Browser extension program
The program enhances browser functionality to facilitate voice input and output interactions with pre-trained AI, addressing usability issues and improving security and customization.
Patent Information
- Application Number
- JP2024005141
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- Filing Date
- 2024-01-17
- Publication Date
- 2025-07-31
AI Technical Summary
Conventional information processing apparatuses require users to manually type text for interaction with pre-trained artificial intelligence, leading to insufficient usability.
A program that extends browser functionality to enable voice input and output interactions with pre-trained artificial intelligence, including voice input icon display, microphone activation, voice-to-text conversion, and voice output, along with features like prompt lists, history logging, prohibited word settings, and IP address restrictions.
Enables seamless voice-based communication with pre-trained AI without typing, enhances security through prohibited word filtering and IP restrictions, and improves usability by allowing voice output and customizable interfaces.
Smart Images

Figure 2025111855000001_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to a program for extending the functions of a browser that displays the interaction between a user terminal and pre-trained artificial intelligence means of a server.
Background Art
[0002] Conventionally, 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 summary basic information that serves as a basis for summary generation, and presentation information including the summary basic information are transmitted to the user terminal. A transmission unit for transmitting, an edit information receiving unit for receiving edit information regarding the presentation information from the user terminal, and a large language model that receives an input of the summary basic information and generates a summary text, and provides the edit information to generate a summary text in which the edit information is reflected. An information processing apparatus including a summary generation unit is known (for example, Patent Document 1).
Prior Art Documents
Patent Documents
[0003]
Patent Document 1
Summary of the Invention
Problems to be Solved by the Invention
[0004] However, since the above-described conventional information processing apparatus only generates a summary of some summary basic information using a large language model, the user has to operate a keyboard or a character key display panel for input and directly enter the text into the input field, resulting in a problem that the usability is not sufficient.
[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 a program for extending the function of a browser that enables a user to communicate with pre-trained artificial intelligence means by voice input without typing, even if the original does not support voice input.
Means for Solving the Problems
[0006] The invention according to claim 1 is a program for extending the function of a browser that displays the interaction between a user terminal and pre-trained artificial intelligence means of a server. In a talk screen where the browser of the user terminal displays the interaction between the pre-trained artificial intelligence means of the server and the user of the user terminal, the method includes: a voice input icon display step of displaying a voice input icon button on the talk screen; a voice input icon operation determination step of determining whether the voice input icon button has been operated; a microphone activation step of activating the microphone of the user terminal when the voice input icon button is operated; and a voice input text conversion and transmission step of converting the input voice data into text data when the user's voice is input through the microphone, inputting the converted text data into an input field, and transmitting it from the user terminal to the pre-trained artificial intelligence means of the server, thereby solving the problems described above.
[0007] In addition to the configuration of the program for extending the function of the browser according to claim 1, the invention according to claim 2 further includes: a voice output icon display step of displaying a voice output icon button corresponding to the generated text data when the browser of the user terminal displays the generated text data generated by the pre-trained artificial intelligence means received from the server on the talk screen; and a voice output step of outputting the content of the generated text data corresponding to the voice output icon button as voice through the speaker of the user terminal when the voice output icon button is operated, thereby further solving the problems described above.
[0008] The invention according to claim 3, in addition to the configuration of the program for extending the functions of the browser described in claim 2, further includes: an instruction command icon display step in which the browser of the user terminal displays an instruction command list display icon button on the talk screen; a prompt list display step in which when the instruction command list display icon button is operated, a prompt list display window for displaying preset and registered standard instruction command information items in a list format is displayed; and a prompt transmission step in which when the instruction command information in the prompt list display window is selected and operated, the selected instruction command information is input into the input field and transmitted from the user terminal to the pre-trained artificial intelligence means of the server, thereby further solving the above-described problems.
[0009] The invention according to claim 4, in addition to the configuration of the program for extending the functions of the browser described in any one of claims 1 to 3, further includes: a log storage step in which the server stores the history data of the interaction between the user terminal and the pre-trained artificial intelligence means on the talk screen of the user terminal; and a history display step in which the browser of the administrator terminal capable of communicating with the server displays the history data of the interaction between the user terminal and the pre-trained artificial intelligence means on the talk screen of the user terminal on the administrator screen based on the operation of the administrator, thereby further solving the above-described problems.
[0010] The invention according to claim 5, in addition to the configuration of the program for extending the functions of the browser described in claim 4, when there is a setting operation for prohibited words or prohibited attributes on the administrator screen, includes a prohibited input setting step of registering the setting for prohibited words or prohibited attributes on the server, a prohibited input determination step of determining whether the content of the text data input in the input field of the talk screen of the user terminal is a preset prohibited word or prohibited attribute, and a transmission operation stop and warning display step of invalidating the transmission operation of the input text data and displaying a warning indicating that a prohibited word or prohibited attribute has been input when the content of the text data input in the input field is a preset prohibited word or prohibited attribute, thereby further solving the above-described problems.
[0011] The invention according to claim 6, in addition to the configuration of the program for extending the functions of the browser described in claim 5, when there is a setting operation for IP address restriction on the administrator screen, includes an IP address restriction setting step of registering the setting for IP address restriction on the server, an IP address condition determination step of determining whether the IP address information of the user terminal included in the request information from the user terminal satisfies the conditions of the preset IP address restriction setting when the browser of the user terminal accesses the server to display the talk screen, a talk screen display step of the server transmitting first response information to the user terminal and the browser of the user terminal displaying the talk screen when the IP address information of the user terminal satisfies the conditions of the preset IP address restriction setting, and an NG display step of the server transmitting second response information different from the first response information to the user terminal and causing the browser to display that the IP address information of the user does not satisfy the conditions of the IP address restriction setting when the IP address information of the user terminal does not satisfy the conditions of the preset IP address restriction setting, thereby further solving the above-described problems.
[0012] The invention according to claim 7, in addition to the configuration of the program for extending the functions of the browser described in claim 6, further includes a mail sending icon display step of displaying a mail sending icon button on the chat screen, a mail sending icon operation determination step of determining whether the mail sending icon button has been operated, and a mail sending step of the server sending the content of the interaction between the pre-trained artificial intelligence means displayed on the chat screen and the user of the user terminal to the e-mail address of the user setting registered in advance when the mail sending icon button has been operated, thereby further solving the above-described problems.
[0013] The invention according to claim 8, in addition to the configuration of the program for extending the functions of the browser described in claim 7, further includes an interface setting operation presence / absence determination step of determining whether there has been a predetermined interface setting operation on the chat screen, an interface setting window display step of displaying an interface setting window in which the colors of characters and the background can be freely selected when the predetermined interface setting operation has been performed, a selection / setting save operation presence / absence determination step of determining whether there has been a selection operation and a setting save operation in the interface setting window, and a color reflection step of displaying the colors of characters and the background on the chat screen based on the selected content when there has been a selection operation and a setting save operation, thereby further solving the above-described problems.
