JP7396762B1 - Information processing system, information processing method, and program - Google Patents

Abstract

The present invention proposes actions based on data input by a user. An information processing system includes a first generation request acquisition unit that acquires first generation request information for generating predetermined content from a user device, a generation request generation rule, a generation request evaluation rule, and a first generation request information. a second generation request generation unit that generates second generation request information that is evaluated to be of higher quality than the first generation request information based on the generation request information; a content acquisition unit that acquires second content generated based on content and second generation request information; a content evaluation unit that evaluates first content and second content based on content evaluation rules; and a content evaluation unit that evaluates first content and second content based on content evaluation rules. The present invention includes an action proposal generation unit that generates action proposal information regarding a proposal of actions related to the first generation request information to the user based on the information, and an action proposal output unit that outputs the action proposal information. [Selection diagram] Figure 1

Description

The present invention relates to an information processing system, an information processing method, and a program.

2. Description of the Related Art In recent years, the use of artificial intelligence (AI) has progressed, and techniques for suggesting to users appropriate ways to use artificial intelligence are known.

For example, the AI service usage support system described in Patent Document 1 supports the usage of an AI service system. The AI service usage support system receives response data by sending inquiry data to the AI service system to be evaluated, and uses the correct answer data corresponding to the inquiry data and the response data to obtain accuracy information of the AI service system. It has an accuracy evaluation processing unit that calculates.

JP 2019-3267 Publication

However, in the AI service usage support system described in Patent Document 1, the accuracy of the AI service system is evaluated by evaluating the output results of the AI service system. Therefore, if the user is unable to obtain the desired content by inputting data into the first AI service system, the AI service usage support system proposes actions (for example, suggestions for correction) regarding the data input by the user. This is not taken into consideration, and the user needs to use the AI service usage support system to search for a second AI service system with higher accuracy.

Therefore, an object of the present invention is to propose actions in response to data input by a user.

An information processing system according to one aspect of the present invention includes a first generation request acquisition unit that acquires first generation request information for generating predetermined content from a user device, and a generation request for generating generation request information. Generate second generation request information that is evaluated to be of higher quality than the first generation request information based on the generation rule, the generation request evaluation rule for evaluating the quality of the generation request information, and the first generation request information. a second generation request generation unit; a content acquisition unit that acquires the first content generated based on the first generation request information; and the second content generated based on the second generation request information; a content evaluation unit that evaluates each of the first content and the second content based on the content evaluation rules of It includes an action suggestion generation section that generates action suggestion information, and an action suggestion output section that outputs the action suggestion information.

According to the present invention, it is possible to propose an action based on data input by a user.

1 is a diagram showing an overview of processing in an information processing system 100 that is an embodiment of the present invention. 1 is a diagram showing the configuration of an information processing system 100 that is an embodiment of the present invention. 5 is a diagram illustrating an example of evaluation result information stored in a storage unit 110. FIG. 3 is a flowchart illustrating an example of processing in the information processing system 100. FIG. 3 is a flowchart illustrating an example of processing in the information processing system 100. FIG. 3 is a flowchart illustrating an example of processing in the information processing system 100. FIG. 7 is a diagram showing an example of the hardware configuration of a computer 700. FIG.

Preferred embodiments of the present invention will be described with reference to the accompanying drawings. FIG. 1 is a diagram showing an overview of processing in an information processing system 100 that is an embodiment of the present invention.

The information processing system 100 acquires first generation request information from a user device, generates second generation request information based on the first generation request information, and generates first content and content generated based on the first generation request information. The information processing system evaluates the second content generated based on the second generation request information, and generates and outputs behavior proposal information regarding the behavior proposal related to the first generation request information.

In recent years, users obtain predetermined content by providing generation request information (for example, a prompt), which will be described later, to a predetermined information processing system (for example, a content generation device 300, which will be described later). If the desired content cannot be obtained, the user may modify the generation request information and provide the modified generation request information to a predetermined information processing system anew. That is, in order to obtain desired content, the user needs to invest resources (eg, time and effort) in repeatedly creating, modifying, and providing generation request information. However, it may be difficult for the user to fully understand the content generation algorithm in the content generation device 300. Therefore, even with the modified generation request information, content desired by the user may not be generated, and lower quality content may be generated based on the modified generation request information. In such a case, the resources invested by the user are wasted.

In this way, acquiring content through generation request information does not necessarily improve the quality of the content that can be acquired in accordance with the input resources. As a result, users have the vague expectation that they will be able to obtain higher quality content if they increase their input resources, but at the same time they are faced with the risk that the quality of the content will not improve even if they increase their input resources. , provides generation request information to the content generation device 300. If content acquisition through generation request information is viewed as an act of investment, in which content is obtained as a return based on the investment of resources into generating, modifying, and providing generation request information, at present, users are It can be said that they are in a situation where they have no choice but to gamble with their time and effort. This can be said to be a situation in which the user is unable to make rational decisions due to low investment efficiency or lack of information for making investment decisions.

Therefore, the information processing system 100 generates second generation request information having a higher quality than the first generation request information created by the user based on the generation request generation rule and the generation request evaluation rule,
By evaluating the quality of the content generated based on each of the first generation request information and the second generation request information, the user can determine whether it is desirable to modify the generation request information or not to modify it. etc., and proposes actions related to the generation request information to the user. That is, for example, if it is determined that the return obtained in response to the resource input of modifying the generation request information is small, the information processing system 100 sends action suggestion information that suggests proceeding with the process without modifying the generation request information. Alternatively, if it is determined that the return obtained by investing resources to modify the generation request information is large, the generation request information may be modified and processed, or the information processing system 100 may process the generated request information. 2. Action suggestion information that suggests proceeding with the process may be generated based on the generation request information.

