Recommendation system, recommendation system program

Abstract

We provide a recommendation system that allows users to understand how to enjoy and what to see based on the negative factors and reasons for the recommendation. [Solution] The recommendation system 100 comprises a user terminal 110 and a server 120, and when a predetermined operation is performed to request a recommendation, the recommendation means uses pre-trained artificial intelligence means to generate a sentence about the activity theme based on the user's attribute information, current user situation information, and external environment information, selects a destination and means of transportation that are in line with the sentence about the activity theme from the server's database 121, evaluates negative elements from the destination information, travel route information by means of transportation, and information used to generate the sentence about the activity theme, uses pre-trained artificial intelligence means to generate a reason explanation text that includes the negative elements and changes them from negative to positive, and displays the generated reason explanation text RE, destination information DN, and travel route information RT on the user terminal.

Description

[Technical Field]

[0001] The present invention relates to a recommendation system that includes a user terminal and a server and recommends destinations to users, and a program for the system. [Background technology]

[0002] Conventionally, a recommendation system is known that takes into consideration the preferences of a user and makes suggestions suited to the user's current situation (for example, Patent Document 1). [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Publication No. 2022-129411 Summary of the Invention [Problem to be solved by the invention]

[0004] However, the conventional recommendation systems mentioned above were only configured to present destinations and means of transportation that took user preferences into consideration, and therefore did not reflect negative factors such as rain, congestion, and inconvenience in the recommendations. As a result, when users actually acted on the recommendations, they felt a gap between the recommendations and the actual negative factors, resulting in low user trust and empathy for the recommendations.

[0005] Therefore, the present invention solves the problems of the prior art as described above. In other words, the object of the present invention is to provide a recommendation system and a program therefor that enable users to easily understand, trust, and empathize with the reasons for the recommendation, including how to enjoy and highlight the negative elements, from the explanation text for the reason for the recommendation. [Means for solving the problem]

[0006] The invention according to claim 1 is a recommendation system that includes a user terminal and a server and recommends destinations to a user, and when a predetermined operation for requesting a recommendation is performed on the user terminal, a recommendation means in the user terminal or the server recommends a destination based on pre-inputted user attribute information, inputted or acquired current user situation information, and, Acquired external environment information At least the user's attribute information entered in advance Based on the above, a sentence about the action theme is generated using a pre-trained artificial intelligence means, a destination and means of transportation that are in line with the sentence about the action theme are selected from the server's database, negative elements are evaluated from the selected destination information and travel route information using the travel route, as well as the information used to generate the sentence about the action theme, and when a reason explanation text recommending the selected destination and travel route is generated using the pre-trained artificial intelligence means based on the selected destination information and travel route information using the travel route, the reason explanation text is generated based on a prompt to generate a reason explanation text that includes negative elements and changes from negative elements to positive elements, and the generated reason explanation text, destination information, and travel route information are displayed on the user terminal, thereby solving the above-mentioned problem.

[0007] The invention according to claim 2 is the configuration of the recommendation system according to claim 1, in addition to the configuration of the recommendation system according to claim 1, wherein the recommendation means is configured to use pre-input user attribute information, input or acquired current user situation information, and, Acquired external environment information At least the external environmental information obtained The system further solves the above-mentioned problem by evaluating any of the following as negative factors: weather information indicating bad weather; information on the selected destination and travel route by the selected means of transportation indicating that the waiting time at the destination will be longer than a specified time; that there are restrictions on payment methods; that construction work is being carried out and the appearance will not be as good as usual; or that the travel route includes a slope or staircase that must be climbed for a specified distance or more; and generating a reason explanation text that includes the evaluated negative factors.

[0008] The invention of claim 3 further solves the above-mentioned problem by being configured such that, in addition to the configuration of the recommendation system described in claim 2, the recommendation means selects multiple sets of destinations and means of transportation that are in line with the sentence on the activity theme from a database on a server, and when the destination information and travel route information by the multiple selected sets are displayed on the user terminal based on the destination information and travel route information by the travel means, a destination icon is placed on a map and a predetermined mark icon is displayed in a position adjacent to the destination icon, and when the predetermined mark icon is operated, the content of the flow that changes from negative elements to positive elements based on the text explaining the reason for creation is displayed.

[0009] The invention of claim 4 further solves the above-mentioned problem by configuring the recommendation system as described in claim 3, in addition to the configuration of the recommendation system, in which the recommendation means uses a pre-trained artificial intelligence means to generate a reason explanation text for recommending a selected destination and travel route with content that changes from negative elements to positive elements, and generates the reason explanation text based on a prompt that the positive content is content that is obtained based on the negative elements.

[0010] The invention of claim 5 further solves the above-mentioned problem by being configured such that, in addition to the configuration of the recommendation system described in any one of claims 1 to 4, when a change or deletion operation is performed on the schedule information on the user terminal, the recommendation means on the user terminal or server searches for free time information in the user's schedule information, and if free time information is detected, generates a sentence about the activity theme using a pre-trained artificial intelligence means based on the length of the free time information.

[0011] The invention of claim 6 further solves the above-mentioned problem by being configured such that, in addition to the configuration of the recommendation system described in any one of claims 1 to 4, when an operation to launch an arbitrary application is performed on the user terminal, if a recommendation means in the user terminal or server detects a change in the operation to launch an arbitrary application that is greater than a predetermined early timing change amount relative to the daily log information on the user terminal, the recommendation means generates a sentence about the behavioral theme using a pre-trained artificial intelligence means based also on the change amount of the operation to launch the arbitrary application.

[0012] The invention according to claim 7 is a program for a recommendation system that recommends destinations to a user, and when a predetermined operation for requesting a recommendation is performed on a user terminal, a recommendation means in the user terminal or server uses pre-input user attribute information, input or acquired current user situation information, and, Acquired external environment information At least the user's attribute information entered in advance The above-mentioned problem is solved by executing the following steps: an action theme generation step, which generates a sentence about an action theme using a pre-trained artificial intelligence means based on the action theme; a destination and means of transportation selection step, which selects a destination and means of transportation that are consistent with the sentence about the action theme from a database on the server; a negative element evaluation step, which evaluates negative elements based on the selected destination information and travel route information by the means of transportation, and the information used to generate the sentence about the action theme; a reason explanation text generation step, which generates a reason explanation text based on a prompt to generate a reason explanation text that includes negative elements and changes from negative elements to positive elements when generating a reason explanation text recommending the selected destination and travel route based on the selected destination information and travel route information by the means of transportation using the pre-trained artificial intelligence means; and a recommendation display step, which displays the generated reason explanation text, destination information, and travel route information on the user terminal. [Effects of the Invention]

[0013] The recommendation system of the present invention is equipped with a user terminal and a server, which not only enables the user terminal to communicate with the server, but also provides the following unique effects.

[0014] According to the recommendation system of the invention of claim 1, when destination information and travel route information are displayed as a suggestion on a user terminal, a text explaining the reason for creation is displayed based on this destination information and travel route information, which may have some negative elements but contains content that changes those negative elements into positive ones. Therefore, the user can easily understand, trust, and sympathize with the reason for the suggestion, which includes ways to enjoy and see the negative elements in accordance with the text explaining the reason for creation, which is the reason for the recommendation. In addition, by including the negative elements in the explanation of the reason for the recommendation, it is more persuasive to the user that they chose this (the proposal they chose) rather than that (the proposal they did not choose), which makes the user feel more convinced. This makes the user more likely to take action based on the content of the suggestion, thereby increasing user acceptance. Furthermore, if the user acts in accordance with the suggestions, the positive aspects of the experience will be outweighed by the negative aspects, resulting in an unforgettable, memorable experience that leaves the user feeling satisfied. In other words, because the negative elements are included in the reasons for proposing the proposal, the user is made aware of the negative elements, but the subsequent positive proposal makes them feel like a small success, which widens the range of their impression, so the user feels that it was a dramatic experience that was at least somewhat conscious of something nice, and is able to feel satisfied.

