Management systems and management programs

Abstract

The purpose is to provide a management system and a management program that can provide useful information about a target. [Solution] The system comprises an acquisition means for acquiring a plurality of pieces of opinion information that indicate the subjective opinions of a subject regarding an object, a grouping means for dividing the plurality of pieces of opinion information acquired by the acquisition means into a plurality of groups based on predetermined criteria, and a management means for managing the plurality of pieces of opinion information acquired by the acquisition means based on the plurality of groups divided by the grouping means, wherein each of the plurality of pieces of object-related information is information that combines first word information corresponding to a subject and second word information corresponding to a predicate, and further comprises an identification means for identifying emotional information that indicates the subject's emotions regarding each of the plurality of pieces of object-related information acquired by the acquisition means, and the grouping means divides the plurality of pieces of object-related information acquired by the acquisition means into a plurality of groups based on the predetermined criteria and the emotional information identified by the identification means.

Description

[Technical Field]

[0001] The present invention relates to a management system and a management program. [Background technology]

[0002] Conventionally, there have been known techniques for performing various processes on content posted by users about a target (see, for example, Patent Document 1). [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Publication No. 2024-069820 Summary of the Invention [Problem to be solved by the invention]

[0004] However, there is a need for a technique for providing useful information about a subject.

[0005] The present invention has been made in view of the above circumstances, and has as its object to provide a management system and a management program that are capable of providing useful information regarding an object. [Means for solving the problem]

[0006] In order to solve the above-mentioned problems and achieve the object, the management system according to claim 1 is a management system comprising: an acquisition means for acquiring a plurality of pieces of object-related information indicating a subject's subjective opinion regarding an object or facts regarding the object; a grouping means for dividing the plurality of pieces of object-related information acquired by the acquisition means into a plurality of groups based on a predetermined criterion; a generation means for generating name information indicating a name of each of the plurality of groups divided by the grouping means based on the object-related information belonging to each of the plurality of groups divided by the grouping means; and a management system for generating the plurality of object-related information acquired by the acquisition means based on the plurality of groups divided by the grouping means and the name information generated by the generation means. and a management means for managing the object-related information, wherein the management system further comprises a selection means for selecting one or more pieces of candidate category information from a plurality of pieces of candidate category information based on the plurality of groups divided by the grouping means and the name information generated by the generation means, wherein the plurality of pieces of candidate category information are information indicating a plurality of candidate categories that are candidates to be set as categories indicating each of the plurality of groups divided by the grouping means, and the management means manages the plurality of pieces of object-related information acquired by the acquisition means based on the plurality of groups divided by the grouping means, the name information generated by the generation means, and the one or more pieces of candidate category information selected by the selection means. 。

[0007] Claim 2 The management system according to claim 1 In the management system described in , each of the plurality of pieces of object-related information acquired by the acquisition means is information that combines first word information corresponding to a subject and second word information corresponding to a predicate.

[0008] Claim 3 The management system according to claim 1In the management system described above, the plurality of pieces of object-related information acquired by the acquisition means indicate at least the subjective opinion of the subject regarding the object, and the management system further includes an identification means for identifying emotional information indicating the subject's emotions regarding each of the plurality of pieces of object-related information acquired by the acquisition means, and the grouping means divides the plurality of pieces of object-related information acquired by the acquisition means into a plurality of groups based on the predetermined criterion and the emotional information identified by the identification means.

[0010] Claim 4 The management program described in is a management program that functions as an acquisition means for acquiring a plurality of pieces of object-related information that indicate a subject's subjectivity regarding an object or facts regarding the object, a grouping means for dividing the plurality of pieces of object-related information acquired by the acquisition means into a plurality of groups based on a predetermined criterion, a generation means for generating name information indicating a name of each of the plurality of groups divided by the grouping means based on the object-related information belonging to each of the plurality of groups divided by the grouping means, and a management means for managing the plurality of pieces of object-related information acquired by the acquisition means based on the plurality of groups divided by the grouping means and the name information generated by the generation means. The management program causes the computer to further function as a selection means that selects one or more pieces of candidate category information from a plurality of pieces of candidate category information based on the plurality of groups divided by the grouping means and the name information generated by the generation means, and the plurality of pieces of candidate category information are information indicating a plurality of candidate categories that are candidates to be set as categories indicating each of the plurality of groups divided by the grouping means, and the management means manages the plurality of pieces of target-related information acquired by the acquisition means based on the plurality of groups divided by the grouping means, the name information generated by the generation means, and the one or more pieces of candidate category information selected by the selection means. [Effects of the Invention]