Specifically, first, the information processing system 100 acquires the first generation request information 101 from the user device 200. The information processing system 100 generates second generation request information 102 based on the first generation request information.

Subsequently, the information processing system 100 acquires the first content 103 generated based on the first generation request information and the second content 104 generated based on the second generation request information. Then, the information processing system 100 evaluates the first content 103 and the second content 104, generates behavior proposal information 105 regarding the behavior proposal related to the first generation request information, and outputs it to the user device 200.

Thereby, the information processing system 100 can control the user's expectations regarding the return obtained in response to resource input. In other words, by investing resources in modifying the generation request information, the user can grasp in advance the level of expectation regarding the quality improvement of the content that can be generated before modifying the generation request information, so the user's level of expectation can be controlled. can do. Furthermore, if it is determined that the quality of content that can be generated will not improve, the user can avoid wasting resources, that is, modifying the generation request information. Furthermore, the user can make rational investment decisions based on the processing results of the information processing system 100, rather than repeatedly modifying the generation request information like gambling, which increases the user's sense of self-efficacy. be able to. Furthermore, by avoiding the input of easy generation request information to the content generation device 300, it is expected that the power consumption of the content generation device 300 will be reduced and the processing speed of the content generation device 300 will be improved. .

FIG. 2 is a diagram showing the configuration of an information processing system 100 that is an embodiment of the present invention. The information processing system 100 is communicably connected to a user device 200 and a content generation device 300 via a network such as the Internet. Details of the information processing system 100 will be described later.

The user device 200 is an information processing device used by a user of the information processing system 100, and is a computer, a smartphone, a tablet terminal, a personal computer, or the like.

The user accesses the information processing system 100 using the user device 200 and provides the information processing system 100 with the first generation request information. Further, the user accesses the information processing system 100 using the user device 200 and obtains action suggestion information from the information processing system 100.

The content generation device 300 is an information processing device that generates content based on generation request information, and includes, for example, a server device that provides so-called generative AI (Generative Artificial Intelligence).

The content generation device 300 may be, for example, an information processing device that receives generation request information as input and outputs content based on natural language processing technology and machine learning technology. The content generation device 300 may input the generation request information and output the content based on a large-scale language model, for example.

The generation request information (for example, first generation request information, second generation request information, and third generation request information, which will be described later) is, for example, a command statement for the content generation device 300 to cause the content generation device 300 to generate content. (e.g., a prompt). The generation request information may include, for example, text data, image data, video data, or audio data.

The content generated by the content generation device 300 is not particularly limited, but may include, for example, text data, image data, video data, or audio data.

Note that although one user device 200 and one content generation device 300 are shown in FIG. 2, there may be a plurality of user devices 200 and a plurality of content generation devices 300, respectively.

Next, details of the information processing system 100 will be explained. The information processing system 100 includes a storage unit 110, a first generation request acquisition unit 120, a user information acquisition unit 130, a second generation request generation unit 140, a content acquisition unit 150, a content evaluation unit 160, a user score calculation unit 170, and a behavior proposal. It includes a generation section 180 and an action suggestion output section 190. Each unit shown in FIG. 2 can be realized, for example, by using a storage area or by having a processor execute a program stored in the storage area.

The storage unit 110 stores information processed in the information processing system 100. The storage unit 110 can store, for example, user information, first generation request information, second generation request information, first content, second content, evaluation result information, user score, and action suggestion information, which will be described later. .

The first generation request acquisition unit 120 acquires first generation request information for generating predetermined content (first content) from the user device 200 and stores the acquired first generation request information in the storage unit 110. do.

The first content may be, for example, "a proposed leaflet encouraging people to undergo cancer screening." In this case, the first generation request information may be, for example, "generate 10 types of leaflet proposals to encourage people to undergo cancer screening."

The user information acquisition unit 130 acquires user information regarding the user and stores the acquired user information in the storage unit 110.

The user information includes, for example, attribute information regarding user attributes. The attribute information may include, for example, information regarding the user's age, occupation, affiliation (for example, company name or department name), or position.

Further, the user information includes, for example, behavior information regarding user behavior based on behavior suggestion information described below. Behavior information will be described later.

The second generation request generation unit 140 generates a first generation request based on a generation request generation rule for generating generation request information, a generation request evaluation rule for evaluating the quality of generation request information, and the first generation request information. Second generation request information that is evaluated to be of higher quality than the request information is generated, and the generated second generation request information is stored in the storage unit 110.

Generation request information that can generate appropriate content includes, for example, a plurality of predetermined elements. The plurality of elements include, for example, "giving a position to the content generation device 300," "conveying information that is not desired to be included in the generated content," and "content size (for example, number of characters, amount of data)." "setting", "showing an example to the content generation device 300", and "clarifying the intention and background of the question". The generation request generation rule is, for example, a rule corresponding to the plurality of elements.

For example, the generation request evaluation rule may be a rule that evaluates the quality of the generation request information as being high if it is within an appropriate amount of data depending on the amount of data (for example, number of characters) of the generation request information. Alternatively, the rule may be such that the quality of the generation request information is evaluated to be high if a predetermined element is included, depending on the presence or absence of a plurality of elements corresponding to the generation request generation rule.

The generation request generation rule and the generation request evaluation rule are based on, for example, the purpose of the generation request information, more specifically, the content to be generated based on the generation request information (specifically, the nature of the content, for example). Depending on the situation, different rules may be used. For example, if the content generated based on the generation request information is text data, the generation request generation rule and the generation request evaluation rule generate and evaluate the generation request information that can generate text data including more specific content. In addition, if the content generated based on the generation request information includes visual content (for example, in the case of flyers, etc.), the generation request generation rule and the generation request evaluation rule may be The rule may be one that can generate and evaluate generation request information that can generate content that appeals to the human's visual sense or content that takes universal devine into consideration.