[0015] According to the recommendation system of the invention of claim 2, in addition to the effects achieved by the invention of claim 1, any of the following items are evaluated as negative elements and displayed in the text explaining the reason for generation: bad weather, waiting time at the destination for more than a specified time, restrictions on payment methods, construction work that makes the location look different than usual, or the route to travel has more than a specified distance of uphill or staircases. These are then reversed to positive elements such as "XXX, but ○○○" and presented as a recommendation, which is more persuasive to the user and allows the user to feel more convinced.

[0016] According to the recommendation system of the invention of claim 3, in addition to the effect achieved by the invention of claim 2, a predetermined mark icon is displayed adjacent to each of the multiple displayed destination icons, and when the predetermined mark icon adjacent to the destination icon is operated, the content of the flow of changing from negative elements to positive elements based on the generation reason explanation text, which is the reason why the destination is recommended, is displayed, so that the user can easily visually understand the relationship between the destination and the content of the flow of changing from negative elements to positive elements.

[0017] According to the recommendation system of the invention of claim 4, in addition to the effect of the invention of claim 3, the positive content in the text explaining the reason for creation is content obtained based on negative elements, and the negative elements and positive content are written in association with each other, so that the user can understand the relationship between the negative elements and the positive content and feel even more convinced. For example, "It looks like it's going to rain, but because it's raining, the observation deck is relatively empty, and the Japanese garden near the observation deck is filled with not only colorful flowers but also umbrellas, creating a slightly different atmosphere in rainy weather than on sunny days, so I highly recommend it." Positive content is derived from negative elements, and the negative elements are written in association with the positive content, so users can understand the relationship between the negative elements and the positive content and feel even more convinced.

[0018] According to the recommendation system of the invention of claim 5, in addition to the effects achieved by the invention of any one of claims 1 to 4, when a free time occurs in the user's schedule, an activity theme is generated based on the length of the free time information, and a destination and means of transportation are selected and displayed as a suggestion on the user's terminal along with the generated course name text, allowing the user to make effective use of the free time in their schedule.

[0019] According to the recommendation system of the invention of claim 6, in addition to the effects achieved by the invention of any one of claims 1 to 4, if a specified operation is performed at a timing earlier than the usual time by a specified amount or more based on the user's daily log, it is determined that free time has occurred, and an action theme is generated based on the amount of change in the specified operation, and a destination and means of transportation are selected and displayed as a suggestion on the user's terminal along with the generated course name text, allowing the user to make effective use of free time in their schedule.

[0020] According to the program of the recommendation system of the invention of claim 7, similar to the effect achieved by the invention of claim 1, when destination information and travel route information are displayed as a suggestion on the user's terminal, a text explaining the reason for creation is displayed based on this destination information and travel route information, which may have some negative elements but contains content that changes those negative elements into positive ones. Therefore, the user can easily understand the reason for the suggestion, which includes ways to enjoy and see the negative elements in accordance with the text explaining the reason for creation, and can trust and sympathize with it. In addition, by including the negative elements in the explanation of the reason for the recommendation, it is more persuasive to the user that they chose this (the proposal they chose) rather than that (the proposal they did not choose), which makes the user feel more convinced. This makes the user more likely to take action based on the content of the suggestion, thereby increasing user acceptance. Furthermore, if the user acts in accordance with the suggestions, the positive aspects of the experience will be outweighed by the negative aspects, resulting in an unforgettable, memorable experience that leaves the user feeling satisfied. In other words, because the negative elements are included in the reasons for proposing the proposal, the user is made aware of the negative elements, but the subsequent positive proposal makes them feel like a small success, which widens the range of their impression, so the user feels that it was a dramatic experience that was at least somewhat conscious of something nice, and is able to feel satisfied. [Brief explanation of the drawings]

[0021] [Figure 1] FIG. 1 is a diagram showing the concept of a recommendation system according to an embodiment of the present invention. [Figure 2] FIG. 1 is a chart showing an example of the operation of a recommendation system according to an embodiment of the present invention. [Figure 3] FIG. 1 is a diagram showing a My Page screen of a recommendation system according to an embodiment of the present invention. [Figure 4] FIG. 10 is a diagram showing a "Your Mood" screen of the recommendation system according to an embodiment of the present invention. [Figure 5] FIG. 1 is a diagram showing a search screen of a recommendation system according to an embodiment of the present invention. [Figure 6] FIG. 10 is a diagram showing an example of data for generating an action theme of a recommendation system according to an embodiment of the present invention. [Figure 7] 1A and 1B are diagrams showing an example of a template for an action theme of a recommendation system according to an embodiment of the present invention, and an example of information inserted into the template. [Figure 8] FIG. 10 is a diagram showing examples of negative elements of a recommendation system according to an embodiment of the present invention. [Figure 9] 1A and 1B are diagrams showing examples of recommendation displays of a recommendation system according to an embodiment of the present invention. [Figure 10] FIG. 1 is a diagram showing an example of a recommendation display of a recommendation system according to an embodiment of the present invention. [Figure 11]1A and 1B are diagrams showing examples of input data of a recommendation system according to an embodiment of the present invention, examples of generated course name text, examples of generated story text with an introduction, development, twist, and conclusion, and examples of generation reason explanation text. [Figure 12] FIG. 10 is a diagram showing an example of a pop-up display when a change of more than a predetermined early timing change amount is detected in the log information of an operation to launch an arbitrary application on a user terminal of a recommendation system according to an embodiment of the present invention. [Figure 13] FIG. 10 is a diagram showing an example of a pop-up display when free time information is searched for in a user's schedule information on the schedule screen of a user terminal of a recommendation system according to an embodiment of the present invention, and free time information is detected. DETAILED DESCRIPTION OF THE INVENTION

[0022] The recommendation system of the present invention comprises a user terminal and a server, and when a predetermined operation for requesting a recommendation is performed on the user terminal, the recommendation means in the user terminal or server uses pre-trained artificial intelligence means to generate a sentence about an action theme based on pre-inputted user attribute information, inputted or acquired current user situation information, and acquired external environment information, selects a destination or service and a means of transportation that matches the sentence about the action theme from the server database, evaluates negative elements from the selected destination information or service information and travel route information by the means of transportation, and information used to generate the sentence about the action theme, and When a pre-trained artificial intelligence means is used to generate a reason explanation text for recommending a selected destination and travel route based on information about the travel route by the means of transportation, the reason explanation text is generated based on a prompt to generate a reason explanation text that includes negative elements and changes from negative elements to positive elements, and the reason explanation text, destination information or service information, and travel route information are displayed on the user terminal, so that the user can easily understand, trust, and sympathize with the reason for the proposal of the proposed content, which includes ways to enjoy and see the negative elements accordingly, from the reason explanation text that is the reason for the recommendation, and can do so.The specific implementation is not limited to any one. Furthermore, the program of the recommendation system of the present invention includes, when a predetermined operation for requesting a recommendation is performed on a user terminal, an action theme generation step in which a recommendation means generates a sentence about an action theme using pre-trained artificial intelligence means on the user terminal or server based on pre-input user attribute information, input or acquired current user situation information, and acquired external environment information; a destination / transportation means selection step in which a destination or service and transportation means in accordance with the sentence about the action theme are selected from a database of the server; a negative element evaluation step in which negative elements are evaluated from the selected destination information or service information and travel route information by the transportation means, and information used to generate the sentence about the action theme; and a negative element evaluation step in which the selected destination information or service information is selected. When generating a reason explanation text for a selected destination and travel route based on information about the travel route by the means of transportation using a pre-trained artificial intelligence means, a reason explanation text generation step is executed in which a reason explanation text is generated based on a prompt to generate a reason explanation text with content that includes negative elements and changes from negative elements to positive elements, and a recommendation display step is executed in which the generated reason explanation text, destination information or service information, and travel route information are displayed on the user terminal.As long as the user can easily understand, trust, and sympathize with the reason for the proposal of the proposed content, which includes ways to enjoy and see the negative elements in accordance with the generated reason explanation text, which is the reason for the recommendation, from the generated reason explanation text, which is the reason for the recommendation, and can do so.