[0012] Claim 1 The management system according to claim 4According to the management program described in the above, multiple pieces of target-related information are divided into multiple groups based on predetermined criteria, and the multiple pieces of target-related information are managed based on the multiple groups. This makes it possible to manage the target-related information on a group basis, for example, and therefore to provide useful information about the target. The management system according to claim 1 and claim 4 According to the management program described in the above, by managing multiple object-related information based on multiple groups and the name information of the groups, it is possible to manage object-related information on a group basis, for example, taking name information into consideration, thereby making it possible to provide useful information about the object. The management system according to claim 1 and claim 4 According to the management program described in the above, by managing multiple pieces of target-related information based on multiple groups, the name information of the groups, and selected candidate category information, it is possible to manage target-related information on a group basis, for example, taking into account the name information and candidate category information, thereby making it possible to provide useful information about the target.

[0013] Claim 2 According to the management system described above, each of the multiple object-related information is a combination of first word information corresponding to a subject and second word information corresponding to a predicate, so that, for example, appropriate information can be used as object-related information, thereby making it possible to provide useful information regarding the object.

[0014] Claim 3 According to the management system described in the above, by dividing multiple pieces of object-related information into multiple groups based on predetermined criteria and emotional information, it is possible to appropriately group the information taking into account emotions, for example, and therefore to provide useful information about the object. [Brief explanation of the drawings]

[0017] [Figure 1] 1 is a block diagram showing a functional concept of an information system according to an embodiment of the present invention; [Figure 2] FIG. 10 is a diagram illustrating an example of post-related information. [Figure 3] FIG. 10 is a diagram illustrating an example of opinion-related information. [Figure 4] 1 is a flowchart of an information management process; [Figure 5] FIG. 10 is an explanatory diagram of information management processing. [Figure 6] 10 is a flowchart of a grouping-related process. [Figure 7] FIG. 10 is an explanatory diagram of opinion information and feature amounts. [Figure 8] FIG. [Figure 9] 10 is a description of selection of setting category information. DETAILED DESCRIPTION OF THE INVENTION

[0018] The following describes in detail an embodiment of a management system and a management program according to the present invention with reference to the accompanying drawings. First, [I] the basic concept of the embodiment is explained, then [II] specific content of the embodiment is explained, and finally, [III] modifications to the embodiment are explained. However, the present invention is not limited to the embodiment.

[0019] [I] Basic Concept of the Embodiment First, the basic concept of the embodiment will be described. The embodiment relates to a management system and a management program. The management system according to the present invention is a system for managing object-related information, and the concept includes, for example, a dedicated system for managing object-related information, or a system realized by implementing a function for managing object-related information in a general-purpose system (for example, a server computer (including a cloud computer), a personal computer, a tablet terminal, etc.).

[0020] "Subject-related information" is information related to a subject, and is a concept that includes, for example, information that indicates the subjective opinion of the subject regarding the subject, or information that indicates facts regarding the subject.

[0021] "Object-related information" is, for example, information that combines information corresponding to a subject (first word information) and information corresponding to a predicate (a phrase corresponding to the aforementioned subject) (second word information). Note that the number of subjects and predicates included in one piece of object-related information is arbitrary. As a variation, "object-related information" is not limited to information that combines this first word information and second word information, but may also be composed of information indicating any word.

[0022] "Object-related information" is a concept that includes, for example, text information such as characters or image information.

[0023] An "object" is a thing or concept that is the object or objective of the subject's perception or fact, and is a concept that includes, for example, a product (examples include a specific application program, a hat, shoes, etc.), a service, or a thing (examples include things other than products, people, or animals, such as monuments at tourist spots, company employees, pets, etc.).