For example, the second generation request generation unit 140 first evaluates the presence or absence of a plurality of elements corresponding to the generation request generation rule in the first generation request information based on the generation request generation rule and the first generation request information, and 1. Identify elements that are not included in the generation request information. Then, the second generation request generation unit 140 generates second generation request information that supplements the missing elements and is evaluated to be of higher quality than the first generation request information based on the generation request evaluation rule. The second generation request information is generated. Thereby, the second generation request generation unit 140 can generate second generation request information that can generate more appropriate content.

Specifically, for example, when the first generation request information is "generate 10 types of leaflet proposals to encourage people to undergo cancer screening," the second generation request generation unit 140 first generates, for example, the first generation request. The element "setting content size (for example, number of characters, amount of data)" is specified as an element not included in the information. Then, the second generation request generation unit 140 generates second generation request information ``setting the content size (for example, number of characters, amount of data)'' supplemented with the element ``set the content size (for example, number of characters, amount of data)''. %, generate 10 types of A4 size poster leaflet proposals to encourage cancer screening.

At this time, the second generation request generation unit 140 generates second generation request information with elements not included in the first request information supplemented by, for example, referring to generation request information that the user has used in the past. You can.

Further, the second generation request generation unit 140 may generate a plurality of pieces of second generation request information based on one piece of first generation request information. Thereby, the information processing system 100 generates action suggestion information, which will be described later, based on variations in the generated plurality of second generation request information (for example, variations in the content of each of the plurality of second generation request information). be able to.

Further, the second generation request generation unit 140 can generate second generation request information further based on the user information. At this time, the second generation request generation unit 140 generates second generation request information in which, for example, elements not included in the first generation request information are supplemented based on the user information. More specifically, for example, if the user information indicates that the user's department is the legal department, the user may be required to use the It is thought that they are looking for content. Therefore, the second generation request generation unit 140 can generate second generation request information for giving the content generation device 300 the role of a "legal expert", for example.

The content acquisition unit 150 acquires the first content generated based on the first generation request information and the second content generated based on the second generation request information, and the acquired first content and second content. The information is stored in the storage unit 110.

The information processing system 100 outputs, for example, first generation request information and second generation request information to the content generation device 300. The content generation device 300 generates first content based on first generation request information, and generates second content based on second generation request information. Then, the content generation device 300 outputs the first content and the second content to the information processing system 100, and the content acquisition unit 150 acquires the first content and the second content.

The content evaluation unit 160 evaluates each of the first content and the second content based on content evaluation rules for evaluating the content, generates evaluation result information regarding the result of the evaluation, and stores it in the storage unit 110. .

For example, the content evaluation unit 160 evaluates that the quality of the content is low when the content (e.g., text) is general content based on content evaluation rules, and If the content is distinctive, the quality of the content may be evaluated as high. Here, whether or not the content is characteristic may be evaluated based on, for example, the proportion of proper nouns included in the content, that is, the proportion of proper nouns included in the content is high. The more distinctive the content, the more distinctive the content may be evaluated.

For example, if the content consists of multiple sentences, the similarity of the meaning of each sentence is low and the density of information is high, or the content has an appropriate proportion of kanji and hiragana and is highly readable. , the content is evaluated to be of high quality if it is appropriate to the context even when translated into another language, or if the content includes fonts, sizes, shapes, or colors that take into account universal design. You can.

The content evaluation rule may be, for example, correspondence information that is set in advance and indicates a correspondence between a plurality of elements for evaluating content and a weight for each element. Further, the content evaluation rule may be, for example, an evaluation model that inputs content and outputs a score indicating the evaluation result of the content.

The content evaluation unit 160 calculates, for example, a score indicating the evaluation result of each of the first content and the second content based on the content evaluation rule.

FIG. 3 is a diagram showing an example of evaluation result information stored in the storage unit 110. The evaluation result information stored in the storage unit 110 includes, for example, an evaluation ID, a first score, and a second score. The evaluation ID is, for example, information that identifies the evaluation result information stored in the storage unit 110. The first score and the second score are, for example, scores indicating the evaluation results of the first content and the second content, respectively. The first score and the second score may be, for example, quantitative scores expressed using numerical values, or may be qualitative scores (for example, "very good", "good", etc.).

The user score calculation unit 170 calculates the user score of the user who uses the user device 200, which is calculated based on the result of the evaluation by the content evaluation unit 160.

The user score calculated by the user score calculation unit 170 is, for example, a score indicating the user's ability to create generation request information for generating appropriate content. That is, the user score calculation unit 170 may calculate a higher user score as the first score calculated by the content evaluation unit 160 is higher. Further, the user score calculation unit 170 may calculate a higher user score, for example, based on the first score and the second score, for example, as the first score is higher than the second score.

The user score calculation unit 170 may calculate the user score, for example, according to the evaluation results for a plurality of contents generated based on each of a plurality of pieces of generation request information created by a user in the past. Thereby, the information processing system 100 can calculate a user score based on the creation record of the user's generation request information.

Further, the user score calculation unit 170 may calculate the user score, for example, according to the evaluation results for the content generated based on the generation request information created by another user. At this time, for example, the user score calculation unit 170 calculates the content created by the first user in the distribution of evaluations for the plurality of contents generated based on the generation request information created by the plurality of users including the predetermined first user. The user score may be calculated according to the statistical value (for example, the degree of deviation from the average value) of the evaluation of the content generated based on the generation request information. Thereby, the information processing system 100 can evaluate the generation request information generation ability of users in a predetermined group as a user score.

