reference range RG using chat GPT, which is an example of the pre-trained artificial intelligence means 131 of the second server 130, and transmits the generated text data GT obtained by this execution to the user terminal 110, the first server 120 is configured to register the question information PT and the generated text data GT as a set in the database 121 as question information QT and answer information AN, thereby making it possible to create a database 121 of FAQs using chat GPT, which is an example of the pre-trained artificial intelligence means 131.

[0077] In addition, the program of the artificial intelligence system 100, which is an embodiment of the present invention, includes a reference range setting step in which a reference range RG of the chat GPT, which is an example of the pre-trained artificial intelligence means 131, is set based on at least one of one or more URL information, site map information, distribution information, document information, and folder designation information specified by the administrator terminal 140; a database setting step in which category information CT, question information QT, and answer information AN in the database 121 of the first server 120 are set by the administrator terminal 140; a talk screen display step S1 in which the user terminal 110 accesses the first server 120, and the browser 111 of the user terminal 110 is able to freely interact with the chat GPT, which is an example of the pre-trained artificial intelligence means 131 of the second server 130, via the first server 120, and displays a FAQ-AI screen 112 as a talk screen having a button 115a for selecting the chat GPT, which is an example of the pre-trained artificial intelligence means 131, and a button 115b for selecting the database 121; a selection operation presence / absence determination step S3, S8 for determining whether or not a selection operation has been performed on a button 115a for selecting chat GPT, which is an example of the pre-trained artificial intelligence means 131, or a button 115b for selecting database 121; a prompt transmission step S4, S5 for the user terminal 110 to transmit the text data as question information PT to the first server 120 when the button 115a for selecting chat GPT, which is an example of the pre-trained artificial intelligence means 131, is selected and text is input; a text generation and transmission step S6 in which text data is generated based on question information PT within the reference range RG using a chat GPT, which is an example of 131, and the generated text data GT obtained by the execution is transmitted to the user terminal 110; a generated text display step S8 in which the user terminal 110 displays the received generated text data GT on the FAQ-AI screen 112 of the browser 111; and when a button 115b for selecting database 121 is selected on the FAQ-AI screen 112, the browser 111 of the user terminal 110By having a database content display step S9 that displays category information CT or question information QT and answer information AN that are preset based on the database 121, the user can select either chat GPT, which is an example of a pre-trained artificial intelligence means 131 that gives flexible answers that may not be 100% correct, or database 121 that is 100% correct with fixed categories and answers that are not flexible, depending on the situation, and the effect is enormous. [Explanation of symbols]

[0078] 100 ··· Artificial Intelligence Systems 110 User terminal 111: Browser (on user device) 112 ··· FAQ-AI screen (talk screen) 113... Item display area 113a... Item 114 ··· Talk display area 114a Input field 114b··· Bot icon 114c··· speaker icon 114d··· Good rating icon 114e··· Bad rating icon 114f··· Operator Icon 115a: Button for selecting pre-trained artificial intelligence means 115b... Database selection button 115c··· Button to chat with an operator 115d: Button to talk to an operator 116 ··· FAQ database screen 117 ··· FAQ detail display window 118 ··· Operator call display window 120 1st Server 121 ··· Database 122 Bot 130 Second Server 131 ··· Pre-trained AI method (Chat GPT) 140 Administrator terminal 141: Browser (on administrator's terminal) 142...Administrator screen RG: Reference range (of pre-trained AI means) QT ··· Question information AN...Answer information CT ··· Category Information PT: (User) question information (instruction information, prompt) GT: Generated text data MG: Message data (between user terminal and operator terminal) GQ: Generated Question Information

Claims

1. An artificial intelligence system comprising a user terminal, a server, and an administrator terminal, the user terminal displaying or outputting sound based on the server, A reference range of the pre-trained artificial intelligence means is preset based on at least one of one or more pieces of URL information, site map information, information to be distributed, document information, and folder designation information that are preset by the administrator terminal; The administrator terminal presets category information, question information, and answer information in the server database; A browser of the user terminal is capable of communicating with the pre-trained artificial intelligence means of the server, and displays a talk screen having a button for selecting the pre-trained artificial intelligence means and a button for selecting a database; When a button for selecting a pre-trained artificial intelligence means is selected on the talk screen and text is input, the user terminal transmits the text data as question information to a server; The server uses the pre-trained artificial intelligence means to generate text data based on question information about the reference range, and transmits the generated text data to a user terminal; The user terminal displays the received generated text data on a chat screen of a browser, An artificial intelligence system characterized in that, when a button for selecting a database is selected on the talk screen, the browser of the user terminal displays category information, or question information and answer information that have been preset based on the database.

2. The talk screen further includes a button for chatting with an operator, The artificial intelligence system described in claim 1, characterized in that when a button indicating to chat with an operator is selected on the talk screen, the talk screen of the user terminal displays message data between the user terminal and the operator terminal.

3. The talk screen further includes a button for talking to an operator, When a button for talking to an operator is selected on the talk screen, a microphone of the user terminal becomes active and the user terminal makes a call to an operator terminal; 3. The artificial intelligence system according to claim 2, wherein when a call operation is performed on the operator terminal, the user terminal and the operator terminal are placed in a call state.

4. The talk screen has a talk display area for displaying question information and generated text data which are an exchange between a user and a pre-trained artificial intelligence means, and an item display area for displaying items of the exchange; The reference range of the pre-trained artificial intelligence means is set in advance for each item, The artificial intelligence system described in any one of claims 1 to 3, characterized in that when an item in the item display area is selected or changed, the browser displays an exchange corresponding to the item in the talk display area.

5. The item is a department name of an organization, The artificial intelligence system described in claim 4, characterized in that the reference range set for each item is handling information by an organizational department corresponding to the item name set based on at least one of URL information, site map information, distribution information, document information, and folder designation information.

6. The artificial intelligence system described in claim 5, characterized in that the server uses the pre-trained artificial intelligence means to generate text data based on question information within the reference range, and when the server transmits the generated text data obtained by the execution to a user terminal, the server registers the question information and the generated text data as a set in a database as question information and answer information.

7. A program of an artificial intelligence system that a user terminal displays or outputs voice based on a server, a reference range setting step in which a reference range of the pre-trained artificial intelligence means is set based on at least one of one or more URL information, site map information, distribution information, document information, and folder designation information designated by an administrator terminal; a database setting step in which category information, question information, and answer information are set in the database of the server by the administrator terminal; a talk screen display step of displaying a talk screen in which the user terminal accesses a server, a browser of the user terminal is capable of communicating with the pre-trained artificial intelligence means of the server, and a button for selecting the pre-trained artificial intelligence means and a button for selecting a database are provided; A selection operation presence / absence determination step of determining whether or not a selection operation of a button for selecting a pre-trained artificial intelligence means or a button for selecting a database has been performed on the talk screen; a prompt transmission step in which, when a button for selecting the pre-trained artificial intelligence means is selected and text is input, the user terminal transmits the text data as question information to a server; a text generation and transmission step in which the server uses the pre-trained artificial intelligence means to generate text data based on question information about the reference range, and transmits the generated text data to a user terminal; a generated text display step in which the user terminal displays the received generated text data on a talk screen of a browser; A program for an artificial intelligence system, characterized in that when a button for selecting the database is selected on the talk screen, the browser of the user terminal displays pre-set category information or question information and answer information based on the database.

Images