[0014] In addition to the configuration of the program for extending the functions of the browser according to claim 8, the invention according to claim 9 further includes a shared prompt setting step of registering the settings for the shared standard instruction command information on the server when there is a setting operation for the shared standard instruction command information on the administrator screen, and a prompt list display step of displaying a prompt list display window that can switchably display in list form the content of the preset personal standard instruction command information items and the content of the preset shared standard instruction command information items when the instruction command list display icon button is operated on the talk screen, and a shared prompt transmission step of inputting the selected shared standard instruction command information into the input field and transmitting it from the user terminal to the pre-trained artificial intelligence means of the server when the shared standard instruction command information in the prompt list display window is selected. By having these steps, the above-mentioned problems can be further solved.
Advantages of the Invention
[0015] The program for extending the functions of the browser of the present invention, when installed on a user terminal, can not only extend the functions of the browser of the user terminal, but also achieve the following specific effects.
[0016] According to the program for extending the functions of the browser of the invention according to claim 1, even on a talk screen that does not originally support voice input, by using the program for extending the functions of the browser, a voice input icon button is displayed on the talk screen. When the voice input icon button is operated and voice input is performed, the voice content is converted into text data, input into the input field, and transmitted. Therefore, the user can communicate with the pre-trained artificial intelligence means by voice input without typing.
[0017] According to the program for extending the functions of the browser of the invention according to claim 2, in addition to the effects achieved by the invention according to claim 1, even for a talk screen that originally does not support voice output, by using the program for extending the functions of the browser, a voice output icon button is displayed corresponding to the generated text data of the talk screen. When the voice output icon button is operated, the content of the corresponding generated text data is voice output. Therefore, the user can understand the answer from the pre-trained artificial intelligence means by voice output without reading the generated text data visually.
[0018] According to the program for extending the functions of the browser of the invention according to claim 3, in addition to the effects achieved by the invention according to claim 2, even for a talk screen that originally does not support a prompt which is stereotyped instruction information, by using the program for extending the functions of the browser, the prompt can be set and registered in advance. When the instruction list display icon button is operated, an instruction list display window is displayed. When the desired instruction information is selected from it and input into the input field and sent to the pre-trained artificial intelligence means, the user can easily communicate with the pre-trained artificial intelligence means using the prompt and obtain an answer in the desired format and character amount from the pre-trained artificial intelligence means.
[0019] According to the program for extending the functions of the browser of the invention according to claim 4, in addition to the effects achieved by the invention according to any one of claims 1 to 3, the log of the interaction between the user terminal and the pre-trained artificial intelligence means is saved and displayed on the administrator screen of the administrator terminal. Therefore, the administrator can manage the user's use from the perspective of security.
[0020] According to the program for extending the functions of the browser of the invention according to claim 5, in addition to the effects achieved by the invention according to claim 4, when a telephone number or an email address, as an example of a prohibited attribute, is entered in the input field by a prohibited word or a prohibited attribute preset on the administrator screen, the transmission operation becomes invalid and the interaction between the user and the pre-trained artificial intelligence means stops and enters a stopped state. Therefore, it is possible to avoid the pre-trained artificial intelligence means from learning personal information and the like. Furthermore, it is also possible to avoid personal information from leaking to other terminals via the pre-trained artificial intelligence means. Also, as an example of a prohibited attribute, when a telephone number or an email address is entered in the input field, a warning is displayed, so that the user can notice that personal information and the like have been entered.
[0021] According to the program for extending the functions of the browser of the invention according to claim 6, in addition to the effects achieved by the invention according to claim 5, for example, only connections in a safe Internet environment are permitted and connections in an Internet environment with unknown security are blocked by the IP address restriction preset on the administrator screen. Therefore, the administrator can easily take security measures for the user's use.
[0022] According to the program for extending the functions of the browser of the invention according to claim 7, in addition to the effects achieved by the invention according to claim 6, even in a chat screen that does not originally support email sending, by using the program for extending the functions of the browser, an email sending icon button is displayed on the chat screen. When the email sending icon button is operated, the content of the interaction displayed on the chat screen is sent to the email address registered in the user settings. Therefore, the user can leave a log as an email and can also share it with other users if necessary.
[0023] According to the program for extending the functions of a browser according to the invention of claim 8, in addition to the effects achieved by the invention of claim 7, even if the color of the chat screen cannot be changed originally, by using the program for extending the functions of the browser, the color of the chat screen can be freely changed by a setting operation (a predetermined interface setting operation, a selection operation, a setting save operation) by the user. Therefore, it can be changed to a color that is easy for the user to see. For example, when the user has characteristics such as color weakness or difficulty seeing specific colors, by changing the setting to a color that is easy for the user to see, the user can make the content of the conversation on the chat screen easier to see for himself / herself.
[0024] According to the program for extending the functions of a browser according to the invention of claim 9, in addition to the effects achieved by the invention of claim 8, even for a chat screen that does not originally correspond to a prompt, which is shared standard instruction information, by using the program for extending the functions of the browser, a shared prompt can be freely set and registered by the administrator in advance. When the prompt list display window is displayed by operating the instruction list display icon button and a desired shared standard instruction information is selected from it, it is input into the input field and sent to the pre-trained artificial intelligence means. Therefore, the user can easily communicate with the pre-trained artificial intelligence means using the shared prompt and obtain an answer in the format and character amount desired from the pre-trained artificial intelligence means. That is, since a list of shared prompts pre-registered by the administrator is displayed in the prompt list display window without the user registering the prompts himself / herself, the user can easily communicate with the pre-trained artificial intelligence means by simply selecting a desired shared prompt from it and obtain an answer in the format and character amount desired from the pre-trained artificial intelligence means.
Brief Description of the Drawings
[0025]
Figure 1
Figure 2
Figure 3
Figure 4
Figure 5
Figure 6
Figure 7
Figure 8
Figure 9
Figure 10
Figure 11
Embodiments for Carrying Out the Invention
[0026] The program for extending the functions of the browser of the present invention includes: on the talk screen where the browser of the user terminal displays the interaction between the pre-trained artificial intelligence means of the server and the user of the user terminal, a voice input icon display step of displaying a voice input icon button on the talk screen; a voice input icon operation determination step of determining whether the voice input icon button has been operated; when the voice input icon button is operated, a microphone activation step of activating the microphone of the user terminal; and when the user's voice is input via the microphone, a voice input text conversion and transmission step of converting the input voice data into text data, inputting the converted text data into the input field, and transmitting it from the user terminal to the pre-trained artificial intelligence means of the server. By having these steps, if the user can interact with the pre-trained artificial intelligence means by voice input without typing, the specific implementation mode can be any.
[0027] For example, the user terminal can send and receive information such as a desktop personal computer terminal, a notebook personal computer terminal, a smartphone terminal, or a tablet terminal, and can be connected to a server through a communication network including a wide area network such as the so-called Internet, a local network, a telephone line, etc., and can be any device as long as it can be connected to the server. Also, the server can be one server or multiple servers on the cloud. The pre-trained artificial intelligence means is composed of, for example, a conversational large language model such as ChatGPT (Generative Pre-trained Transformer) (hereinafter referred to as ChatGPT), etc., and can also be configured to be provided on one server or multiple servers on the cloud.