The behavior proposal generation unit 180 generates behavior suggestion information regarding a behavior proposal related to the first generation request information for the user using the user device 200 based on the evaluation result by the content evaluation unit 160, and generates behavior proposal information regarding a behavior proposal related to the first generation request information. The information is stored in the storage unit 110.

The action suggestion information generated by the action suggestion generation unit 180 includes, for example, action suggestion information that suggests proceeding with the process based on the first generation request information, and action suggestion information that suggests proceeding with the process based on the second generation request information. It may be at least one of proposal information and action proposal information that proposes modifying the first generation request information.

The action suggestion information that suggests proceeding with the process based on the first generation request information may, for example, prompt the user to press a "Continue" button on the user device 200 to prompt the user to proceed with the process based on the first generation request information. It may also include display information for display. At this time, the first content generated based on the first generation request information may be displayed on the user device 200 by pressing the "Continue" button by the user.

Further, the action suggestion information may include display information for displaying the first score. In addition, the action suggestion information displays information indicating whether the first generation request information is capable of generating content of a predetermined quality (for example, content whose evaluation result based on the content evaluation rule is equal to or higher than a predetermined threshold). It may also include display information for. As a result, the user viewing the display of the action suggestion information can use the first score and the first generation request information to determine whether or not it is possible to generate content of a predetermined quality, as a criterion for pressing the "Continue" button. At least one of the indicated information can be referred to.

The action suggestion information that suggests proceeding with the process based on the second generation request information may be used to display, for example, on the user device 200 a "refer to modification proposal" button that prompts the user to refer to the second generation request information. may also include display information. At this time, the second generation request information may be displayed on the user device 200 when the user presses the "Refer to modification proposal" button. This allows the user to use the second generation request information that is expected to generate more appropriate content. Further, the action suggestion information may include display information for displaying the second score. Further, at this time, the information processing system 100 may provide the second generation request information to the user on the condition that the user provides a predetermined consideration (for example, money, points, etc.). Thereby, the administrator of the information processing system 100 can receive benefits based on the provision of the second generation request information.

The action suggestion information that suggests modifying the first generation request information is, for example, display information for displaying on the user device 200 a "modify" button that prompts the user to modify the first generation request information. May include. At this time, a modification field for modifying the first generation request information may be displayed on the user device 200 by pressing the "modify" button by the user. The action suggestion information that suggests modifying the first generation request information may be, for example, information that does not include the content of the second generation request information (for example, the content of the prompt). Thereby, the information processing system 100 can prompt the user to voluntarily modify the first generation request information. Further, the action suggestion information may include display information for displaying a third score indicating the evaluation result of the third content that can be generated based on the third generation request information described later. In addition, the action suggestion information may be information indicating an element to be modified or added in the first generation request information based on the generation request evaluation rule (for example, ``Better generation results can be obtained by additionally inputting information regarding element A.'') It may also include display information for displaying ".").

Next, a specific example of the action proposal information generation process by the action proposal generation unit 180 will be described.

If the first score is higher than the second score, the action suggestion generation unit 180 can generate action suggestion information that suggests proceeding with the process based on the first generation request information, for example.

Further, when the second score is higher than the first score, the action suggestion generation unit 180 can generate action suggestion information that suggests proceeding with the process based on the second generation request information, for example.

Further, when the second score is higher than the first score, the action suggestion generation unit 180 can generate action suggestion information that suggests modifying the first generation request information, for example.

Further, the action suggestion generation unit 180 may generate action suggestion information based on the quality of the first generation request information and the quality of the second generation request information based on the generation request evaluation rule.

Specifically, if the quality of the first generation request information is evaluated to be higher than the quality of the second generation request information based on the generation request evaluation rule, the action proposal generation unit 180, for example, Action suggestion information that suggests proceeding with processing based on the information can be generated.

Furthermore, if the quality of the second generation request information is evaluated to be higher than the quality of the first generation request information based on the generation request evaluation rule, the action suggestion generation unit 180, for example, It is possible to generate action suggestion information that suggests proceeding with the process, or action suggestion information that suggests modifying the first generation request information.

Further, the action suggestion generation unit 180 may generate action suggestion information based on the degree of improvement of the first generation request information, which is evaluated based on the generation request evaluation rule. Here, the expected degree of improvement of the first generation request information is, for example, the first generation request information corresponding to the first generation request information in the third content corresponding to the third generation request information that can be generated by modifying the first generation request information. This is the degree of improvement in the quality of the content.

For example, when the quality of the second generation request information is evaluated to be higher than the quality of the first generation request information by a predetermined threshold value or more, the information processing system 100 evaluates that the degree of improvement in the first generation request information is high. Also, for example, if the quality of the second generation request information is higher than the quality of the first generation request information, but the difference is smaller than a predetermined threshold (in other words, the first generation request information and the second generation request information If the quality of the information is the same), it may be evaluated that the degree of improvement in the first generation request information is low.

Further, if the first generation request information is evaluated as having a high degree of improvement based on the generation request evaluation rule, the action proposal generation unit 180 proposes, for example, proceeding with the process based on the second generation request information. It is possible to generate action suggestion information that suggests modifying the first generation request information or action suggestion information that suggests modifying the first generation request information. Thereby, the information processing system 100 allows the user to select the second generation request information or the first generation request information, so to speak, when generation request information that can generate content of higher quality than the first content is assumed. Can suggest modifications to the requested information.