[0023] For example, the user terminal may be any device that has a display and an operating unit and can send and receive information, such as a notebook-type personal computer terminal, a smartphone terminal, a tablet terminal, a wristwatch-type terminal, or an eyeglass-type terminal, and 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. Furthermore, the server may be a cloud server created in a cloud environment, and the number of physical servers that make up the server may be one or more. Furthermore, the pre-trained artificial intelligence means may be composed of a text generation means that generates text data, such as an interactive ChatGPT (Generative Pre-trained Transformer) (hereinafter referred to as ChatGPT), which is also called a large-scale language model, an image generation means that generates image data such as photographs, illustrations, and videos, or a voice generation means that generates voice data, and may be configured to be installed on one server or multiple servers on the cloud. [Example]

[0024] A recommendation system 100 according to an embodiment of the present invention will be described below with reference to FIGS. 1 is a diagram showing the concept of a recommendation system 100 according to an embodiment of the present invention, FIG. 2 is a chart showing an example of the operation of the recommendation system 100 according to an embodiment of the present invention, FIG. 3 is a diagram showing a My Page screen 112 of the recommendation system 100 according to an embodiment of the present invention, FIG. 4 is a diagram showing a Your Mood screen 113 of the recommendation system 100 according to an embodiment of the present invention, FIG. 5 is a diagram showing a search screen 114 of the recommendation system 100 according to an embodiment of the present invention, FIG. 6 is a diagram showing an example of data for generating an action theme AT of the recommendation system 100 according to an embodiment of the present invention, FIG. 7(A) is a diagram showing an example of a template of an action theme AT of the recommendation system 100 according to an embodiment of the present invention, FIG. 7(B) is a diagram showing an example of information inserted into the template, and FIG. 8 is a diagram showing a negative of the recommendation system 100 according to an embodiment of the present invention. 9A to 9C are diagrams showing examples of recommendation displays of the recommendation system 100 according to an embodiment of the present invention; FIG. 11 is a diagram showing examples of input data of the recommendation system 100 according to an embodiment of the present invention, an example of generated course name text CS, an example of generated story text SR with an introduction, development, twist, and conclusion, and an example of generation reason explanation text RE; FIG. 12 is a diagram showing an example of a pop-up display PP when a change of more than a predetermined early timing change amount is detected in the operation of launching an arbitrary application in the log information in the user terminal 110 of the recommendation system 100 according to an embodiment of the present invention; and FIG. 13 is a diagram showing an example of a pop-up display PP when free space information in the user's schedule information is searched for and free space information is detected on the schedule screen 117 of the user terminal 110 of the recommendation system 100 according to an embodiment of the present invention.

[0025] As shown in FIG. 1, a recommendation system 100 according to an embodiment of the present invention includes a user terminal 110 and a server 120. The user terminal 110 is provided so as to be able to communicate freely with the server 120 . The server 120 includes a database 121 . The recommendation system 100 is configured to recommend destinations to users based on a database 121 that registers the latest information on products, services, locations, etc. input from a store terminal (not shown).

[0026] More specifically, first, the user logs in to the recommendation system 100 using the user terminal 110 . The user then inputs the user's attribute information US and preference information LK in advance. For example, on a user registration screen (not shown) of the recommendation system 100 displayed on the display unit 111 of the user terminal 110, the user inputs and registers the user's attribute information US in an input field.

[0027] In addition, the "Favorite Things" screen (not shown) of the recommendation system 100 displayed on the display unit 111 of the user terminal 110 has categories such as gourmet, cafes, outdoors, shopping, going out, entertainment, fashion, and indoors, and within each category, multiple photos of content related to that category are displayed. The user then selects one or more photos that show something they like. Each photo is tagged with information that indicates what is shown in the photo. When a photo is selected by the user, the tag information of the selected photo is added as preference information LK, which is information about the user's favorite things.

[0028] The user may also select icons, keywords, sounds, etc. in addition to photos. Icons, keywords, and sounds represent something. In the case of audio, a button is displayed so that when the button is operated, something is output as audio, and when the button is selected, tag information of the button content is added as preference information LK. In addition, the user may use the microphone on the user terminal 110 to input voice input about what they like, or may use the input keys to input text about what they like in sentences or words, and this information may be added as preference information LK.

[0029] Furthermore, a "Your Personality" screen (not shown) of the recommendation system 100 displayed on the display unit 111 of the user terminal 110 includes a plurality of simple questions. For example, questions include, "I like meeting new people and can enjoy conversations with people even when meeting them for the first time." As an example, the system is designed so that users can intuitively answer each question using a five-point scale, such as completely not applicable, hardly applicable, neither applicable nor unapplicable, somewhat applicable, and completely applicable. When the user inputs these answers, a recommendation means having artificial intelligence in the user terminal 110 or the server 120 analyzes the answer information and obtains user type tendency information TP. It is assumed that an operation for requesting a recommendation is then performed as a predetermined operation. Here, operations to request a recommendation include launching the recommendation system 100 app on the user terminal 110, operating the "Leave it to AI" button 114a described below, or inputting the user's current situation information ST. Then, when an operation to request a recommendation is performed as a predetermined operation, the recommendation means in the user terminal 110 or the server 120 generates a sentence about the behavioral theme AT using a pre-trained artificial intelligence means based on at least the user's attribute information US that has been input in advance.

[0030] Furthermore, the recommendation means selects a destination and a means of transportation that are in line with the sentence of the activity theme AT from the database 121 of the server 120. Next, the recommendation means generates a text explaining the reasons for recommending the selected destination and travel route based on the selected destination information DN and travel route information RT by travel means using the pre-trained artificial intelligence means. At this time, the recommendation means generates the reason explanation text based on a prompt to generate a reason explanation text including negative elements and changing from negative elements to positive elements. The recommendation means is configured to display the creation reason explanation text RE, the destination information DN, and the travel route information RT on the display unit 111 of the user terminal 110.

[0031] As a result, when the destination information DN and travel route information RT are displayed as a proposal on the user terminal 110, a creation reason explanation text RE based on this destination information DN and travel route information RT, which has some negative elements but has the content of the flow of changing the negative elements into positive ones, is displayed. As a result, users can easily understand the reason for the recommendation, including how to enjoy the content and highlights according to the negative elements, from the generated reason explanation text RE, which is the reason for the recommendation, and can trust and sympathize with it.

[0032] In addition, by including the negative elements in the RE, which explains the reason for the recommendation, it is possible to convey to the user in a more persuasive way that they chose this (the proposal they chose) rather than that (the proposal they did not choose). As a result, users can feel more satisfied. Users are then more likely to be motivated to take action based on the suggestions. As a result, user acceptance can be increased.

[0033] Furthermore, if the user acts in accordance with the suggestions, the positive aspects of the experience will outweigh the negative aspects. As a result, users are provided with an unforgettable and memorable experience, leaving them feeling satisfied. In other words, because the negative elements are included in the reasons for proposing the proposal, the user is made aware of the negative elements, but the subsequent positive proposal makes them feel like a small success, widening the range of their impression. As a result, users feel that the experience is at least somewhat dramatic and well-thought-out, and they are left feeling satisfied.

[0034] Next, the operation of the recommendation system 100 will be described in detail. As shown in FIG. 2, in step S1, as a step of determining whether or not a recommendation request operation has been performed as a predetermined operation, the user terminal 110 or the server 120 determines whether or not an operation requesting a recommendation has been performed as the predetermined operation described above. As mentioned above, operations to request recommendations as a predetermined operation include launching the recommendation system 100 app, operating the "Leave it to AI" button 114a described below, or inputting the user's current situation information ST.

[0035] Here, before performing the operation to request recommendations, the user registers by entering the user's attribute information US in the input field on the user registration screen (not shown) of the recommendation system 100 displayed on the display unit 111 of the user terminal 110. Furthermore, the user may select a photo on the "Favorite Things" screen described above, for example. Then, the user terminal 110 or the server 120 registers the user's preference information LK in the database 121 in the server 120.