Example
[0028] Next, a program for extending the function of a browser 111 that displays the interaction between a user terminal 110, which is an example of an embodiment of the present invention, and ChatGPT, which is an example of the pre-trained artificial intelligence means 121 of the server, will be described with reference to FIGS. 1 to 11(B). Here, FIG. 1 is a diagram showing the concept of an artificial intelligence system 100 installed with a program for extending the functions of a browser 111 according to an embodiment of the present invention. FIG. 2 is a chart diagram showing an operation example of the artificial intelligence system 100 installed with a program for extending the functions of a browser 111 according to an embodiment of the present invention. FIG. 3(A) is a diagram showing an example of an initial screen of an original talk screen TK before the program for extending the functions of a browser 111 according to an embodiment of the present invention is installed. FIG. 3(B) is a diagram showing an example of an initial screen of a talk screen 112 whose functions are extended after the program for extending the functions of a browser 111 according to an embodiment of the present invention is installed. FIG. 4(A) is a diagram showing an example of a conversation on an original talk screen TK before the program for extending the functions of a browser 111 according to an embodiment of the present invention is installed. FIG. 4(B) is a diagram showing an example of a conversation on a talk screen 112 whose functions are extended after the program for extending the functions of a browser 111 according to an embodiment of the present invention is installed. FIG. 5(A) is a diagram showing a state in which a prompt list display window 113 is displayed on a talk screen 112 whose functions are extended after the program for extending the functions of a browser 111 according to an embodiment of the present invention is installed. FIG. 5(B) is a diagram showing a state in which a personal fixed instruction information item 113c is selected from the prompt list display window 113. FIG. 6 is a diagram showing a state in which conversation history data LG is displayed on an administrator screen 142 of a browser 141 of an administrator terminal 140 of the artificial intelligence system 100 installed with a program for extending the functions of a browser 111 according to an embodiment of the present invention. FIG. 7(A) is a diagram showing a state in which a prohibited word BW and a prohibited attribute BA are set on an administrator screen 142 of an administrator terminal 140. FIG. 7(B) is a diagram showing a state when a prohibited word BW and a prohibited attribute BA are input on a talk screen 112 of a user terminal 110. FIG. 8(A) is a diagram showing a state in which a restriction condition of an IP address information AD of a user terminal 110 that can use the artificial intelligence system 100 and a user ID that can use the artificial intelligence system 100 are set on an administrator screen 142 of an administrator terminal 140. FIG. 8(B) isFIG. 9(A) is a diagram showing a state when the user terminal 110 accesses the server without satisfying the IP address restriction condition. FIG. 9(A) is a diagram showing a state when the mail sending icon button 112e is operated on the talk screen 112 of the user terminal 110. FIG. 9(B) is a diagram showing a state when an e-mail ML including data of the content of the communication is received on the received mail screen 117 of the browser 111 of the user terminal 110. FIG. 10(A) is a diagram showing a state when the interface setting window 119 is displayed and the setting is changed on the talk screen 112 of the user terminal 110. FIG. 10(B) is a diagram showing a state when the interface setting is reflected on the talk screen 112 of the user terminal 110. FIG. 11(A) is a diagram showing a state when the sharing instruction command information PT is set on the administrator screen 142 of the administrator terminal 140. FIG. 11(B) is a diagram showing a state when the sharing instruction command information PT is displayed in the prompt list display window 113 of the talk screen 112 of the user terminal 110.,
[0029] As shown in FIG. 1, the artificial intelligence system 100 installed with the function expansion program of the browser 111 according to the embodiment of the present invention includes a user terminal 110, a first server 120 and a second server 130 as servers, and an administrator terminal 140 as an example. Among them, the first server 120 has ChatGPT which is an example of pre-trained artificial intelligence means 121, and provides the service of the talk screen TK(112). In other words, the first server 120 provides the service on the talk screen TK(112) to the user terminal 110 as an http response in response to an http request from the user terminal 110. The second server 130 has a database 131 and a bot which is also a display program. As a technical idea, the first server 120 and the second server 130 may have a physically common configuration with each other, or may have a physically different configuration.
[0030] When the user terminal 110 accesses the first server 120 based on predetermined URL information, the browser 111 of the user terminal 110 displays a talk screen TK(112) that can execute an interaction with ChatGPT, which is an example of the pre-trained artificial intelligence means 121 of the server, between the user terminal 110 and the server. When the program for extending the function of the browser 111 of the present invention is installed in the user terminal 110, the function of the talk screen 112 displayed on the browser 111 of the user terminal 110 is configured to be extended compared to the original talk screen TK. In addition, when the function of the talk screen 112 displayed on the browser 111 of the user terminal 110 is extended, the administrator terminal 140 is configured to be able to obtain the log in the talk screen 112 of the user terminal 110 via the second server 130 and to be able to freely set and restrict the user's use.
[0031] More specifically, when the program (plugin) for extending the function of the browser 111 is installed, in the talk screen 112 where the browser 111 of the user terminal 110 displays the interaction between ChatGPT, which is an example of the pre-trained artificial intelligence means 121 of the server, and the user of the user terminal 110, a voice input icon button 112b is displayed in the input field 112a as an example of the talk screen 112. Regarding the display position of the voice input icon button 112b, it may be anywhere within the talk screen 112 and is not limited to within the input field 112a. For example, it may be adjacent to the input field 112a. When the voice input icon button 112b is operated, the microphone of the user terminal 110 is activated. Furthermore, when the user's voice is input via the microphone, the input voice data is converted into text data TX, and the converted text data TX is input into the input field 112a. At the same time, it is configured to be transmitted from the user terminal 110 to ChatGPT, which is an example of the pre-trained artificial intelligence means 121 of the server.
[0032] Next, the operation of the artificial intelligence system 100 installed with the program for extending the functions of the browser 111 will be described in detail. As shown in FIG. 2, in step S1, as a talk screen display determination step, the user terminal 110 installed with the program for extending the functions of the browser 111 accesses the first server 120, and the browser 111 of the user terminal 110 determines whether it is about to display the original talk screen TK (see FIG. 3(A)). That is, when the user terminal 110 sends an http request to the first server 120 and receives an http response from the first server 120, it is determined that the original talk screen TK (see FIG. 3(A)) is about to be displayed. If it is determined that the talk screen TK is about to be displayed, the process proceeds to step S2. On the other hand, if it is determined that it is not yet, step S1 is repeated.
[0033] In step S2, as a voice input icon display step, the browser 111 of the user terminal 110 displays a voice input icon button 112b in the input field 112a of the talk screen 112 that displays the conversation between the user of the user terminal 110 and ChatGPT, which is an example of the pre-trained artificial intelligence means 121 of the server, on the talk screen 112. Here, in FIG. 3(A), the state where the browser 111 of the user terminal 110 in which the program for extending the functions of the browser 111 is not installed displays the original talk screen TK is shown. The voice input icon button 112b is not displayed on the original talk screen TK, and the original talk screen TK does not support voice input.