[0024] The term "subject" refers to a person who has their own thoughts regarding the subject, and is a concept that indicates, for example, a person who uses the subject (user, employer, boss, customer, owner, etc.).

[0025] "Information that indicates the subject's subjective opinion regarding the subject" is information that indicates the subject's thoughts (opinions) regarding the subject, and is a concept that includes, for example, information that indicates the subject's perception, impressions, or evaluation of the subject.

[0026] "Information indicating facts about an object" is information indicating matters about an object, and is a concept that includes, for example, information indicating the quality, progress, or status of an object.

[0027] In the following embodiment, an example will be described in which the "target" is a predetermined application program, and the "target-related information" is text information that indicates the subjective opinion of the target user regarding the target.

[0028] [II] Specific details of the embodiment Next, specific details of the embodiment will be described.

[0029] (composition) First, the configuration of the information system according to this embodiment will be described. Fig. 1 is a block diagram showing the functional concept of the information system according to this embodiment.

[0030] The information system 100 is a system including a management system, and includes, for example, a terminal device 1 and a server device 2.

[0031] (Configuration - Terminal Device) The terminal device 1 in FIG. 1 is a device used by various users, such as a tablet terminal or a smartphone, and includes, for example, a communication unit 11, a touchpad 12, a display 13, a recording unit 14, and a control unit 15.

[0032] It should be noted that other terminal devices such as a personal computer may also be used as the terminal device 1. The number of terminal devices 1 is arbitrary, but in this embodiment, the one illustrated in FIG.

[0033] (Configuration - Terminal Device - Communication Unit) The communication unit 11 is a communication means for communicating with an external device (for example, the server device 2). The specific type and configuration of the communication unit 11 are arbitrary, but it can be configured using, for example, a known communication circuit or the like.

[0034] (Configuration - Terminal Device - Touchpad) The touchpad 12 is an operation means that receives various operation inputs from the user when pressed by the user's finger, etc. The specific configuration of the touchpad 12 is arbitrary, but for example, a known touchpad equipped with an operation position detection means using a resistive film method, a capacitance method, or the like can be used.

[0035] (Configuration - Terminal Equipment - Display) The display 13 is a display means for displaying various images under the control of the control unit 15. The specific configuration of the display 13 is arbitrary, and for example, a known flat panel display such as a liquid crystal display or an organic EL display can be used. The touch pad 12 and the display 13 may be superimposed on each other to be integrally formed as a touch panel.

[0036] (Configuration - Terminal Device - Recording Unit) The recording unit 14 is a recording means for recording programs and various data required for the operation of the terminal device 1, and can be configured using, for example, a flash memory or the like (the same applies to the recording units of other devices).

[0037] (Configuration - Terminal Device - Control Unit) The control unit 15 is a control means for controlling the terminal device 1, and is specifically a computer including a CPU, various programs interpreted and executed on the CPU (including basic control programs such as an OS and application programs that are started on the OS and realize specific functions), and an internal memory such as RAM for storing programs and various data (the same applies to the control units of other devices). In particular, the program according to the embodiment is installed on the terminal device 1 via an arbitrary recording medium or a network, thereby substantially configuring each unit of the control unit 15 (the same applies to the control units of other devices). The processing of each unit of the control unit 15 will be described later.

[0038] (Configuration - Server Device) The server device 2 in FIG. 1 is a management system, and includes, for example, a communication unit 21, a recording unit 22, and a control unit 23.

[0039] (Configuration - Server Device - Communication Unit) The communication unit 21 is a communication means for communicating with an external device (for example, the terminal device 1 or another device not shown). The specific type and configuration of the communication unit 21 are arbitrary, but it can be configured in the same way as the communication unit 11, for example.

[0040] (Configuration - Server Device - Recording Unit) The recording unit 22 is a recording means for recording programs and various data required for the operation of the server device 2, and stores, for example, post-related information and opinion-related information.