[0036] Then, as shown in FIG. 3, once the user's preference information LK is registered, the user's preference information LK is displayed on a My Page screen 112 of the recommendation system 100 displayed on the display unit 111 of the user terminal 110. More specifically, the tag information of the photograph selected by the user is registered as the preference information LK, and the tag information is displayed in the "Favorite things and events" section.

[0037] In addition, the recommendation means may be configured to display the user type tendency information TP obtained based on the answer information provided on the aforementioned ``Your Personality'' screen (not shown) in the ``Your Personality'' section of the My Page screen 112. As an example, the user type tendency information TP has items such as openness, conscientiousness, extroversion, agreeableness, and neuroticism, and is configured to be displayed as a level value for each item. The user type tendency information TP is configured to be registered in the database 121 in the server 120, similar to the user preference information LK.

[0038] Then, as an example of the operation for requesting a recommendation as a predetermined operation, the user terminal 110 or the server 120 determines whether or not the user has performed an operation to input the current situation information ST of the user in the user terminal 110. For example, as shown in FIG. 4, the user's current situation information ST is freely inputtable on the "Your Mood" screen 113 of the recommendation system 100 displayed on the display unit 111 of the user terminal 110.

[0039] The "Your Mood" screen 113 has fields for inputting information about people you are with, free time information, appetite and material desire information, and activity and rest desire information. For example, in the section about information about people you are with, options such as alone, couple, friends, and family are set to indicate whether you are alone or who you are with. Furthermore, the item of free time information is provided so that the current amount of free time can be freely input. In addition, for the appetite and material desire information items, you can freely input which of the following is given more weight. Furthermore, for the active / rest desire information items, you can freely input which of the two you would like to prioritize. When an input operation is performed on these items, the user terminal 110 or the server 120 determines that the user's current situation information ST has been input.

[0040] Also, as shown in Figure 5, as an example of an operation to request recommendations as a specified operation, the user terminal 110 or the server 120 may determine whether or not the "Leave it to AI" button 114a displayed on the search screen 114 of the recommendation system 100 of the user terminal 110 has been operated. If it is determined that an operation for which a recommendation is requested has been performed, the process proceeds to step S2, and if it is determined that no operation has been performed yet, step S1 is repeated.

[0041] In step S2, as an action theme generation step, the recommendation means in the user terminal 110 or the server 120 generates sentences about the action theme AT using pre-trained artificial intelligence means based on at least the user attribute information US input in advance. Here, as shown in FIG. 6, the data used to generate the behavioral theme AT may include preference information LK about hobbies and preferences entered on the My Page screen 112 in FIG. 3, situation information ST entered on the Your Mood screen 113 in FIG. 4, comment information entered by the user in chat or the like on a social networking service (SNS), location information of the user terminal 110 from a Global Positioning System (GPS), log information about the operation and location of the user terminal 110, and information entered via the user terminal 110, such as vital information such as heart rate, blood pressure, and sleep status obtained from a watch-type terminal worn on the user's wrist, etc.

[0042] In addition, external environment information EF, which is external environment data such as coupon information for any spot, information on means of transportation, traffic congestion information, store congestion information, event information such as concerts, and weather information, may be used as data for generating behavioral themes AT. Furthermore, as data for generating the behavioral theme AT, training data using collaborative filtering technology or statistical analysis technology, such as spot matching trend data that accumulates previous recommendation information and user satisfaction information (feedback information), may be used.

[0043] We explain the generation of behavioral themes AT. Here, the behavioral theme AT refers to an expression of what kind of person is considering a destination and a means of transportation under what conditions. For example, as shown in Figure 7(A), the recommendation means may insert weather information, transportation information, hobby and preference information LK, vital signs information, situation information ST which is your mood, and user attribute information US into a template, and generate sentences about an action theme AT using a pre-trained artificial intelligence means.

[0044] More specifically, as shown in Figure 7(B), the recommendation means may insert weather information, transportation information, hobby and preference information LK, vital signs information, situation information ST (your mood), and user attribute information US into a template, and generate sentences about the behavioral theme AT using a pre-trained artificial intelligence means. As an example, as shown in Figure 7(B), the recommendation means inserts each piece of information into the template shown in Figure 7(A) to generate a sentence about the behavioral theme AT, such as "In the weather where there is a 50% chance of rain, a person with the user attributes of a 'female office worker in her 30s' who is a type that 'likes emotional expressions' and tends to 'like castles' is considering a destination and mode of transportation when they feel like 'relaxing for about three hours' in an area that has 'spot data that can be enjoyed even on a rainy day', with the travel options of 'walking or train'."

[0045] Although the sentences about the behavioral theme AT are generated using the template, it goes without saying that the sentences about the behavioral theme AT may be generated without using the template. Here, the reason for using a template as an example is that by using a template, the format of the text about the behavioral theme AT is established. As a result, when selecting candidate destinations and modes of transportation in the next step, the accuracy of the generated sentences about the action theme AT can be improved when vectorized as an example, thereby improving the accuracy of selection and making suggestions that suit the user.

[0046] In step S3, as a destination and transportation means selection step, the recommendation means selects a destination and transportation means that are in line with the sentence of the activity theme AT from the database 121 of the server 120. In this case, the recommendation means selects a destination and means of transportation that are in line with the text of the behavioral theme AT from the database 121 of the server 120 using an estimation model that estimates, as a concept of technical idea, that the user will be satisfied by providing an output in response to the input, for example, having a relationship between output information such as destination information DN and means of transportation information that correspond to input information such as the user's attribute information US, situation information ST, and preference information LK. In order to improve the accuracy of the output when making this selection, as one example, the recommendation means may be configured to vectorize (embedding process) the generated sentences about the behavioral theme AT.

[0047] In step S4, as a negative element evaluation step, the recommendation means evaluates negative elements from the selected destination information DN and travel route information RT by the travel means, as well as the information used to generate the sentence about the activity theme AT. More specifically, as shown in Figure 8, for example, if the weather forecast for the destination or travel route indicates rain, the negative factor is evaluated in that the rain will reduce the enjoyment of travel and the scenery. Regarding the evaluation of these negative elements, things may be evaluated as "+ (positive)" or "- (negative)," and the recommendation means may be configured to focus on those with an evaluation value of "- (negative)."

[0048] Furthermore, if the waiting time at the destination is longer than a predetermined value, a negative factor is evaluated in that there is a queue at a popular spot (taking extra time). Furthermore, if the route includes uphill slopes or stairs, the negative factors are evaluated in terms of the physical strain caused by steep uphill slopes or stairs. Additionally, if there are payment restrictions at the destination and the only payment method is cash, the fact that cashless payment is not possible is evaluated as a negative factor. Furthermore, if there is news about a destination and the exterior is undergoing construction, the negative element is evaluated in that the construction makes it look unsightly.

[0049] In step S5, as a reason explanation text generation step, the recommendation means uses the pre-trained artificial intelligence means to generate a reason explanation text in natural language that recommends the selected destination and travel route based on the selected destination information DN and travel route information RT by the travel means. At this time, the recommendation means generates the reason explanation text based on a prompt to generate a reason explanation text including negative elements and changing from negative elements to positive elements. For example, as shown in Figure 8, the recommendation means uses pre-trained artificial intelligence means to generate a positive image of "a quiet experience of the sound of rain, windows, and sweets" in response to the negative element of rain reducing the enjoyment of travel and scenery, based on destination information DN and travel route information RT by means of transportation.

[0050] In addition, the recommendation method uses pre-trained artificial intelligence to generate positive impressions such as "numbered ticket system + ability to move around the area, valuable experience" in contrast to negative elements such as long lines at popular spots (which takes extra time). Furthermore, the recommendation means uses pre-trained artificial intelligence means to generate positive elements such as "shade, benches, and scenery along the way" in response to negative elements such as steep uphill slopes and stairs that are physically demanding. In addition, the recommendation method uses pre-trained artificial intelligence to generate positive impressions such as "the charm of a retro experience / limited quantity" in contrast to the negative element of not being able to pay with cash. Furthermore, the recommendation means uses a pre-trained artificial intelligence means to generate positive comments such as "less crowded and the view remains attractive" in response to negative factors such as the unsightly appearance due to construction.