[0034] In contrast, as shown in FIG. 3(B), when the browser 111 of the user terminal 110 installed with the program for extending the functions of the browser 111 of the present invention displays the talk screen TK based on the http response from the first server 120, a talk screen 112 obtained by extending the functions of the original talk screen TK is displayed. More specifically, as shown in FIG. 3(B), the browser 111 of the user terminal 110 displays a voice input icon button 112b in the input field 112a as an example of the talk screen 112 on the talk screen 112. When the program for extending the function of the browser 111 is installed and pinned in the display settings and becomes active, as an example, a function extension icon 111a is displayed on the browser 111.
[0035] In step S3, as a voice input icon operation determination step, the browser 111 determines whether or not the voice input icon button 112b has been operated. More specifically, as shown in FIG. 3(B), it is determined whether or not the voice input icon button 112b has been clicked by the user's mouse operation. If it is determined that it has been operated, the process proceeds to step S4. On the other hand, if it is determined that it has not been operated yet, step S3 is repeated.
[0036] In step S4, as a microphone activation step, the browser 111 activates the microphone of the user terminal 110. Note that the microphone is not limited to the one provided in the main body of the user terminal 110, and may be one connected to the user terminal 110. In step S5, as a voice input presence determination step, the browser 111 determines whether or not the user's voice has been input via the microphone. If it is determined that there is voice input, the process proceeds to step S6. On the other hand, if it is determined that there is still no voice input, step S5 is repeated.
[0037] In step S6, as a voice input text conversion and transmission step, the browser 111 converts the input voice data into text data TX. Then, the browser 111 inputs the converted text data TX into the input field 112a. At the same time, the browser 111 transmits the converted text data TX from the user terminal 110 to ChatGPT, which is an example of the pre-trained artificial intelligence means 121 of the first server 120.
[0038] As a result, even in the talk screen TK that originally did not support voice input, by using the program for extending the function of the browser 111, a voice input icon button 112b is displayed in the input field 112a as an example of the talk screen 112. When the voice input icon button 112b is operated and voice input is performed, the voice content is converted into text data TX, input into the input field 112a, and then transmitted. As a result, the user can communicate with ChatGPT, which is an example of the pre-trained artificial intelligence means 121, by voice input without typing characters.
[0039] Here, in FIG. 4(A), a state is shown in which the original talk screen TK of the browser 111 of the user terminal 110 in which the program for extending the function of the browser 111 is not installed displays the communication between the user and ChatGPT, which is an example of the pre-trained artificial intelligence means 121. In the original talk screen TK, a voice output icon button 112c is not displayed at the position of the answer from ChatGPT, which is an example of the pre-trained artificial intelligence means 121.
[0040] Furthermore, in this embodiment, as shown in FIG. 4(B), when the browser 111 of the user terminal 110 in which the program for extending the function of the browser 111 is installed displays the generated text data GT generated by ChatGPT, which is an example of the pre-trained artificial intelligence means 121, received from the first server 120 on the talk screen 112, a voice output icon button 112c is configured to be displayed corresponding to the generated text data GT.
[0041] More specifically, as the voice output icon display step, when the browser 111 of the user terminal 110 displays the generated text data GT generated by ChatGPT, which is an example of the pre-trained artificial intelligence means 121 received from the first server 120, on the talk screen 112, a voice output icon button 112c is displayed corresponding to the generated text data GT. And when the voice output icon button 112c is operated, as the voice output step, the content of the generated text data GT corresponding to the voice output icon button 112c is output as voice through the speaker of the user terminal 110.
[0042] As a result, even on the talk screen TK that originally did not support voice output, by using the function extension program of the browser 111, a voice output icon button 112c is displayed corresponding to the generated text data GT on the talk screen 112. And when the voice output icon button 112c is operated, the content of the corresponding generated text data GT is output as voice. As a result, the user can understand the answer from ChatGPT, which is an example of the pre-trained artificial intelligence means 121, by voice output without reading the generated text data GT with their eyes.
[0043] Also, in this embodiment, as shown in FIG. 5(A), the browser 111 of the user terminal 110 is configured to display an instruction command list display icon button 112d on the talk screen 112. More specifically, as the instruction command icon display step, the browser 111 of the user terminal 110 displays an instruction command list display icon button 112d on the talk screen 112.
[0044] And when the instruction command list display icon button 112d is operated, as the prompt list display step, the browser 111 displays a prompt list display window 113 (display column) that displays a personal fixed instruction command information item 113c, which is an example of a fixed instruction command information item registered in advance, in a list format. Furthermore, as shown in FIG. 5(B), when the instruction command information PT in the prompt list display window 113 is selected and operated, as a prompt transmission step, the selected instruction command information PT is input into the input field 112a and transmitted from the user terminal 110 to ChatGPT, which is an example of the pre-trained artificial intelligence means 121 of the first server 120.
[0045] As a result, even for a talk screen TK that does not correspond to a prompt that is originally a fixed instruction command information PT, the prompt can be preset and registered by using the function extension program of the browser 111. Then, the prompt list display window 113 is displayed by operating the instruction command list display icon button 112d. Furthermore, when the desired instruction command information PT is selected from among them, it is input into the input field 112a and transmitted to ChatGPT, which is an example of the pre-trained artificial intelligence means 121. As a result, the user can easily communicate with ChatGPT, which is an example of the pre-trained artificial intelligence means 121, using the prompt and obtain an answer in the desired format and number of characters obtained from ChatGPT, which is an example of the pre-trained artificial intelligence means 121. For example, if a limit of 300 characters or less is set as a condition for the answer and transmitted, an answer within 300 characters can be obtained. In addition, if conditions (restrictions) such as a table format, a markdown format, an email format that can be used as it is for email, and a report format that can be used as it is for a report are set as conditions for the answer and transmitted, an answer in the format along with the conditions can be obtained.
[0046] Note that the prompt list display window 113 is provided with a "personal use" tab 113a and a "shared" tab 113b. When the "personal use" tab 113a is selected, personal fixed instruction command information items 113c are displayed in the prompt list display window 113. Regarding the personal fixed instruction command information item 113c, when the "New Prompt Addition" button is operated, the user can input and set-register a prompt by themselves. And the personal fixed instruction command information item 113c is provided so that the prompt items registered by the user are displayed. On the other hand, when the "Share" tab 113b is selected, it is provided so that the shared fixed instruction command information item 113d, which will be described later, is displayed. The shared fixed instruction command information item 113d is provided so that the items of the shared instruction command information PT, which is a shared prompt registered by the administrator, are displayed.
[0047] Furthermore, in this embodiment, as an example, the history data LG of the interaction between the user terminal 110 and ChatGPT, which is an example of the pre-trained artificial intelligence means 121, on the talk screen 112 of the user terminal 110 is transmitted from the user terminal 110 or the first server 120 to the second server 130. And the second server 130 as a server is configured to store the history data LG of the interaction between the user terminal 110 and ChatGPT, which is an example of the pre-trained artificial intelligence means 121, on the talk screen 112 of the user terminal 110. More specifically, as a log storage step, the second server 130 as a server stores the history data LG of the interaction between the user terminal 110 and ChatGPT, which is an example of the pre-trained artificial intelligence means 121, on the talk screen 112 of the user terminal 110. And as shown in FIG. 6, as a history display step, the browser 141 of the administrator terminal 140 that can communicate with the second server 130 displays the administrator screen 142 based on the operation of the administrator.