[0041] (Configuration - Server Device - Recording Unit - Posting Related Information) Figure 2 is a diagram showing an example of post-related information. "Post-related information" is information posted by a target person, and as shown in Figure 2, for example, is information in which product name information, date information, and post content information are mutually associated. Note that the information specifically shown in Figure 2 is given for the sake of convenience (the same applies to Figure 3 described below).

[0042] ===Product name information=== The product name information in FIG. 2 is information indicating the name of the product (the predetermined application program as the "target" mentioned above) (such as "AAA" in FIG. 2).

[0043] ===Date Information=== The date information in FIG. 2 is information indicating the date posted by the subject (e.g., "March 1, 2024" in FIG. 2).

[0044] ===Post content information=== The posting content information in Figure 2 is information indicating the content of the information posted by the subject, and is a concept that includes, for example, one or more subject-related pieces of information (in Figure 2, "The app crashes easily, you have to pay..., I think the image quality is good, and..." etc.).

[0045] ===Explanation of each information=== The information at the top of Figure 2 shows that on March 1, 2024, the subject posted the following regarding a product called "AAA": "The app crashes easily, but you have to pay... I think the image quality is good. And..."

[0046] Although there are actually many other pieces of posted content information related to "AAA" in addition to those shown in FIG. 2, the following description will mainly focus on those shown in the figure.

[0047] Furthermore, the method for storing this posting content information is arbitrary, but for example, the information may be stored by an administrator entering the information into the server device 2, or the information may be recorded in one or more external DBs (databases), and the server device 2 may periodically access the external DBs to obtain and store the information from the external DBs.

[0048] (Configuration - Server Device - Recording Unit - Opinion-Related Information) Fig. 3 is a diagram showing an example of opinion-related information. "Opinion-related information" is information related to the subjective opinion of a subject, and as shown in Fig. 3, it is information in which product name information, date information, opinion information, post content information, emotion information, emotion likelihood information, group name information, group likelihood information, representative opinion flag information, and set category information are mutually associated.

[0049] ===Product name information, date information, and post content information=== The product name information, date information, and posting content information in FIG. 3 are the same as the information of the same names in FIG.

[0050] ===Opinion information=== The opinion information in Fig. 3 is the above-mentioned object-related information (information indicating the subject's subjective opinion regarding the object), and is, for example, information indicating the subject's thoughts (also referred to as "opinion") regarding the object (e.g., "The image quality is beautiful" in Fig. 3). The opinion information in Fig. 3 is information indicating the opinion indicated by the posted content information.

[0051] The opinion information in this embodiment is information indicating one opinion. "One opinion" may be interpreted as indicating, for example, one sentence including a combination of one or more subjects and one or more predicates corresponding to the subjects, or as indicating a unit of opinion determined according to any other rule.

[0052] As a variation, information indicating two or more opinions may be used as the opinion information in FIG.

[0053] ===Emotional information=== The emotion information in FIG. 3 is information indicating the emotion of the subject related to the opinion indicated by the opinion information.

[0054] The specific content of the emotional information is arbitrary, but in this embodiment, for example, we will explain the use of "positive" which indicates a positive, affirmative, or proactive emotion, "negative" which indicates a negative, negative, or passive emotion (i.e., an emotion opposite to "positive"), and "neutral" which indicates the absence of emotion or an emotion other than "positive" or "negative."

[0055] ===Emotion Likelihood Information=== The emotion likelihood information in Figure 3 is information that indicates the likelihood (reliability, which is the degree of reliability) of emotion information associated with opinion information, and more specifically, information that indicates the reliability of the emotion indicated by emotion information related to the opinion indicated by the opinion information.

[0056] The specific content of the emotion likelihood information is arbitrary, but in this embodiment, we will explain an example in which the emotion likelihood information is expressed as a numerical value ranging from 0 to 1, with the value increasing as the reliability increases (the same applies to group likelihood information, which will be described later).

[0057] ===Group Name Information=== The group name information in Figure 3 is information (name information) that indicates the name of the group to which the opinion information belongs when the opinion information is grouped (in Figure 3, the name of the group to which the opinion information in the first row of the figure belongs is ``The image quality is good,'' and the name of the group to which the opinion information in the second row of the figure belongs is ``Strong pressure to charge,'' etc.).