[0051] In addition, examples of generating reason explanation texts include, for example, based on the destination information DN and the travel route information RT by means of transportation, messages such as "For those of you who like temples, even though it is raining now, I would like you to experience a course where you can feel the smell of rainbow after the rain, so I recommend this!", "For those of you who like Japanese sweets, even though the weather is a little chilly, I would like you to enjoy a course of walking that will warm you up and matcha sweets, so I recommend this!", or "For those of you who like tradition and Japanese things, even though you cannot drink alcohol, I would like you to experience a course of visiting a sake brewery and shopping where you can buy wonderful things made with koji, so I recommend this!".

[0052] In step S6, as a recommendation display step, the recommendation means displays the creation reason explanation text RE, the destination information DN, and the travel route information RT on the user terminal 110. More specifically, as shown in Figures 9(A) to 10, the recommendation means displays, as the generation reason explanation text RE, destination information DN and travel route information RT on the recommendation display screen 115 of the display unit 111 of the user terminal 110, such as "For those of you who like temples, even though it's raining now, I would like you to experience a course where you can feel the smell of after the rain and a rainbow, so I recommend this!", "For those of you who like Japanese sweets, even though the weather is a little chilly, I would like you to enjoy a course of walking and matcha sweets that will warm you up, so I recommend this!", or "For those of you who like tradition and Japanese things, even though you can't drink alcohol, I would like you to experience a course of visiting a sake brewery and shopping where you can buy wonderful things made with koji, so I recommend this!".

[0053] As a result, as described above, when the destination information DN and travel route information RT are displayed as a suggestion on the user terminal 110, a creation reason explanation text RE based on this destination information DN and travel route information RT, which has some negative elements but contains the content of the flow of changing those negative elements into positive ones, is displayed. As a result, users can easily understand the reason for the recommendation, including how to enjoy the content and highlights according to the negative elements, from the generated reason explanation text RE, which is the reason for the recommendation, and can trust and sympathize with it. In addition, by including the negative elements in the RE, which explains the reason for the recommendation, it is possible to convey to the user in a more persuasive way that they chose this (the proposal they chose) rather than that (the proposal they did not choose). As a result, users can feel more satisfied.

[0054] Users are then more likely to be motivated to take action based on the suggestions. As a result, user acceptance can be increased. Furthermore, if the user acts in accordance with the suggestions, the positive aspects of the experience will outweigh the negative aspects. As a result, users are provided with an unforgettable and memorable experience, leaving them feeling satisfied.

[0055] In other words, because the negative elements are included in the reasons for proposing the proposal, the user is made aware of the negative elements, but the subsequent positive proposal makes them feel like a small success, widening the range of their impression. As a result, users feel that the experience is at least somewhat dramatic and well-thought-out, and they are left feeling satisfied. Note that the recommendation means may be configured to generate image data for the contents of the generation reason explanation text RE using image generation means, and display the image data on the recommendation display screen 115 or the detail display window WD.

[0056] Furthermore, in this embodiment, as shown in Figure 8, the recommendation means evaluates as negative factors any of the following from the pre-input user attribute information US, the situation information ST which is input or acquired current status information of the user, and the acquired external environment information EF: weather information indicating bad weather; selected destination information DN and travel route information RT by travel means indicating that there is a waiting time at the destination for more than a predetermined time, that there are restrictions on payment methods, that there is construction work and the appearance is not as usual, or that there is a slope or staircase climbing for more than a predetermined distance on the travel route. The recommendation means is configured to generate a reason explanation text including the evaluated negative elements.

[0057] As a result, any of the following will be evaluated as negative factors: the weather report is bad, there will be a waiting time at the destination that exceeds a certain time, there are restrictions on payment methods, construction work is occurring and the view will not be as usual, or there will be a certain distance or more of uphill or staircases along the route, and these will be displayed in the generation reason explanation text RE. Then, they will be reversed to a positive one, such as "XXX, but ○○○", and proposed as a recommendation, making the message more persuasive to the user. As a result, users can feel more satisfied.

[0058] In this embodiment, as shown in FIGS. 9(A) to 10, the recommendation means selects a plurality of sets of destinations and means of transportation that are in line with the sentences in the activity theme AT from the database 121 of the server 120. Then, the recommendation means displays the destination information DN and the travel route information RT on the user terminal 110 based on the multiple sets of selected destination information DN and travel route information RT by travel means. At this time, the recommendation means places the destination icon IC on the map and displays a predetermined mark icon ICa in a position adjacent to the destination icon IC. When the predetermined mark icon ICa is operated, the recommendation means is configured to display the contents of the flow of change from negative elements to positive elements based on the creation reason explanation text RE.

[0059] As a result, a predetermined mark icon ICa adjacent to each of the multiple displayed destination icons IC is displayed, and when the predetermined mark icon ICa adjacent to the destination icon IC is operated, the content of the flow of changing from negative elements to positive elements based on the creation reason explanation text RE, which is the reason why the destination is recommended, is displayed. As a result, users can easily visually understand the relationship between the destination and the content of the flow that changes from negative to positive.

[0060] For example, as shown in FIGS. 8 and 9(A), in the case of a negative factor of rain, an umbrella mark as an example of a predetermined mark icon ICa is displayed on the upper right of the destination icon IC on the map. Then, by tapping on the umbrella mark, which is an example of a predetermined mark icon ICa, a speech bubble saying "Why rainy days are the best days to relax" may appear, or immediately after making a recommendation, the AI ​​may say on the chat screen, "It's raining, but there's actually a reason why I recommend this."

[0061] Furthermore, as shown in Figures 8 and 9(B), in the case of a negative factor such as a queue, the recommendation means may be configured to display a navigational message on the map of the recommendation display screen 115 saying "Get a numbered ticket → Go to the tour course," or to back up the value of waiting in line with reviews such as "8 stars" or "There are ○ people waiting in line, but I'm satisfied." Furthermore, as shown in Figures 8 and 9(B), in the case of negative factors such as steep slopes and a large physical burden, the recommendation display screen 115 may be configured to display "rest points along the way" or "benchmarks" on the course, which is the travel route information RT on the map, or to provide positive supplements such as "calories burned ○ kcal" with a health tag.

[0062] Furthermore, as shown in Figures 8 and 9(B), in the case of a negative factor such as not being able to pay cashlessly, the payment conditions can be clearly indicated by an icon, such as a predetermined mark icon ICa, and the AI ​​can be configured to provide conversational supplementary information via chat, such as "This printing shop takes cash, but they will be sold out by the afternoon, so I recommend it now." Furthermore, as shown in FIGS. 8 and 10, in the case of a negative factor such as poor appearance due to construction, an "under construction" mark badge as an example of a predetermined mark icon ICa is displayed at the spot on the map of the recommendation display screen 115. When the "under construction" mark badge, which is an example of a predetermined mark icon ICa, is operated, a comment such as "It's under construction, but now is the time to enjoy it quietly" is displayed, or the configuration may be such that it is complemented by a coupon link (e.g., a limited-time discount during the construction period).

[0063] Furthermore, in this embodiment, the recommendation means uses pre-trained artificial intelligence means to generate text explaining the reasons for recommending the selected destination and travel route, with content that changes from negative elements to positive ones. In this case, the recommendation means is configured to generate a reason explanation text using the pre-trained artificial intelligence means based on a prompt that the positive content is content obtained based on the negative elements.

[0064] As a result, the positive content in the creation reason explanation text RE becomes content obtained based on the negative elements, and the negative elements and the positive content are written in association with each other. As a result, users can understand the relationship between negative elements and positive content, which gives them a greater sense of satisfaction. For example, as shown in Figure 11, "It looks like it's going to rain, but because it's raining, the observation deck is relatively empty, and the Japanese garden near the observation deck is filled with not only colorful flowers but also umbrellas, creating a slightly different atmosphere that can only be enjoyed on rainy days compared to sunny days, so I highly recommend it." Positive content is derived from negative elements, and negative elements are associated with positive content. As a result, users can understand the relationship between negative elements and positive content, which gives them a greater sense of satisfaction.