[0048] The history data LG of the interaction between the user terminal 110 and ChatGPT, which is an example of the pre-trained artificial intelligence means 121, on the talk screen 112 of the user terminal 110 is displayed on the administrator screen 142. As shown in FIG. 6, when an arbitrary log location is selected by an administrator through a double - click or the like, the browser 141 opens the log detail display window 143. The log detail display window 143 describes details of the transaction history data LG of the selected log.
[0049] For example, access date and time, user ID, result, action type, type of the model (engine) of ChatGPT which is an example of the pre - learned artificial intelligence means 121, number of input characters, input text, response time of ChatGPT which is an example of the pre - learned artificial intelligence means 121, user evaluation of the answer, and detailed information about the response text from ChatGPT which is an example of the pre - learned artificial intelligence means 121 are described.
[0050] As a result, the log of the interaction between the user terminal 110 and ChatGPT which is an example of the pre - learned artificial intelligence means 121 is saved and displayed on the administrator screen 142 of the administrator terminal 140. As a result, the administrator can manage the user's use from the security perspective.
[0051] Also, in this embodiment, as shown in FIG. 7(A), the administrator screen 142 is provided so that the prohibited word BW or the prohibited attribute BA can be freely set. More specifically, when there is a setting operation for the prohibited word BW or the prohibited attribute BA on the administrator screen 142, as the prohibited input setting step, the setting of the prohibited word BW or the prohibited attribute BA is registered in the second server 130 which is an example of a server. And, as the prohibited input determination step, the browser 111 of the user terminal 110 determines whether the content of the text data TX input in the input field 112a of the talk screen 112 of the user terminal 110 is the preset prohibited word BW or prohibited attribute BA. For example, assume that the user inputs "03 - 1234 - 5678" as an example of the phone number of the set prohibited attribute BA by voice input or direct text input. Then, the text data TX of "03-1234-5678" is input into the input field 112a of the chat screen 112. In this case, when the content of the text data TX input in the input field 112a is the preset prohibited attribute BA, the browser 111 of the user terminal 110 determines this.
[0052] And when the content of the text data TX input in the input field 112a is the preset prohibited word BW or the prohibited attribute BA, as the transmission operation stop and warning display step, the transmission operation of the input text data TX is invalidated, and a warning indicating that the prohibited word BW or the prohibited attribute BA has been input is displayed. More specifically, as shown in FIG. 7(B), when the content of the text data TX input in the input field 112a is the preset prohibited word BW or the prohibited attribute BA, even if the user operates the transmission icon button SD as an example of the transmission operation, the transmission icon button SD is configured not to react. At the same time, the browser 111 displays a message "Inappropriate expression is included" as an example of the warning display 114 for inappropriate input.
[0053] As a result, when a telephone number or an email address is input into the input field 112a as an example of the prohibited attribute BA by the preset prohibited word BW or the prohibited attribute BA on the administrator screen 142, the transmission operation becomes invalid, and the interaction between the user and ChatGPT, which is an example of the pre-trained artificial intelligence means 121, does not proceed and stops. That is, the text data TX input in the input field 112a is not sent to ChatGPT, which is an example of the pre-trained artificial intelligence means 121. As a result, it is possible to avoid the pre-trained artificial intelligence means 121, such as ChatGPT, from learning personal information and the like. Furthermore, it is also possible to avoid personal information from leaking to other terminals via ChatGPT, which is an example of the pre-trained artificial intelligence means 121. Also, as an example of the prohibited attribute BA, when a telephone number or email address is entered in the input field 112a, a warning is displayed. As a result, the user can notice that they have entered personal information or the like.
[0054] Furthermore, in this embodiment, as shown in FIG. 8(A), on the administrator screen 142, a setting for IP address restriction is provided so that it can be freely set. More specifically, when there is a setting operation for IP address restriction on the administrator screen 142, as the IP address restriction setting step, the setting for IP address restriction is registered in a second server 130 which is an example of a server. And when the browser 111 of the user terminal 110 accesses the first server 120 in order for the user terminal 110 to display the talk screen 112, as the IP address condition determination step, the browser 111 determines whether the IP address information AD of the user terminal 110 included in the request information from the user terminal 110 satisfies the preset IP address restriction setting conditions.
[0055] Here, when the browser 111 makes a determination, in order to refer to the setting for IP address restriction, the user terminal 110 communicates not only with the first server 120 but also with the second server 130. And as shown in FIG. 3(B) described above, when the IP address information AD of the user terminal 110 satisfies the preset IP address restriction setting conditions, as the talk screen display step, the first server 120 transmits the first response information to the user terminal 110 and the browser 111 of the user terminal 110 displays the talk screen 112. On the other hand, as shown in FIG. 8(B), when the IP address information AD of the user terminal 110 does not satisfy the preset IP address restriction setting conditions, as the NG display step, the second server 130 transmits second response information different from the first response information to the user terminal 110 and causes the browser 111 to display that the user's IP address information AD does not satisfy the IP address restriction setting conditions.
[0056] More specifically, the browser 111 displays, as an example of a warning display 115 indicating that the IP address restriction condition is not satisfied, a message such as "The IP address is restricted. Please access again using the set IP address!" At this time, the browser 111 may or may not display the talk screen 112 itself. The reason is that even if it is displayed, the use of the talk screen 112 is restricted and it does not need to be available to the user.
[0057] As a result, due to the IP address restriction set in advance on the administrator screen 142, for example, only connections in a safe Internet environment are permitted, and connections in an Internet environment with unknown security are blocked. That is, the use on the talk screen 112 is restricted and further communication is blocked. As a result, the administrator can easily take security measures regarding the user's use.
[0058] Note that, as shown in FIG. 8(A), the administrator screen 142 is provided so that the restriction of users who can use the talk screen 112 can be freely set. As a result, only the users permitted by the administrator can communicate freely with ChatGPT, which is an example of pre-trained artificial intelligence means 121, on the talk screen 112. As a result, the administrator can easily take security measures regarding the user's use.
[0059] Here, in FIGS. 3(A) and 4(A), in the original talk screen TK of the browser 111 of the user terminal 110 in which the function expansion program of the browser 111 is not installed, the mail sending icon button 112e is not displayed. And in this embodiment, as shown in FIG. 9(A), the browser 111 of the user terminal 110 installed with the program for extending the functions of the browser 111 is configured to display a mail sending icon button 112e on the talk screen 112.
[0060] More specifically, as the mail sending icon display step, when the browser 111 of the user terminal 110 attempts to display the talk screen TK based on the http response from the first server 120, the mail sending icon button 112e is displayed on the talk screen 112. Furthermore, as the mail sending icon operation determination step, the browser 111 determines whether the mail sending icon button 112e has been operated. Here, it is assumed that the user clicks the mail sending icon button 112e as an example. Then, the browser 111 determines that the mail sending icon button 112e has been operated.