[0058] ===Group Likelihood Information=== The group likelihood information in Figure 3 is information indicating the likelihood (reliability, which is the degree of trustworthiness) of the group to which the opinion information belongs, and more specifically, information indicating the reliability that the opinion information belongs to the group corresponding to the group name information (i.e., the degree of suitability of the group to which the opinion information belongs).

[0059] ===Representative opinion flag information=== The representative opinion flag information in FIG. 3 is information indicating whether the opinion indicated by the opinion information is a representative opinion (in FIG. 3, it is "TRUE" indicating that it is a representative opinion and "FALSE" indicating that it is not a representative opinion).

[0060] A "representative opinion" is a concept that indicates a predetermined number of opinions (e.g., a predetermined number such as one, two, or three) that represent the entire group and represent one or more opinions indicated by one or more opinion information belonging to the group.

[0061] ===Settings Category Information=== The set category information in Fig. 3 is information indicating the category of the group indicated by the group name information. Note that, since opinion information belongs to a group, the set category information may be interpreted as information indicating the category of the opinions belonging to the group. More specifically, this set category information is information indicating a category corresponding to one or more candidate categories selected from a plurality of predetermined candidate categories (categories that are set as candidates).

[0062] ===Explanation of each information=== The information at the top of Figure 3 shows the opinion information indicating one opinion corresponding to the subject's post on March 1, 2024, regarding a product named "AAA," which reads, "The app crashes easily, but I have to pay... I think the image quality is good. And..." The information is "The image quality is good." It also shows that the emotion corresponding to the opinion information "The image quality is good" is "positive," and the emotion likelihood information is "0.9" (relatively high reliability). It also shows that the name of the group to which the opinion information belongs is "The image quality is good and preferable," and the group likelihood information is "0.9" (the group of "The image quality is good and preferable" is relatively suitable as a group to which "The image quality is good" belongs). It also shows that the opinion indicated by the opinion information is a representative opinion. It also shows that the category indicated by the group of "The image quality is good and preferable" (i.e., the category of "The image quality is good" that belongs to the group of "The image quality is good and preferable") is "UI" (user interface).

[0063] The opinion-related information may be stored by any method, for example, by executing an information management process described later.

[0064] (Configuration - Server Device - Control Unit) The control unit 23 is a control means that controls the server device 2. The control unit 23 conceptually includes, for example, an acquisition means, a grouping means, a management means, a specification means, a generation means, and a selection means.

[0065] ===Acquisition method=== The acquisition means is a means for acquiring a plurality of pieces of object-related information that indicate the subject's subjective opinion regarding the object or facts regarding the object.

[0066] For example, each of the plurality of object-related information acquired by the acquisition means is information that combines first word information corresponding to a subject and second word information corresponding to a predicate. Also, for example, the plurality of object-related information acquired by the acquisition means indicates at least the subject's subjectivity regarding the object.

[0067] ===Grouping Methods=== The grouping means is a means for dividing the plurality of pieces of object-related information acquired by the acquisition means into a plurality of groups based on a predetermined criterion. The grouping means divides the plurality of pieces of object-related information acquired by the acquisition means into a plurality of groups based on, for example, the predetermined criterion and the emotion information identified by the identification means.

[0068] ===Management means=== The management means is a means for managing the plurality of pieces of object-related information acquired by the acquisition means based on the plurality of groups divided by the grouping means. The management means manages the plurality of pieces of object-related information acquired by the acquisition means, for example, based on the plurality of groups divided by the grouping means and name information generated by the generation means. The management means manages the plurality of pieces of object-related information acquired by the acquisition means, for example, based on the plurality of groups divided by the grouping means, name information generated by the generation means, and one or more pieces of candidate category information selected by the selection means.

[0069] ===Specific means=== The specifying means is means for specifying emotion information indicating the emotion of the subject for each of the plurality of subject-related information acquired by the acquiring means.

[0070] ===Generation means=== The generating means is a means for generating name information indicating the name of each of the plurality of groups divided by the grouping means, based on the object-related information belonging to each of the plurality of groups divided by the grouping means.