[0065] Furthermore, in this embodiment, the recommendation means acquires the user's physical condition information and mood information from the user's attribute information US that has been input in advance, situation information ST that is the input or acquired current status information of the user, and the acquired external environment information EF, and generates a sentence about the behavioral theme AT including the user's physical condition information and mood information using the pre-trained artificial intelligence means. For example, the recommendation means generates a sentence about the behavioral theme AT using a prompt to the pre-trained artificial intelligence means to "generate a sentence about the behavioral theme AT including the user's physical condition information and mood information." Furthermore, when the recommendation means generates a reason explanation text using a pre-trained artificial intelligence means based on the selected destination information DN and travel route information RT by the travel means, the recommendation means is configured to generate a reason explanation text with content that touches on the user's physical condition information and mood information. For example, the recommendation means generates a reason explanation text by using a prompt to the pre-trained artificial intelligence means to "generate a reason explanation text that touches on the user's physical condition information and mood information."

[0066] More specifically, as shown in FIG. 11, the user attribute information US includes information such as "resident of Nagoya city, 34 years old, office work, accounting, arrives at work at 10:00 on weekdays, arrives at Nagoya Station at 8:30 in the morning, and usually goes straight to work." Furthermore, the situation information ST includes information such as "Mood this morning: good. Yesterday, because of the rain, I stayed at home and did not go out." Additionally, vital signs log information includes information such as "the number of steps taken recently has been low and there are concerns about his health." Then, as shown in FIG. 7(A) and FIG. 7(B), an action theme AT including the user's physical condition information and mood information is generated, and then a destination and a means of transportation are selected. The creation reason explanation text RE is displayed, for example, in a detail display window WD of the recommendation display screen 115 shown in FIGS. 9(A) to 10. FIG.

[0067] This generates a sentence for the behavioral theme AT that reflects the user's physical condition information and mood information, and based on this, a destination and means of transportation are selected, and the reason for selecting that destination and transportation route is expressed in the generation reason explanation text RE. As a result, users can easily understand that the suggestion is based on a thorough understanding of them, and they can trust, be convinced, and be satisfied. This makes users more likely to take action based on the suggestions they receive. As a result, user acceptance can be further increased.

[0068] In this embodiment, the recommendation means acquires weather information from the acquired external environment information EF, and generates sentences about the behavioral theme AT including the weather information using the pre-trained artificial intelligence means. For example, the recommendation means generates a sentence about the behavioral theme AT using a prompt to the pre-trained artificial intelligence means to "generate a sentence about the behavioral theme AT including weather information." Next, when the recommendation means generates a reason explanation text using the pre-trained artificial intelligence means based on the selected destination information DN and travel route information RT by the travel means, it is configured to generate a reason explanation text that includes content that touches on weather information. For example, the recommendation means generates a reason explanation text by using a prompt to the pre-trained artificial intelligence means to "generate a reason explanation text that includes content that touches on weather information."

[0069] More specifically, as shown in FIG. 11, the weather information includes information that states "rain, temperature 25°C." Then, as shown in Fig. 7(A) and Fig. 7(B), an activity theme AT including weather information is generated, and then a destination and a means of transportation are selected. Next, as shown in Figure 9(A), for example, the recommendation means generates a recommendation reason explanation text RE that touches on meteorological information, such as, "If you like temples, even though it's raining now, I would like you to experience a course where you can feel the smell of rainbows after the rain, so I recommend this course!" Then, the recommendation means displays the creation reason explanation text RE on the recommendation display screen 115 of the display unit 111 of the user terminal 110.

[0070] This generates a sentence for the behavioral theme AT that reflects meteorological information including the weather, temperature, humidity, wind, etc., and based on this, a destination and means of transportation are selected, and the reason for choosing that destination and transportation route is expressed in the generation reason explanation text RE. As a result, users can easily understand that the system is making suggestions based on an understanding of the weather, temperature, humidity, wind, etc., and can trust, understand, and be more satisfied. This makes users more likely to take action based on the suggestions they receive. As a result, user acceptance can be further increased.

[0071] Furthermore, in this embodiment, the recommendation means generates course name text for the course name using the pre-trained artificial intelligence means based on the selected destination information DN and travel route information RT by the travel means. More specifically, the recommendation means uses a pre-trained artificial intelligence means to generate, for example, a "Temple of XX Temple Course," a "Sweets and Walking Course," or a "Shopping Course in the Morning Sky" based on the destination information DN and the travel route information RT by the means of transportation. As shown in FIG. 11, the generated course name text CS may be something along the lines of "A wonderful hidden course around Nagoya Castle that is perfect for a rainy day." The recommendation means is configured to display the generated course name text CS, the destination information DN, and the travel route information RT on the display unit 111 of the user terminal 110. More specifically, as shown in Figures 9(A) to 10, the recommendation means displays the generated course name text CS, such as "Temple of XX Course," "Sweets and Walking Course," or "Shopping Course in the Morning Sky," as well as the destination information DN and travel route information RT on the recommendation display screen 115 of the display unit 111 of the user terminal 110.

[0072] As a result, when the destination information DN and travel route information RT are displayed as suggestions on the user terminal 110, a generation reason explanation text RE based on the destination information DN and travel route information RT is displayed, and the generated course name text CS is displayed as a consistent theme title. As a result, users can easily understand the appeal of the proposed content from the set of the generated course name text CS, which is a consistent theme title, and the generation reason explanation text RE. Users are then more likely to be motivated to take action based on the suggestions. As a result, users can feel a sense of excitement from the moment they start taking action.

[0073] In addition, in this embodiment, when the recommendation means generates course name text data for the course name using the pre-trained artificial intelligence means, it generates a story text with an introduction, development, twist, and conclusion using the pre-trained artificial intelligence means based on the selected destination information DN and travel route information RT by the travel means, and is configured to display the generated story text SR on the display unit 111 of the user terminal 110.

[0074] More specifically, as shown in FIG. 11, the user attribute information US includes information such as "resident of Nagoya city, 34 years old, office work, accounting, arrives at work at 10:00 on weekdays, arrives at Nagoya Station at 8:30 in the morning, and usually goes straight to work." Then, it is assumed that the recommendation means generates a generated course name text CS to the effect of "A wonderful hidden course around Nagoya Castle that is perfect for a rainy day" using the pre-trained artificial intelligence means. As an example, the recommendation means uses a pre-trained artificial intelligence means to generate a generated story text SR with the following content: "Introduction: Take the subway from Chikusa Station near your home to Nagoya Castle Station without getting wet," "Development: Since it is raining, go to the observation deck of Nagoya Castle, which is relatively empty even on weekends," "Turnaround: Enjoy the unique atmosphere of the Japanese garden in the rain and have a meal at a teahouse," and "Conclusion: Take a short walk and appreciate the sculpture exhibition at the Nagoya City Art Museum." The generated story text SR is displayed, for example, in a detail display window WD of the recommendation display screen 115 shown in FIGS.

[0075] As a result, the generated story text SR with a specific beginning, development, twist and conclusion, which is fitted with not only the theme title of the data of the generated course name text CS and the text explaining the reason for generation RE, but also the destination information DN and the travel route information RT by means of transportation, is displayed on the user terminal 110. As a result, the user can understand the relationship between the reason for the proposal and the story of the proposal content from the set of the generation reason explanation text RE and the generated story text SR, and can gain a deeper understanding of the appeal of the proposal content.