[0061] And when the mail sending icon button 112e is operated, as the mail sending step, the second server 130 as a server sends the content of the conversation between ChatGPT, which is an example of the pre-trained artificial intelligence means 121 displayed on the talk screen 112, and the user of the user terminal 110 to the email address of the user set in advance. At this time, the second server 130 receives from the user terminal 110 the content of the conversation between ChatGPT, which is an example of the pre-trained artificial intelligence means 121 displayed on the talk screen 112 of the user terminal 110, and the user of the user terminal 110. And it is sent to the user as an example to the email address of the user set in advance. Also, together with this, the browser 111 displays a message indicating that the mail has been sent, such as "The mail has been sent."
[0062] As a result, even on the talk screen TK which originally did not support email sending, by using the program for extending the functions of the browser 111, an email sending icon button 112e is displayed on the talk screen 112. When the email sending icon button 112e is operated, the content of the conversation displayed on the talk screen 112 is sent to the email address registered in the user settings. As a result, the user can leave a log as an email and can also share it with other users as needed. More specifically, as shown in FIG. 9(B), when the receiving email screen 117 is displayed as an email browser in the browser 111 of the user terminal 110, an email ML including the data of the content of the conversation sent from the second server 130 is displayed. The email ML including the data of the content of this conversation can be left as a log.
[0063] Here, in FIGS. 3(A) and 4(A), in the original talk screen TK of the browser 111 of the user terminal 110 in which the program for extending the functions of the browser 111 is not installed, the interface is set so that it cannot be changed. On the other hand, in this embodiment, as shown in FIG. (A), in the talk screen 112, the interface is provided so that it can be freely set. More specifically, as an interface setting operation presence / absence determination step, in the talk screen 112, the browser 111 determines whether or not there has been a predetermined interface setting operation. When there has been a predetermined interface setting operation, as an interface setting window display step, an interface setting window 119 in which the color of characters and the color of the background can be freely selected is displayed.
[0064] As an example of a predetermined interface setting operation, when the "?" mark button 112f on the talk screen 112 is operated, the browser 111 displays a menu item window 118. The menu item window 118 is provided with items such as "Help", "Release Notes", "Keyboard Shortcuts", and "UI Settings" as an example. And, as an example of an operation for a predetermined interface setting, when "UI Settings" in the menu item window 118 is selected, the browser 111 displays an interface setting window 119. Here, the interface setting window 119 is provided so that the text color and background color of the talk screen 112 can be freely selected and set respectively.
[0065] Subsequently, as a step for determining the presence or absence of a selection / setting save operation, it is determined whether there is a selection operation for the color and a setting save operation in the interface setting window 119. More specifically, as a selection operation in the interface setting window 119, the user selects a desired color. And, as an example of a setting save operation, when the "Save" button is operated, the selected color setting information is saved in the second server 130. As shown in FIG. 10(B), when there is a selection operation and a setting save operation, as a color reflection step, the browser 111 displays the text color and background color on the talk screen 112 based on the selected content. At this time, the browser 111 displays the text color and background color on the talk screen 112 based on the color setting information saved in the second server 130 as an example.
[0066] As a result, even if the color of the talk screen TK cannot be changed originally, by using the function expansion program of the browser 111, the color of the talk screen 112 can be freely changed by the user's setting operations (predetermined interface setting operation, selection operation, setting save operation). As a result, it can be changed to a color that is easy for the user to see. For example, when a user has characteristics such as color blindness or difficulty seeing specific colors, by changing the settings to colors that are easier for the user to see, the user can make the content of the conversation on the talk screen 112 easier for themselves to view.
[0067] Also, in this embodiment, as shown in FIG. 11(A), on the administrator screen 142, settings for the shared standard instruction command information PT can be freely set. More specifically, when there is a setting operation for the shared standard instruction command information PT on the administrator screen 142, as the shared prompt setting step, the setting for the shared standard instruction command information PT is registered in the second server 130 as the server. Then, as shown in FIG. 11(B), when the instruction command list display icon button 112d is operated on the talk screen 112, as the prompt list display step, the browser 111 displays a prompt list display window 113 that can freely switch and display in list form the content of the personal standard instruction command information item 113c set and registered in advance and the content of the shared standard instruction command information item 113d set and registered in advance. At this time, based on the setting information for the shared standard instruction command information PT registered in the second server 130, the browser 111 displays the content of the shared standard instruction command information item 113d in the prompt list display window 113. Furthermore, when the shared standard instruction command information PT in the prompt list display window 113 is selected and operated, as the shared prompt transmission step, the browser 111 inputs the selected shared standard instruction command information PT into the input field 112a and transmits it from the user terminal 110 to ChatGPT, which is an example of the pre-trained artificial intelligence means 121 of the server.
[0068] Thereby, even for a talk screen TK that does not originally correspond to the prompt that is the shared standard instruction command information PT, by using the function expansion program of the browser 111, the shared prompt can be freely set and registered in advance by the administrator. Then, the prompt list display window 113 is displayed by operating the instruction command list display icon button 112d. Furthermore, when a desired shared standard instruction command information PT is selected from among them, it is input into the input field 112a and sent to ChatGPT, which is an example of the pre-trained artificial intelligence means 121. As a result, the user can easily communicate with ChatGPT, which is an example of the pre-trained artificial intelligence means 121, using the shared prompt and obtain a response in the desired format and word count from ChatGPT, which is an example of the pre-trained artificial intelligence means 121. That is, a list of shared prompts pre-registered by the administrator is displayed in the prompt list display window 113 without the user having to register the prompt themselves. As a result, the user can easily communicate with ChatGPT, which is an example of the pre-trained artificial intelligence means 121, simply by selecting a desired shared prompt from among them and obtain a response in the desired format and word count from ChatGPT, which is an example of the pre-trained artificial intelligence means 121.
[0069] Note that the browser 111 of the user terminal 110 installed with the program for extending the function of the browser 111 may be configured to display the talk sharing icon button 112g on the talk screen 112. When the talk sharing icon button 112g is operated, the second server 130 generates shared URL information having the content of the conversation on the talk screen 112. Then, the second server 130 sends it to the user terminal 110. Then, the browser 111 of the user terminal 110 displays the received shared URL information on the talk screen 112. When the user clicks on the displayed shared URL information, the shared URL information is temporarily saved in the clipboard. The user can send the shared URL information to other users in an email or a chat screen to share the content of the conversation on the talk screen 112 that they have had with other users.
[0070] The program for extending the functions of the browser 111, which is an embodiment of the present invention obtained in this way, enables the browser 111 of the user terminal 110 to display a voice input icon button 112b on a talk screen 112 that displays the conversation between the user of the user terminal 110 and ChatGPT, which is an example of the pre-trained artificial intelligence means 121 of the server, in a voice input icon display step S2, a voice input icon operation determination step S3 for determining whether the voice input icon button 112b has been operated, a microphone activation step S4 for activating the microphone of the user terminal 110 when the voice input icon button 112b is operated, and when voice data of the user is input via the microphone, converting the input voice data into text data TX, inputting the converted text data TX into the input field 112a, and transmitting it from the user terminal 110 to ChatGPT, which is an example of the pre-trained artificial intelligence means 121 of the server, in voice input text conversion and transmission steps S5 and S6. Thus, even on a talk screen TK that does not originally support voice input, by using the program for extending the functions of the browser 111, the user can communicate with ChatGPT, which is an example of the pre-trained artificial intelligence means 121, by voice input without typing text.