Furthermore, if the first generation request information is evaluated as having a low possibility of improvement based on the generation request evaluation rule, the action proposal generation unit 180 proposes, for example, proceeding with the process based on the first generation request information. It is possible to generate action suggestion information. As a result, the information processing system 100 can, for example, when the quality of the first generation request information and the second generation request information are the same, the quality of the first generation request information is lower than the quality of the second generation request information. Even in such a case, it is possible to determine that there is no major problem with the quality of the first content and generate action suggestion information that suggests proceeding with the process based on the first generation request information.

Further, the behavior suggestion generation unit 180 can generate behavior suggestion information further based on the user score.

Specifically, for example, when the user score is higher than a predetermined threshold and the second score is higher than the first score, the action suggestion generation unit 180 generates an action that suggests modifying the first generation request information. Proposal information can be generated. Thereby, the information processing system 100 can generate action suggestion information according to the user score. For example, when the user score is high, that is, when the user's ability to create and modify the generation request information is high, , the user can be prompted to voluntarily modify the first generation request information.

Further, for example, when the user score is lower than a predetermined threshold and the second score is higher than the first score, the action suggestion generation unit 180 performs processing based on the first generation request information or the second generation request information. Action suggestion information that suggests proceeding can be generated. Thereby, the information processing system 100 can generate action suggestion information according to the user score. For example, when the user score is low, that is, when the user's ability to create generation request information is low, It is possible to judge that the possibility of modification of the first generation request information is low and proceed with the process based on the first generation request information, or to inform the user of the second generation request information that can be evaluated as an example of modification, so to speak, and to proceed with the process based on the first generation request information. It is possible to propose proceeding with the process based on the generation request information.

Further, the action suggestion generation unit 180 can generate action suggestion information further based on the generation request evaluation rule and the user score. At this time, the action suggestion generation unit 180 can generate action suggestion information based on, for example, the degree of improvement of the first generation request information evaluated based on the generation request evaluation rule and the user score.

Specifically, in the case where the first generation request information is evaluated as having a low degree of improvement possibility based on the generation request evaluation rule, if the user score is equal to or higher than a predetermined threshold, the action suggestion generation unit 180: For example, action suggestion information that suggests modifying the first generation request information can be generated. As a result, the information processing system 100 determines that a user with a user score equal to or higher than a predetermined threshold is highly likely to modify the first generation request information, and requests the user to voluntarily modify the first generation request information. , it is possible to prompt the user to modify the first generation request information.

Further, in the case where the first generation request information is evaluated as having a low possibility of improvement based on the generation request evaluation rule, if the user score is less than or equal to a predetermined threshold, the action suggestion generation unit 180, for example, Action suggestion information that proposes proceeding with the process based on the first generation request information or action suggestion information that proposes proceeding with the process based on the second generation request information can be generated. As a result, the information processing system 100 takes into account the possibility that the quality of the content may deteriorate due to modification of the first generation request information for a user with a user score below a predetermined threshold. It is possible to determine that the possibility of modifying the first generation request information is small and to proceed with the process as is with the first generation request information, or to proceed with the process based on the second generation request information.

Further, the action suggestion generation unit 180 generates a probability of improvement evaluated based on the user score from the first generation request information of the third generation request information that can be generated by modifying the first generation request information by the user. Further, action suggestion information can be generated based on the degree of likelihood of improvement.

Specifically, first, the action suggestion generation unit 180 generates third generation request information based on the first generation request information and the user score. At this time, the action proposal generation unit 180 predicts, so to speak, third generation request information that may be generated by the user modifying the first generation request information, based on the user score, and generates the third generation request information. generate.

Here, the action suggestion generation unit 180 may generate the third generation request information, for example, by transmitting and receiving data to and from the prompt generation device. The prompt generation device is an information processing device that generates generation request information different from the generation request information based on the generation request information, and includes, for example, a server device that provides so-called generative AI (Generative Artificial Intelligence). In this case, the action suggestion generation unit 180 may, for example, transmit the first generation request information to the prompt generation device and receive the third generation request information generated in the prompt generation device from the prompt generation device.

Further, the action suggestion generation unit 180 may, for example, set in advance the correspondence between the user score and the modification pattern of the generation request information, and transmit the correspondence and the first generation request information to the prompt generation device. . Thereby, the third generation request information can be predicted by, so to speak, simulating the modification by the user according to the user score.

Subsequently, the action proposal generation unit 180 evaluates the degree of improvement expected from the first generation request information based on the first generation request information and the third generation request information. At this time, the action suggestion generation unit 180 evaluates the quality of the first generation request information and the quality of the third generation request information based on the generation request evaluation rule, and makes the first generation request of the third generation request information. The likelihood of improvement based on the information may be evaluated. Specifically, when the quality of the third generation request information is higher than the quality of the first generation request information, it may be evaluated that the degree of improvement of the third generation request information from the first generation request information is high. Then, the action suggestion generation unit 180 can generate action suggestion information that suggests modifying the first generation request information when the degree of improvement possibility is high. Thereby, the information processing system 100 can, so to speak, predict the third generation request information that may be generated by the user, and generate the action suggestion information.

Further, the action suggestion generation unit 180 may generate action suggestion information based on the cost of modifying the first generation request information for generating the third generation request information, which is evaluated based on the user score. can. Here, the cost of modification is, for example, the time required to modify the first generation request information, or the amount of data (for example, number of characters) that is added, modified, or deleted when generating the third generation request information. It may be.