[0076] In this embodiment, the recommendation means selects a plurality of sets of destinations and means of transportation that are in line with the sentences in the activity theme AT from the database 121 of the server 120 . Furthermore, the recommendation means uses a pre-trained artificial intelligence means to generate a text explaining the reason for the suggestion, a course name text about the course name, and a story text with an introduction, development, twist, and conclusion based on the multiple sets of selected destination information DN and travel route information RT by the means of travel. As shown in Figures 9(A) to 10, the recommendation means may be configured to display each generation reason explanation text RE, generated course name text CS, generated story text SR, destination information DN, and travel route information RT on the recommendation display screen 115 of the display unit 111 of the user terminal 110.

[0077] As a result, a plurality of generated course name texts CS, generated story texts SR, and generation reason explanation texts RE are displayed on the user terminal 110 as suggestions. As a result, the user can compare and select the generated course name text CS, generated story text SR, and generated reason explanation text RE that interest him / her from the options. That is, not only a plurality of destinations and travel routes but also a plurality of generated course name texts CS, generated story texts SR, and generation reason explanation texts RE are displayed. As a result, rather than simply comparing destinations, users can compare and select the destinations and means of transportation that are in line with the action theme AT by comparing the excitement gained from the generated course name text CS, generated story text SR, and generated reason explanation text RE.

[0078] Furthermore, in this embodiment, as shown in Figures 9(A) to 10, when the destination information DN and travel route information RT are displayed on the recommendation display screen 115 of the display unit 111 of the user terminal 110, an icon IC for the destination and the travel route are displayed on the map. The content of the icon IC may be configured to include story content relating to the destination corresponding to the icon IC in the generated story text SR. Regarding the contents of the icon IC, the recommendation means may generate them using an image generation means that generates image data such as photographs, illustrations, and videos, which is an example of a pre-trained artificial intelligence means, or may be configured to display the icon IC based on image data on the Internet. This allows the contents of the icon IC to indicate what is relevant to the content of the generated story text SR at the destination. As a result, users can visually and easily understand the relationship between the content of the generated story text SR and the destination.

[0079] In this embodiment, when the icon IC is selected, the user terminal 110 may be configured to display on the recommendation display screen 115 the story content relating to the destination corresponding to the icon IC in the generated story text SR. As a result, a part corresponding to the introduction, development, turn, and conclusion icons IC in the generated story text SR is displayed. As a result, users can easily understand the story's beginning, development, twist, and conclusion by switching between icon IC selections, and can gain a deeper understanding of the appeal of the proposed content by matching it with the order of the story. It should be noted that the recommendation means may be configured to generate image data for the content of the generated story text SR using the image generation means, and display the image data on the recommendation display screen 115 or the detailed display window WD.

[0080] Furthermore, in this embodiment, as shown in FIG. 12, when an operation to start an arbitrary application is performed on the user terminal 110 as an example of a predetermined operation, the user terminal 110 starts the arbitrary application and displays the arbitrary application screen 116. At this time, in the user terminal 110 or the server 120, when the recommendation means detects a change of more than a predetermined early timing change amount in the operation of launching an arbitrary application as an example of a predetermined operation for daily log information in the user terminal 110, the recommendation means is configured to generate a sentence about the behavioral theme AT using the pre-trained artificial intelligence means based also on the change amount of the operation of launching the arbitrary application.

[0081] More specifically, when a change equal to or greater than a predetermined early timing change amount is detected in the launch operation of an arbitrary application as an example of a predetermined operation, the user terminal 110 displays a pop-up PP on the display unit 111, for example, saying, "Hello! It's an hour earlier than usual! Would you like to take an hour to change your mood?" When this pop-up display PP is operated, it is determined that this is the predetermined operation for requesting a recommendation in step S1 described above, and the recommendation means generates a sentence about the action theme AT. Subsequently, the user terminal 110 or the server 120 executes steps S2 to S6.

[0082] As a result, if an operation to start an arbitrary application is performed at a timing earlier than the normal time by a predetermined time or more based on the user's daily log, it is determined that free time has occurred, and an action theme AT is generated based on the amount of change in the operation to start the arbitrary application, and a destination and means of transportation are selected and displayed as a proposal on the user terminal 110 together with the generated course name text CS. As a result, users can make effective use of free time in their schedules.

[0083] In this embodiment, as shown in FIG. 13, it is assumed that an operation to arbitrarily change, delete, or cancel a schedule on the schedule screen 117 is performed as an example of a predetermined operation on the user terminal 110. At this time, in the user terminal 110 or the server 120, the recommendation means searches for free time information in the user's schedule information, and if free time information is detected, it is configured to generate a sentence about the behavioral theme AT using a pre-trained artificial intelligence means based on the length of the free time information.

[0084] More specifically, an example of a predetermined operation is an operation to arbitrarily change, delete, or cancel a schedule on the schedule screen 117, and the recommendation means searches for available information in the user's schedule information, and when available information is detected, the user terminal 110 displays a pop-up PP on the display unit 111, for example, saying, "Hello! You have about two hours free! Would you like to take a break for about 1.5 hours?" When this pop-up display PP is operated, it is determined that this is the predetermined operation for requesting a recommendation in step S1 described above, and the recommendation means generates a sentence about the action theme AT. Subsequently, the user terminal 110 or the server 120 executes steps S2 to S6.

[0085] As a result, when a free time occurs in the user's schedule, an action theme AT is generated based on the length of free time information, a destination and a means of transportation are selected, and are displayed as a proposal on the user terminal 110 together with the generated course name text CS. As a result, users can make effective use of free time in their schedules.

[0086] The recommendation system 100 thus obtained as an embodiment of the present invention comprises a user terminal 110 and a server 120, and when a predetermined operation for requesting a recommendation is performed on the user terminal 110, a recommendation means in the user terminal 110 or the server 120 generates a sentence about the activity theme AT using pre-trained artificial intelligence means based on the user's attribute information US input in advance, situation information ST which is input or acquired current user situation information, and external environment information EF, selects a destination and a means of transportation that are in line with the sentence about the activity theme AT from a database 121 of the server 120, evaluates negative elements from the selected destination information DN and travel route information RT using the travel route information RT, and the information used to generate the sentence about the activity theme AT, and performs a recommendation process using the pre-trained artificial intelligence means based on the selected destination information DN and travel route information RT using the travel route information RT. When artificial intelligence means is used to generate a reason explanation text for recommending the selected destination and travel route, the reason explanation text is generated based on a prompt to generate a reason explanation text that includes negative elements and changes from negative elements to positive, and the generated reason explanation text RE, destination information DN, and travel route information RT are displayed on the user terminal 110.This allows the user to easily understand, trust, and sympathize with the reasons for the proposal of the proposed content, which includes ways to enjoy and highlights in accordance with the negative elements, from the generated reason explanation text RE, which is the reason for the recommendation, and allows the user to feel more convinced and more receptive, allowing the user to have an unforgettable and memorable experience and feel satisfied, and the user to feel that it is a dramatic experience that is at least somewhat conscious of being wonderful, and feel satisfied.

[0087] Furthermore, the recommendation means evaluates any of the following as negative factors based on the pre-input user attribute information US, the situation information ST which is the input or acquired current user status information, and the acquired external environment information EF: that the weather information indicates bad weather; that the selected destination information DN and the travel route information RT by the travel means indicates that the waiting time at the destination is longer than a specified time, that there are restrictions on payment methods, that there is construction work and the appearance is not as usual, or that there is a slope or staircase climbing distance longer than a specified distance on the travel route; and generates a reason explanation text including the evaluated negative factors, thereby allowing the user to feel more convinced.

[0088] Furthermore, when the recommendation means selects multiple sets of destinations and means of transportation that are in line with the text of the behavioral theme AT from the database 121 of the server 120, and displays the destination information DN and travel route information RT on the user terminal 110 based on the multiple sets of selected destination information DN and travel route information RT by the travel means, a destination icon IC is placed on the map and a predetermined mark icon ICa is displayed in a position adjacent to the destination icon IC, and when the predetermined mark icon ICa is operated, the content of the flow that changes from negative elements to positive elements based on the generation reason explanation text RE is displayed, so that the user can easily visually understand the relationship between the destination and the content of the flow that changes from negative elements to positive elements.