[0071] Furthermore, when the browser 111 of the user terminal 110 displays the generated text data GT generated by ChatGPT, which is an example of the pre-trained artificial intelligence means 121 received from the server, on the talk screen 112, there are a voice output icon display step of displaying a voice output icon button 112c corresponding to the generated text data GT, and a voice output step of outputting the content of the generated text data GT corresponding to the voice output icon button 112c through the speaker of the user terminal 110 when the voice output icon button 112c is operated. By having these steps, even on the talk screen TK that originally did not support voice output, by using the function expansion program of the browser 111, the user can understand the answer from ChatGPT, which is an example of the pre-trained artificial intelligence means 121, through voice output without reading the generated text data GT visually.
[0072] Also, the browser 111 of the user terminal 110 has an instruction command icon display step of displaying an instruction command list display icon button 112d on the talk screen 112, a prompt list display step of displaying a prompt list display window 113 that displays, in list form, a personal fixed instruction information item 113c, which is an example of a fixed instruction information item set and registered in advance, when the instruction command list display icon button 112d is operated, and a prompt transmission step of inputting the selected instruction information PT into the input field 112a and transmitting it from the user terminal 110 to ChatGPT, which is an example of the pre-trained artificial intelligence means 121 of the server, when the instruction information PT in the prompt list display window 113 is selected. By having these steps, even on the talk screen TK that originally did not support the prompt of the fixed instruction information PT, by using the function expansion program of the browser 111, the user can easily communicate with ChatGPT, which is an example of the pre-trained artificial intelligence means 121, using the prompt and obtain the answer in the desired format and word count from ChatGPT, which is an example of the pre-trained artificial intelligence means 121.
[0073] Furthermore, the server has a log storage step of storing the history data LG of the interaction between the user terminal 110 and ChatGPT, which is an example of the pre-trained artificial intelligence means 121, on the chat screen 112 of the user terminal 110, and a history display step in which the browser 141 of the administrator terminal 140 capable of communicating with the server displays the history data LG of the interaction between the user terminal 110 and ChatGPT, which is an example of the pre-trained artificial intelligence means 121, on the chat screen 112 of the user terminal 110 based on the operation of the administrator. Thus, the administrator can manage the user's use from the perspective of security.
[0074] Also, when there is a setting operation for the prohibited word BW or the prohibited attribute BA on the administrator screen 142, a prohibited input setting step of registering the setting of the prohibited word BW or the prohibited attribute BA in the server, a prohibited input determination step of determining whether the content of the text data TX input in the input field 112a of the chat screen 112 of the user terminal 110 is the preset prohibited word BW or prohibited attribute BA, and when the content of the text data TX input in the input field 112a is the preset prohibited word BW or prohibited attribute BA, a transmission operation stop and warning display step of invalidating the transmission operation of the input text data TX and displaying a warning indicating that the prohibited word BW or the prohibited attribute BA has been input. Thus, it is possible to avoid the pre-trained artificial intelligence means 121, such as ChatGPT, from learning personal information and the like, and it is also possible to avoid personal information from leaking to other terminals via the pre-trained artificial intelligence means 121, such as ChatGPT, and the user can be aware that personal information and the like have been input.
[0075] Furthermore, when there is a setting operation for IP address restriction on the administrator screen 142, an IP address restriction setting step for registering the IP address restriction setting on the server, and when the user terminal 110 accesses the server for the browser 111 of the user terminal 110 to display the talk screen 112, an IP address condition determination step for determining whether the IP address information AD of the user terminal 110 included in the request information from the user terminal 110 satisfies the conditions of the preset IP address restriction setting, and when the IP address information AD of the user terminal 110 satisfies the conditions of the preset IP address restriction setting, a talk screen display step in which the server sends the first response information to the user terminal 110 and the browser 111 of the user terminal 110 displays the talk screen 112, and when the IP address information AD of the user terminal 110 does not satisfy the conditions of the preset IP address restriction setting, an NG display step in which the server sends second response information different from the first response information to the user terminal 110 and causes the browser 111 to display that the user's IP address information AD does not satisfy the conditions of the IP address restriction setting. By further having these steps, the administrator can easily take security measures regarding user usage.
[0076] Also, on the talk screen 112, a mail sending icon display step for displaying the mail sending icon button 112e, a mail sending icon operation determination step for determining whether the mail sending icon button 112e has been operated, and when the mail sending icon button 112e has been operated, a mail sending step in which the server sends the content of the conversation between the ChatGPT, which is an example of the pre-trained artificial intelligence means 121 displayed on the talk screen 112, and the user of the user terminal 110 to the electronic mail address of the user setting registered in advance. By further having these steps, even on the talk screen TK that does not originally support mail sending, by using the function expansion program of the browser 111, the user can leave a log as an electronic mail and can also share it with other users if necessary.
[0077] Furthermore, on the talk screen 112, there are further steps including an interface setting operation presence / absence determination step for determining whether there has been a predetermined interface setting operation, an interface setting window display step for displaying an interface setting window 119 in which the colors of characters and the background can be freely selected when there has been a predetermined interface setting operation, a selection / setting save operation presence / absence determination step for determining whether there has been a selection operation and a setting save operation in the interface setting window 119, and a color reflection step for displaying the colors of characters and the background on the talk screen 112 based on the selected content when there has been a selection operation and a setting save operation. By having these steps, even if the color of the original talk screen TK cannot be changed, it can be changed to a user-friendly color by using the program for extending the functions of the browser 111. For example, when the user has characteristics such as color weakness or difficulty seeing specific colors, by changing the setting to a user-friendly color, the user can make the content of the conversation on the talk screen 112 easier to see for themselves.
[0078] Also, when there is a setting operation for the shared standard instruction command information PT on the administrator screen 142, a shared prompt setting step for registering the setting of the shared standard instruction command information PT on the server, and when the instruction command list display icon button 112d is operated on the talk screen 112, the content of the personal standard instruction command information item 113c set and registered in advance and the content of the shared standard instruction command information item 113d set and registered in advance are displayed in a list format in a switchable manner. A prompt list display step of displaying a prompt list display window 113, and when the shared standard instruction command information PT in the prompt list display window 113 is selected and operated, the selected shared standard instruction command information PT is input into the input field 112a, and at the same time, it is transmitted from the user terminal 110 to ChatGPT, which is an example of the pre-trained artificial intelligence means 121 of the server. By further having a shared prompt transmission step, even on a talk screen TK that does not originally correspond to a prompt that is the shared standard instruction command information PT, by using the program for extending the function of the browser 111, the user can easily select a desired shared prompt from the list of shared prompts pre-registered by the administrator without registering the prompt by himself / herself, and communicate with ChatGPT, which is an example of the pre-trained artificial intelligence means 121, and obtain a response in the desired format and word count from ChatGPT, which is an example of the pre-trained artificial intelligence means 121. The effect is very significant.