[0071] ===Selection Method=== The selection means is means for selecting one or more pieces of candidate category information from the plurality of pieces of candidate category information based on the plurality of groups divided by the grouping means and the name information generated by the generation means. Note that the plurality of pieces of candidate category information are information indicating a plurality of candidate categories that are candidates to be set as categories indicating each of the plurality of groups divided by the grouping means.

[0072] The specific processing executed by each means will be described later.

[0073] (process) Next, we will explain the information management process executed by the information system 100. Figure 4 is a flowchart of the information management process (in the following explanation of each process, steps are abbreviated as "S"), and Figure 5 is an explanatory diagram of the information management process. Note that Figure 5 shows an overview of the process.

[0074] The "information management process" is a process for managing opinion-related information, and is, for example, a process executed by the server device 2. The timing for executing this information management process is arbitrary, but for example, it is assumed that the process starts when the user performs the following predetermined operation, and the description will begin from the point where the process starts.

[0075] In addition, a case will be described in which a large amount of information other than the information specifically shown in the top row is already stored in the post-related information in Fig. 2. Also, some elements in Fig. 5 will be described with reference numerals.

[0076] ===SA1=== At SA1 in Fig. 4, the control unit 23 of the server device 2 acquires opinion information. Specifically, although this is optional, for example, the control unit 23 acquires post content information in the post related information in Fig. 2, and then acquires opinion information based on the acquired post content information.

[0077] In detail, for example, a method using an opinion-related trained model may be used.

[0078] The "opinion-related trained model" is a model that, when post content information is input, outputs opinion information corresponding to the post content information. This opinion-related trained model is a model that has been trained in advance to output, as opinion information, information that combines first word information corresponding to a subject and second word information corresponding to a predicate, as information indicating the subject's opinion about the target, with respect to the input post content information, and is recorded, for example, in the recording unit 22.

[0079] The opinion information to be output here is, for example, basically information corresponding to a sentence (a sentence including a subject and a predicate) included in the text indicated by the input post content information, but specifically, the sentence itself included in the input post content information or information indicating a sentence corresponding to that sentence is expected. That is, in addition to the sentence itself included in the post content information, it is also expected that a sentence with a similar meaning corresponding to that sentence will also be included in the opinion information. Furthermore, the opinion-related trained model is expected to output only one piece of opinion information or two or more pieces of opinion information depending on the content of the input post content information.

[0080] ==Processing== In this method, the control unit 23 acquires post content information from the post-related information in the figure and inputs the post content information to the opinion-related trained model. In this case, the opinion-related trained model outputs one or more pieces of opinion information corresponding to the input post content information, and the control unit 23 identifies and acquires the output opinion information.

[0081] Then, by executing the above process for all the post content information included in the post related information in FIG. 2, a plurality of pieces of opinion information are acquired.

[0082] Here, for example, the post content information in Figure 2, "The app crashes easily, but if you don't pay... I think the image quality is good, and..." is acquired, and the acquired post content information is input into the opinion-related trained model. Then, when the opinion-related trained model outputs opinion information such as "The image quality is good" (see Figure 3) and "The app crashes easily," these "The image quality is good," "The app crashes easily," etc. are identified and acquired as opinion information.

[0083] In the explanatory diagram of FIG. 5, the processing here results in obtaining opinion information 81 to 83 and the like based on posted content information 91 and the like.

[0084] == Variations == As a variation, the processing may be configured to use a generative model stored outside the server device 2 (such as a GPT (Generative Pretrained Transformer) model that outputs answer information corresponding to query information when query information is input), or the processing may be configured to use a predetermined algorithm without using a model (the same applies to SA2 described below).

[0085] As a specified algorithm, for example, an algorithm may be applied that breaks down the post content information into sentence units, determines whether each sentence unit is a sentence that indicates the subject's opinion about the target based on whether it contains a specified word, and obtains the sentence that indicates the subject's opinion about the target as opinion information.