[0089] Furthermore, when the recommendation means uses a pre-trained artificial intelligence means to generate a text explaining the reasons for recommending the selected destination and travel route in a flow that changes from negative elements to positive, the text explaining the reasons is generated based on a prompt that the positive content is content obtained based on the negative elements. This allows the user to understand the relationship between the negative elements and the positive content and feel even more convinced. For example, the positive content may be obtained based on the negative elements, such as "It looks like it's raining, but because it's raining, the observation deck is relatively empty, and the Japanese garden near the observation deck is filled with not only colorful flowers but also umbrellas, so you can enjoy a slightly different atmosphere on rainy days than on sunny days, so I recommend it." The user can understand the relationship between the negative elements and the positive content and feel even more convinced.

[0090] In addition, when a change or deletion operation is performed on the schedule information on the user terminal 110, the recommendation means on the user terminal 110 or the server 120 searches for free time information in the user's schedule information, and if free time information is detected, the recommendation means uses a pre-trained artificial intelligence means to generate a sentence about the action theme AT based on the length of the free time information, thereby allowing the user to make effective use of free time in their schedule.

[0091] Furthermore, when an operation to launch an arbitrary application is performed on the user terminal 110, if the recommendation means on the user terminal 110 or the server 120 detects a change in the operation to launch an arbitrary application relative to the daily log information on the user terminal 110 that is greater than a predetermined early timing change amount, the recommendation means generates a sentence about the behavioral theme AT using a pre-trained artificial intelligence means based on the change amount of the operation to launch the arbitrary application, thereby allowing the user to make effective use of free time in their schedule.

[0092] Furthermore, the program of the recommendation system 100 according to the embodiment of the present invention includes the following steps: when a predetermined operation for requesting a recommendation is performed on the user terminal 110, the recommendation means in the user terminal 110 or the server 120 generates a sentence about the action theme AT using pre-trained artificial intelligence means based on the user's attribute information US input in advance, situation information ST which is input or acquired current user situation information, and external environment information EF; a destination and means of transportation selection step S3 which selects a destination and means of transportation that are consistent with the sentence about the action theme AT from the database 121 of the server 120; a negative element evaluation step S4 which evaluates negative elements from the selected destination information DN and travel route information RT by the travel means, and the information used to generate the sentence about the action theme AT; and a negative element evaluation step S5 which uses the pre-trained artificial intelligence means to select a destination and means of transportation that are consistent with the sentence about the action theme AT based on the selected destination information DN and travel route information RT by the travel means, and the information used to generate the sentence about the action theme AT. When generating a reason explanation text for recommending a selected destination and travel route, a reason explanation text generation step S5 is executed to generate a reason explanation text based on a prompt to generate a reason explanation text that includes negative elements and changes from negative elements to positive elements, and a recommendation display step S6 is executed to display the generated reason explanation text RE, destination information DN, and travel route information RT on the user terminal 110.By doing so, the user can easily understand, trust, and sympathize with the reason for the recommendation, including ways to enjoy and highlights that correspond to the negative elements, from the generated reason explanation text RE, which is the reason for the recommendation, and can feel more convinced and more receptive to the user.The user can have an unforgettable and memorable experience and feel satisfied.The effects are enormous, such as the user feeling that it is a dramatic experience that is at least somewhat conscious of nice things and feeling satisfied. [Explanation of symbols]

[0093] 100 ··· Recommendation System 110 User terminal 111... Display section 112 ··· My Page screen 113 ··· Your Mood Screen 114 ··· Search screen 114a··· "Leave it to AI" button 115 ··· Recommendation display screen 116 ··· Any app screen 117 ··· Schedule screen 120 Server 121 ··· Database PP ··· Pop-up display US ··· User demographic information ST: Situation information (situation information, physical condition information, mood information) LK...Preference information RT ··· Travel route information DN ··· Destination Information IC icon ICa ··· Predefined mark icon AT ··· Action Theme CS ··· Generate course name text SR ··· Generated story text TP User type trend information RE... Explanation of the reason for creation WD...Detailed View Window EF: External environmental information (external environmental data, weather information, traffic congestion information, traffic information)

Claims

1. A recommendation system that includes a user terminal and a server and recommends destinations to a user, When a predetermined operation requesting a recommendation is performed on the user terminal, a recommendation means in the user terminal or server uses pre-trained artificial intelligence means to generate a sentence about an action theme based on at least the pre-input user attribute information among pre-input user attribute information, input or acquired current user situation information, and acquired external environment information, selects a destination and means of transportation that are in line with the sentence about the action theme from the server's database, evaluates negative elements from the selected destination information and travel route information by the travel means, and the information used to generate the sentence about the action theme, and when using the pre-trained artificial intelligence means to generate a reason explanation text for recommending the selected destination and travel route based on the selected destination information and travel route information by the travel means, generates the reason explanation text based on a prompt to generate a reason explanation text that includes negative elements and changes from negative elements to positive elements, and displays the generated reason explanation text, destination information, and travel route information on the user terminal.

2. The recommendation system of claim 1, characterized in that the recommendation means evaluates any of the following as negative factors from at least the acquired external environmental information out of the pre-input user attribute information, the input or acquired current user situation information, and the acquired external environmental information: weather information indicating bad weather; selected destination information and travel route information by travel means indicating that there will be a waiting time at the destination of a predetermined time or more; that there are restrictions on payment methods; that there is construction work and the appearance is not as usual; or that there is a slope or staircase climbing of a predetermined distance or more on the travel route; and generates a reason explanation text including the evaluated negative factors.

3. The recommendation system described in claim 2 is characterized in that the recommendation means selects multiple sets of destinations and means of transportation that are in line with the sentence on the activity theme from the server's database, and when the destination information and travel route information by the selected multiple sets are displayed on the user terminal based on the destination information and travel route information by the travel means, a destination icon is placed on the map and a predetermined mark icon is displayed in a position adjacent to the destination icon, and when the predetermined mark icon is operated, the content of the flow that changes from negative elements to positive elements based on the text explaining the reason for creation is displayed.

4. The recommendation system described in claim 3, characterized in that the recommendation means uses a pre-trained artificial intelligence means to generate a reason explanation text for recommending a selected destination and travel route with content that changes from negative elements to positive elements, and generates the reason explanation text based on a prompt that the positive content is content that is obtained based on the negative elements.

5. A recommendation system as described in any one of claims 1 to 4, characterized in that when a change or deletion operation is performed on schedule information on the user terminal, a recommendation means on the user terminal or server searches for free time information in the user's schedule information, and if free time information is detected, generates a sentence about the activity theme using a pre-trained artificial intelligence means based on the length of the free time information.

6. A recommendation system as described in any one of claims 1 to 4, characterized in that when an operation to launch an arbitrary application is performed on the user terminal, a recommendation means in the user terminal or server detects a change in the operation to launch an arbitrary application that is greater than a predetermined early timing change amount relative to the daily log information on the user terminal, and generates a sentence about the behavioral theme using a pre-trained artificial intelligence means based also on the change amount of the operation to launch the arbitrary application.

7. A program for a recommendation system that recommends destinations to users, an action theme generation step in which, when a predetermined operation for requesting a recommendation is performed on the user terminal, a recommendation means in the user terminal or the server generates a sentence about an action theme using a pre-trained artificial intelligence means based on at least the pre-input user attribute information among pre-input user attribute information, input or acquired current user situation information, and acquired external environment information; a destination and transportation means selection step of selecting a destination and transportation means according to the sentence of the activity theme from a database of the server; a negative element evaluation step of evaluating negative elements based on the selected destination information, travel route information by the means of transportation, and information used to generate sentences about the activity theme; a reason explanation text generation step of generating a reason explanation text based on a prompt to generate a reason explanation text including negative elements and a flow of content that changes from negative elements to positive elements when generating a reason explanation text for recommending the selected destination and travel route using a pre-trained artificial intelligence means based on the selected destination information and travel route information by the travel means; and a recommendation display step of displaying a text explaining the reason for creation, destination information, and travel route information on a user terminal.

Images