Explanation of Signs
[0079] 100 ··· Artificial intelligence system 110 ··· User terminal 111 ··· (Of the user terminal) Browser 111a··· Function extension icon 112 ··· (After the function extension of the present invention) Talk screen 112a··· Input field 112b··· Voice input icon button 112c··· Voice output icon button 112d ··· Indicator command list display icon button 112e ··· Email sending icon button 112f ··· "?" Mark button 112g ··· Talk sharing icon button 113 ··· Prompt list display window (display column) 113a ··· "Personal" tab 113b ··· "Shared" tab 113c ··· (Personal) Fixed indicator command information item 113d ··· (Shared) Fixed indicator command information item 114 ··· Warning display for inappropriate input 115 ··· Warning display for not meeting the IP address restriction conditions 116 ··· Display indicating that the email has been sent 117 ··· Received email screen 118 ··· Menu item window 119 ··· Interface setting window 120 ··· First server 121 ··· Pre-trained artificial intelligence means (ChatGPT) 130 ··· Second server 131 ··· Database 140 ··· Administrator terminal 141 ··· (Administrator terminal) Browser 142 ··· Administrator screen 143 ··· Log details display window TK ··· (Chat screen before function expansion as a comparison target) ML ··· Email containing data on the content of the conversation TX ··· Text data GT ··· Generated text data PT ··· Indicator command information LG ··· Conversation history data BW ··· Prohibited words BA ··· Prohibited attributes AD ··· IP address information SD ··· Transmission icon button
Claims
1. A program for extending the functionality of a browser that displays the interaction between a user terminal and pre-trained artificial intelligence means of a server, in a chat screen where the browser of the user terminal displays the interaction between the pre-trained artificial intelligence means of the server and the user of the user terminal, an audio input icon display step of displaying an audio input icon button on the chat screen; an audio input icon operation determination step of determining whether or not the audio input icon button has been operated; a microphone activation step of activating the microphone of the user terminal when the audio input icon button is operated; and an audio input text conversion and transmission step of converting the input audio data into text data when the user's voice is input via the microphone, inputting the converted text data into an input field, and transmitting it from the user terminal to the pre-trained artificial intelligence means of the server. A program for extending the functionality of a browser, characterized by having these steps.
2. when the browser of the user terminal displays the generated text data generated by the pre-trained artificial intelligence means received from the server on the chat screen, an audio output icon display step of displaying an audio output icon button corresponding to the generated text data; and an audio output step of outputting the content of the generated text data corresponding to the audio output icon button as audio via the speaker of the user terminal when the audio output icon button is operated. The program for extending the functionality of a browser according to Claim 1, further characterized by having these steps.
3. an instruction command icon display step of displaying an instruction command list display icon button on the chat screen by the browser of the user terminal; a prompt list display step of displaying a prompt list display window that displays preset and registered standard instruction command information items in a list format when the instruction command list display icon button is operated; and a prompt transmission step of inputting the selected instruction command information into an input field and transmitting it from the user terminal to the pre-trained artificial intelligence means of the server when the instruction command information in the prompt list display window is selected. The program for extending the functionality of a browser according to Claim 2, further characterized by having these steps.
4. A log storage step in which the server stores history data of the interaction between the user terminal and the pre-trained artificial intelligence means on the talk screen of the user terminal; The browser of the administrator terminal capable of communicating with the server further has a history display step in which, based on the operation of the administrator, the history data of the interaction between the user terminal and the pre-trained artificial intelligence means on the talk screen of the user terminal is displayed on the administrator screen. The program for extending the function of the browser according to any one of claims 1 to 3, characterized in that.
5. A prohibited input setting step in which, when there is a setting operation for a prohibited word or prohibited attribute on the administrator screen, the setting for the prohibited word or prohibited attribute is registered in the server; A prohibited input determination step of determining whether the content of the text data input in the input field of the talk screen of the user terminal is a preset prohibited word or prohibited attribute; When the content of the text data input in the input field is a preset prohibited word or prohibited attribute, the program for extending the function of the browser according to claim 4, further comprising a transmission operation stop / warning display step of invalidating the transmission operation of the input text data and displaying a warning indicating that the prohibited word or prohibited attribute has been input.
6. An IP address restriction setting step in which, when there is a setting operation for IP address restriction on the administrator screen, the setting for IP address restriction is registered in the server; An IP address condition determination step of determining whether the IP address information of the user terminal included in the request information from the user terminal satisfies the conditions of the preset IP address restriction setting when the user terminal accesses the server in order for the browser of the user terminal to display the talk screen; A talk screen display step in which, when the IP address information of the user terminal satisfies the conditions of the preset IP address restriction setting, the server transmits first response information to the user terminal and the browser of the user terminal displays the talk screen. When the IP address information of the user terminal does not meet the conditions of the preset IP address restriction setting, the server sends second response information different from the first response information to the user terminal, and has an NG display step of causing the browser to display that the IP address information of the user does not meet the conditions of the IP address restriction setting. The program for extending the function of the browser according to claim 5, characterized in that it further has this.
7. In the chat screen, an email sending icon display step of displaying an email sending icon button, and An email sending icon operation determination step of determining whether or not the email sending icon button has been operated, and When the email sending icon button is operated, the server further has an email sending step of sending the content of the conversation between the pre-trained artificial intelligence means displayed on the chat screen and the user of the user terminal to the email address of the user setting registered in advance. The program for extending the function of the browser according to claim 6, characterized in that it has this.
8. In the chat screen, an interface setting operation presence determination step of determining whether or not there has been a predetermined interface setting operation, and When there has been the predetermined interface setting operation, an interface setting window display step of displaying an interface setting window in which the color of characters and the color of the background can be freely selected, and A selection / setting save operation presence determination step of determining whether or not there has been a selection operation and a setting save operation in the interface setting window, and When there has been the selection operation and the setting save operation, the program for extending the function of the browser according to claim 7, characterized in that it further has a color reflection step of displaying the color of characters and the color of the background on the chat screen based on the selection content.
9. When there is a setting operation for the shared standard instruction command information on the administrator screen, a shared prompt setting step of registering the setting for the shared standard instruction command information in the server, and When the instruction command list display icon button is operated on the chat screen, a prompt list display step of displaying a prompt list display window that can freely switch between the content of the personal standard instruction command information items set and registered in advance and the content of the shared standard instruction command information items set and registered in advance in a list format. When the shared standard instruction information in the prompt list display window is selected, a shared prompt transmission step of inputting the selected shared standard instruction information into the input field and transmitting it from the user terminal to the pre-trained artificial intelligence means of the server is further included, and the program for extending the function of the browser according to claim 8 is characterized by this.
Citation Information
Patent Citations
Information processing apparatus, information processing method, and program for information processing
JP2023073095A