[0086] ===SA2=== At SA2 in FIG. 4, the control unit 23 of the server device 2 identifies emotion information and emotion likelihood information for each piece of opinion information acquired at SA1.

[0087] In detail, for example, a method using an emotion-related trained model may be used.

[0088] The "emotion-related trained model" is a model that, when one piece of opinion information is input, outputs emotion information and emotion likelihood information corresponding to the opinion information. This emotion-related trained model is a model that has been trained in advance to output emotion information ("positive," "negative," or "neutral") indicating the emotion of a subject corresponding to the opinion indicated by the input opinion information and emotion likelihood information related to the emotion information (a numerical value ranging from 0 to 1, the higher the value becomes as the reliability increases), and is recorded in, for example, the recording unit 22.

[0089] ==Processing== In this method, the control unit 23 selects one piece of opinion information from the opinion information acquired in SA1 and inputs the selected piece of opinion information to the emotion-related trained model. In this case, the emotion-related trained model outputs emotion information indicating the emotion corresponding to the input opinion information and emotion likelihood information related to the emotion information, and the control unit 23 acquires and identifies the output emotion information and emotion likelihood information.

[0090] Then, by executing the above process for all the opinion information acquired in SA1, emotion information and emotion likelihood information corresponding to all the opinion information are identified.

[0091] Here, for example, if SA1 acquires opinion information such as "The image quality is beautiful" (see Figure 3) or "The app crashes easily," first, "The image quality is beautiful" is selected and input into the emotion-related trained model. Then, if "positive" and "0.9" are output from the emotion-related trained model as emotion information and emotion likelihood information, these "positive" and "0.9" are acquired and identified as emotion information and emotion likelihood information corresponding to "The image quality is beautiful." Then, the same process is performed for "The app crashes easily," etc.

[0092] In the explanatory diagram of FIG. 5, the processing here identifies emotion information 71-73 etc. based on opinion information 81-83 etc., for example.

[0093] ===SA3=== In SA3 of Fig. 4, the control unit 23 of the server device 2 executes grouping-related processing. Fig. 6 is a flowchart of the grouping-related processing.

[0094] The "grouping-related processing" is processing for dividing opinion information into groups (grouping), and in this embodiment, for example, is processing that includes processing for dividing multiple pieces of opinion information acquired in SA1 into multiple groups based on predetermined criteria and emotion information identified in SA2.

[0095] ===SB1=== In SB1 of Fig. 6, the control unit 23 of the server device 2 acquires opinion information in which "positive" has been specified as emotion information in SA2 from the opinion information acquired in SA1 of Fig. 4. Then, in the following SB2 to SB5, processing is performed on this "positive" opinion information.

[0096] Here, for example, among the comments obtained in SA1 such as "The image quality is beautiful" (see Figure 3) and "The app crashes easily," "The image quality is beautiful" is a "positive" comment, so this "image quality is beautiful" is obtained.

[0097] ===SB2=== In SB2 of FIG. 6, the control unit 23 of the server device 2 identifies the feature amount of the "positive" opinion information acquired in SB1.

[0098] A "feature" is a concept related to AI (Artificial Intelligence) or machine learning. Specifically, it is a quantitative representation of data characteristics, which can be identified, for example, using a specified embedding model.

[0099] Specifically, although the details are arbitrary, for example, the control unit 23 applies the opinion information acquired by SB1 to a predetermined embedding model, thereby acquiring and identifying, as a feature, numerical vector information of a predetermined number of dimensions that indicates the opinion information converted by the model.

[0100] Fig. 7 is an explanatory diagram of opinion information and feature quantities. Here, for example, a case where opinion information such as "The image quality is beautiful," "The addition of new functions has improved work efficiency," and "The new design is very easy to use" is acquired in SB1 as shown in Fig. 7 will be described.

[0101] In this case, when each piece of opinion information illustrated in Figure 7 is converted into numerical vector information such as [-0.28, 0.31, -0.35, -0.40, 0.15, -0.05] by applying a predetermined embedding model to each piece of opinion information, these [-0.28, 0.31, -0.35, -0.40, 0.15, -0.05] etc. are acquired and identified as features of each piece of opinion information.