Specifically, the information processing system 100 may evaluate, for example, that the higher the user score, the lower the cost of modification, and the lower the user score, the higher the cost of modification. Then, for example, the higher the user score, the behavior suggestion generation unit 180 can generate the behavior suggestion information that suggests modifying the first generation request information, and the lower the user score, the more the first generation request information can be generated. Alternatively, it is possible to propose proceeding with the process based on the second generation request information.

Further, the action suggestion generation unit 180 can generate action suggestion information based on variations in the plurality of pieces of second generation request information generated by the second generation request generation unit 140.

Here, the variation in the plurality of second generation request information may be, for example, the variation in the content of each of the plurality of second generation request information. The variation in the content of each of the plurality of second generation request information may be, for example, the presence or absence of a plurality of elements corresponding to the generation request generation rule, or the amount of content (for example, the amount of characters) of each of the plurality of elements. Good too. The variation in content of each of the plurality of pieces of second generation request information may be calculated based on existing technology (for example, natural language processing technology).

If the plurality of pieces of second generation request information have large variations, the action suggestion generation unit 180 can generate, for example, action suggestion information that suggests modifying the first generation request information. As a result, the information processing system 100 determines that there is a high possibility of modifying the first generation request information when the dispersion of the plurality of second generation request information is large, and requests the user to voluntarily modify the first generation request information. It is possible to prompt the user to modify the generation request information. In addition, when the variation in the plurality of pieces of second generation request information is small, the action suggestion generation unit 180 generates, for example, action suggestion information that proposes proceeding with the process based on the first generation request information, or second generation request information. It is possible to generate action suggestion information that suggests proceeding with the process based on. As a result, when the variation in the plurality of pieces of second generation request information is small, the information processing system 100 determines that there is little possibility of modifying the first generation request information and continues processing with the first generation request information. Alternatively, it is possible to propose proceeding with the process based on the second generation request information.

Furthermore, the behavior suggestion generation unit 180 can generate behavior suggestion information based on the behavior information.

Specifically, the information processing system 100 first generates action suggestion information for the user. Subsequently, the user refers to the action suggestion information and performs an action based on the action suggestion information (for example, modifying the generation request information, not modifying it, etc.), and the information processing system 100 performs an action based on the action proposal information. Obtain behavioral information about user behavior based on. Then, the behavior suggestion generation unit 180 generates behavior suggestion information based on the behavior information. That is, the behavior suggestion generation unit 180 can generate behavior suggestion information based on the user's behavior performance based on the behavior suggestion information provided to the user in the past.

More specifically, for example, based on the action suggestion information, the action suggestion generation unit 180 may modify the first generation request information for a user who has a large track record of taking actions in accordance with the action suggestion information. You may generate action suggestion information that suggests.

Here, the behavioral information includes, for example, the time spent referring to the behavioral suggestion information, eye movements, cursor movement, presence or absence of operations such as clicks when referring to the behavioral suggestion information, and evaluation of the behavioral suggestion information. (For example, information regarding whether or not a "like" button has been clicked) may be used.

Furthermore, the behavior suggestion generation unit 180 can generate behavior suggestion information based on the behavior information of other users. Thereby, the information processing system 100 can generate action suggestion information in consideration of other users' action records.

Specifically, the behavior proposal generation unit 180 generates behavior information of other users with similar user scores (for example, the rate at which the first generation request information has been modified), and information regarding the quality of content acquired by the other users. (For example, the rate at which the content was improved by modifying the first generation request information, or the rate at which the content obtained by modifying the first generation request information was given a positive evaluation, such as a “like” rating) , may be generated as action suggestion information.

In this way, when content acquisition through generation request information is viewed as an act of investment, in which content is acquired as a return obtained in response to the input of resources into the generation, modification, and provision of generation request information, action proposals can be made. The generation unit 180 can be considered to generate action proposal information as information for making investment decisions.

In other words, the action proposal generation unit 180 makes the user's investment decision that the investment of resources for modifying the first generation request information or creating new generation request information is not commensurate with the return obtained in accordance with the resource input. For example, action suggestion information that suggests proceeding with the process can be generated based on the first generation request information. Furthermore, the action proposal generation unit 180 allows the user to understand that investing resources in utilizing the second generation request information is a more rational investment decision than investing resources in modifying the first generation request information. As information for making the determination, action suggestion information that suggests proceeding with the process can be generated based on the second generation request information. In addition, the action proposal generation unit 180 modifies the first generation request information as information for the user to determine that investing resources in modifying the first generation request information is a rational investment decision. Action suggestion information that suggests things to do can be generated. In other words, the action proposal generation unit 180 generates action proposal information by functioning as an ROI predictor that predicts the return on investment (ROI) that predicts the degree of return according to the user's resource input. be able to.

The behavior suggestion output unit 190 outputs behavior suggestion information.

The behavior suggestion output unit 190 can output behavior suggestion information to the user device 200, for example.

Further, the behavior suggestion output unit 190 may output behavior suggestion information including display information for displaying a button on the screen of the user device 200, for example. The user device 200 acquires the display information and displays, for example, a “Continue”, “Refer to modification proposal”, or “Modify” button on the screen of the user device 200.

FIG. 4 is a flowchart illustrating an example of processing in the information processing system 100.

First, the first generation request acquisition unit 120 acquires first generation request information (S401). The second generation request generation unit 140 generates second generation request information (S402). The content acquisition unit 150 acquires the first content and the second content (S403). The content evaluation unit 160 evaluates the first content and the second content (S404). The behavior suggestion generation unit 180 generates behavior suggestion information, and the behavior suggestion output unit 190 outputs the behavior suggestion information (S405).

FIG. 5 is a flowchart illustrating an example of processing in the information processing system 100. The flowchart shown in FIG. 5 shows an example of generation processing of second generation request information based on user information.

First, the user information acquisition unit 130 acquires user information (S501). Then, the second generation request generation unit 140 generates second generation request information further based on the user information (S502).

FIG. 6 is a flowchart illustrating an example of processing in the information processing system 100. The flowchart shown in FIG. 6 shows an example of the process of generating action suggestion information based on the user score.

First, the user score calculation unit 170 calculates a user score based on the results of the evaluation by the content evaluation unit 160 (S601). The behavior suggestion generation unit 180 generates behavior suggestion information based on the user score, and the behavior suggestion output unit 190 outputs the behavior suggestion information (S602).

Next, with reference to FIG. 7, an example of a hardware configuration when the information processing system 100 is implemented by the computer 700 will be described. FIG. 7 is a diagram showing an example of the hardware configuration of the computer 700.

As shown in FIG. 7, the computer 700 includes, for example, a processor 701, a memory 702, a storage device 703, an input I/F section 704, a data I/F section 705, a communication I/F section 706, and a display device 707. .

Computer 700 can be, for example, a server computer, a personal computer (e.g., desktop, laptop, tablet, etc.), a media computer platform (e.g., cable, satellite set-top box, digital video recorder, etc.), a handheld computer device (e.g., PDA, e-mail client, etc.) or other type of computer or communication platform.

The processor 701 is a control unit that controls various processes in the computer 700 by executing programs stored in the memory 702.

The memory 702 is, for example, a storage medium such as a RAM (Random Access Memory). The memory 702 temporarily stores program codes of programs executed by the processor 701 and data required when executing the programs.

The storage device 703 is a nonvolatile storage medium such as a hard disk drive (HDD) or flash memory. The storage device 703 stores an operating system and various programs for realizing each of the above configurations.

The input I/F unit 704 is a device for receiving input from the user. The input I/F unit 704 is, for example, a keyboard, a mouse, a touch panel, various sensors, a wearable device, or the like. Input I/F section 704 may be connected to computer 700 via an interface such as a USB (Universal Serial Bus).

Data I/F section 705 is a device for inputting data from outside of computer 700. The data I/F unit 705 is, for example, a drive device or the like for reading data stored in various storage media. Data I/F section 705 may be provided outside computer 700. When the data I/F unit 705 is provided outside the computer 700, the data I/F unit 705 is connected to the computer 700 via an interface such as a USB.

The communication I/F unit 706 is a device for performing wired or wireless data communication with a device external to the computer 700 via a network such as the Internet. Communication I/F section 706 may be provided outside computer 700. When the communication I/F unit 706 is provided outside the computer 700, the communication I/F unit 706 is connected to the computer 700 via an interface such as a USB.

The display device 707 is a device for displaying various information. The display device 707 is, for example, a liquid crystal display, an organic EL (Electro-Luminescence) display, a wearable device display, or the like. Display device 707 may be provided outside computer 700. When the display device 707 is provided outside the computer 700, the display device 707 is connected to the computer 700 via, for example, a display cable. Furthermore, when a touch panel is employed as the input I/F section 704, the display device 707 may be configured to be integrated with the input I/F section 704.

An embodiment of the present invention has been described above. The information processing system 100 acquires first generation request information from the user device 200, generates second generation request information based on the generation request generation rule, the generation request evaluation rule, and the first generation request information, and generates the first content and the second content, evaluate each, and generate and output action suggestion information based on the evaluation results. Thereby, the information processing system 100 can propose actions in response to data input by the user.

Further, the information processing system 100 can generate action suggestion information based on the quality of the first generation request information and the quality of the second generation request information based on the generation request evaluation rule. Thereby, the information processing system 100 can generate action suggestion information, for example, based on the degree of improvement expected in the first generation request information.

Furthermore, the information processing system 100 can calculate a user score based on the evaluation result, and generate action suggestion information further based on the user score. Thereby, the information processing system 100 can propose an action for the data input by the user, taking into account the user's ability to create the first generation request information.

The information processing system 100 also determines the likelihood of improvement evaluated based on the user score, and the third generation request information that can be generated by the user's modification of the first generation request information, from the first generation request information. Action suggestion information can be generated further based on the degree of likelihood of improvement. Thereby, the information processing system 100 can propose an action for the data input by the user, taking into account the user's ability to create the first generation request information and taking into account the return obtained from the user's resource investment. .

Furthermore, the information processing system 100 can generate action suggestion information based on the cost of modifying the first generation request information to generate the third generation request information, which is evaluated based on the user score. . Thereby, it is possible to propose an action for the data input by the user, taking into consideration the user's ability to create the first generation request information and taking into account the cost of modification by the user.

Furthermore, the information processing system 100 can acquire user information and generate second generation request information based on the user information. Thereby, the information processing system 100 can supplement the first generation request information based on the user information and generate the second generation request information that can generate higher quality content.

Further, the information processing system 100 acquires behavior information regarding the behavior of at least one first user based on the behavior suggestion information proposed to the at least one first user, and further based on the behavior information, the information processing system 100 generates behavior suggestion information. can be generated. Thereby, the information processing system 100 can learn the user's behavior based on the behavior suggestion information and generate the behavior suggestion information.

Note that this embodiment is provided to facilitate understanding of the present invention, and is not intended to be interpreted as limiting the present invention. The present invention may be modified/improved without departing from its spirit, and the present invention also includes equivalents thereof.

Furthermore, in the present invention, a "section" does not simply mean a physical means, but also includes a case where the function of the "section" is realized by software. Furthermore, even if the functions of one "section" or device are realized by two or more physical means, devices, or software, the functions of two or more "sections" or devices are realized by one physical means, device, or software. , or may be realized by software.

100 information processing system, 110 storage unit, 120 first generation request acquisition unit, 130 user information acquisition unit, 140 second generation request generation unit, 150 content acquisition unit, 160 content evaluation unit, 170 user score calculation unit, 180 action suggestion generation unit, 190 action suggestion output unit, 200 user device, 300 content generation device

Claims (9)

a first generation request acquisition unit that acquires first generation request information for generating predetermined content from a user device;
A generation request generation rule for generating generation request information that includes information regarding a plurality of elements to be included in the generation request information , the presence or absence of the plurality of elements in the generation request information, and at least one of the data amount of the generation request information. at least one element that is not included in the first generation request information among the plurality of elements based on a generation request evaluation rule for evaluating the quality of the generation request information according to the first generation request information, and the first generation request information . a second generation request generation unit that generates second generation request information that is supplemented second generation request information and whose evaluation based on the generation request evaluation rule is improved compared to the first generation request information;
a content acquisition unit that acquires a first content generated based on the first generation request information and a second content generated based on the second generation request information;
Each of the first content and the second content is evaluated based on a content evaluation rule for evaluating content , and a first score corresponding to the first content and a second score corresponding to the second content are determined. A content evaluation unit that calculates
an action proposal generation unit that generates action proposal information regarding a proposal of action related to the first generation request information to a user who uses the user device , based on a magnitude relationship between the first score and the second score ;
an action suggestion output unit that outputs the action suggestion information;
An information processing system equipped with. The action suggestion generation unit is configured to generate the first action proposal according to at least one of the presence or absence of the plurality of elements included in the generation request information and the amount of data of the generation request information, which are calculated based on the generation request evaluation rule. The information processing system according to claim 1, wherein the action suggestion information is generated further based on a magnitude relationship between an evaluation of the generation request information and an evaluation of the second generation request information. further comprising a user score calculation unit that calculates a user score indicating the ability to create generation request information for generating appropriate content based on a magnitude relationship between the first score and the second score,
The action suggestion generation unit generates the action suggestion information further based on a magnitude relationship between the user score and a predetermined threshold .
The information processing system according to claim 1. The information processing system calculates, based on the generation request evaluation rule, the degree to which the evaluation of the first generation request information based on the generation request evaluation rule is likely to be improved by modification of the first generation request information by the user. Calculated based on the magnitude relationship between the evaluation of the first generation request information and the evaluation of the second generation request information calculated by
The information processing system according to any one of claims 1 to 3, wherein the action suggestion generation unit generates action suggestion information further based on the magnitude relationship between the degree of likelihood of improvement and a predetermined threshold . The information processing system calculates the cost of modifying the first generation request information by the user based on the magnitude relationship between the user score and a predetermined threshold;
The information processing system according to claim 3 , wherein the action proposal generation unit generates the action proposal information further based on a magnitude relationship between the cost of the modification and a predetermined threshold . further comprising a user information acquisition unit that acquires first information regarding at least one of the plurality of elements and regarding the user;
The second generation request generation unit generates the second generation request information supplemented with at least one element not included in the first generation request information among the plurality of elements, further based on the first information. do,
The information processing system according to any one of claims 1 to 3 . further comprising a user information acquisition unit that acquires first information regarding the behavior of the at least one first user based on the behavior suggestion information proposed to the at least one first user;
The action proposal generation unit generates the action proposal information that proposes an action to modify the first generation request information when the number of actual actions of the first user indicated by the first information exceeds a predetermined threshold. ,
The information processing system according to any one of claims 1 to 3 . The computer is
Obtaining first generation request information for generating predetermined content from the user device,
A generation request generation rule for generating generation request information that includes information regarding a plurality of elements to be included in the generation request information , the presence or absence of the plurality of elements in the generation request information, and at least one of the data amount of the generation request information. at least one element that is not included in the first generation request information among the plurality of elements based on a generation request evaluation rule for evaluating the quality of the generation request information according to the first generation request information, and the first generation request information . Generate second generation request information that is the supplemented second generation request information and has an improved evaluation based on the generation request evaluation rule as compared with the first generation request information,
obtaining a first content generated based on the first generation request information and a second content generated based on the second generation request information;
Each of the first content and the second content is evaluated based on a content evaluation rule for evaluating content, and a first score corresponding to the first content and a second score corresponding to the second content are determined. Calculate,
generating action proposal information regarding a proposal of action related to the first generation request information to a user using the user device based on a magnitude relationship between the first score and the second score ;
outputting the action suggestion information;
Information processing method. to the computer,
a first generation request acquisition unit that acquires first generation request information for generating predetermined content from a user device;
A generation request generation rule for generating generation request information that includes information regarding a plurality of elements to be included in the generation request information , the presence or absence of the plurality of elements in the generation request information, and at least one of the data amount of the generation request information. at least one element that is not included in the first generation request information among the plurality of elements based on a generation request evaluation rule for evaluating the quality of the generation request information according to the first generation request information, and the first generation request information . a second generation request generation unit that generates second generation request information that is supplemented second generation request information and has an improved evaluation based on the generation request evaluation rule as compared with the first generation request information;
a content acquisition unit that acquires a first content generated based on the first generation request information and a second content generated based on the second generation request information;
Each of the first content and the second content is evaluated based on content evaluation rules for evaluating content , and a first score corresponding to the first content and a second score corresponding to the second content are determined. A content evaluation unit that calculates
an action proposal generation unit that generates action proposal information regarding a proposal of action related to the first generation request information to a user who uses the user device , based on a magnitude relationship between the first score and the second score ;
an action suggestion output unit that outputs the action suggestion information;
A program to make this happen.

Images