[0102] ===SB3=== In SB3 of FIG. 6, the control unit 23 of the server device 2 groups the "positive" opinion information acquired in SB1 based on the feature amount identified in SB2.

[0103] Specifically, although the specific method is arbitrary, for example, the control unit 23 performs grouping to group opinion information whose feature values ​​are close to each other (i.e., whose meanings are similar) into the same group. Note that the criterion for this grouping may be interpreted as corresponding to the "predetermined criterion."

[0104] In detail, points corresponding to the numerical vector information, which is the feature amount identified in SB2, are arranged in space, and a grouping process is performed so that opinion information of each point corresponding to a point group that is arranged close to each other and has a density higher than a predetermined value is grouped into the same group. By processing in this way, opinion information having similar features (i.e., similar meanings) is grouped into the same group. Note that, as a method for realizing this process, a method using a predetermined clustering model or other methods may be applied.

[0105] Fig. 8 is an explanatory diagram of groups. Here, for example, if feature amounts such as "The image quality is beautiful" and "The new design is very easy to use" in Fig. 7 have values ​​that are relatively close to each other (i.e., they are feature points that are located relatively close when arranged in space), and feature amounts such as "The addition of new functions has improved work efficiency" have values ​​that are relatively far from the aforementioned features (i.e., they are feature points that are located relatively far when arranged in space), then, as shown in Fig. 8, opinion information such as "The image quality is beautiful" and "The new design is very easy to use" will be grouped together in the same group, and opinion information such as "The addition of new functions has improved work efficiency" will be grouped in another group different from the aforementioned group.

[0106] In reality, it is expected that the opinions will be grouped into groups other than the above two groups depending on the number of pieces of opinion information, the dispersion or density of each feature amount, and the like.

[0107] In the explanatory diagram of FIG. 5, by this processing, for example, opinion information 81, 83, etc. are grouped into the same group.

[0108] == Variations == As a variation, the distance between each feature may be calculated, and opinion information whose calculated distance is equal to or less than a predetermined distance may be grouped together, or any other method may be used for grouping.

[0109] ===SB4=== At SB4 in FIG. 6, the control unit 23 of the server device 2 identifies group likelihood information for each piece of opinion information grouped at SB3.

[0110] Specifically, although this is optional, for example, the control unit 23 calculates the distance (i.e., the difference) between the reference feature of each group grouped in SB3 and the feature of each opinion information, and identifies the group likelihood information based on the calculated distance.

[0111] The "reference feature" is the central feature of the features of opinion information belonging to each group, and is a concept that indicates, for example, the feature that corresponds to the center when each feature is arranged in space.Also, as a variation, it can be interpreted as indicating the feature of opinion information that has the shortest total distance from other features.

[0112] In detail, the group likelihood information is determined so that the shorter the calculated distance (i.e., the smaller the difference between the feature amounts), the closer the feature amount is to the central reference feature amount and the higher the reliability. For example, the group likelihood information may be determined using various information (such as a mathematical formula, a conversion table, or a threshold value) that indicates the relationship between the distance and the group likelihood information.

[0113] Here, for example, if the distance between the reference feature of the group to which "The image quality is good" in Figure 8 belongs and the feature of "The image quality is good" is relatively close (the difference amount is relatively small), this "The image quality is good" is considered to be a group to which this "The image quality is good" belongs to a relatively high degree as the group (the group grouped by SB3), and the group likelihood information is specified as "0.9".

[0114] ===SB5=== At SB5 in FIG. 6, the control unit 23 of the server device 2 identifies representative opinion flag information for each piece of opinion information grouped at SB3.

[0115] Specifically, although this is optional, for example, a predetermined number of pieces of opinion information are selected from among the pieces of opinion information belonging to each group grouped in SB3 as opinion information indicating a representative opinion, and the representative opinion flag information for the selected opinion information is identified as "TRUE," and the representative opinion flag information for opinion information other than the selected opinion information (i.e., opinion information indicating opinions other than the representative opinion) is identified as "FALSE."

[0116] Here, for example, with regard to "The image quality is beautiful" and "The new design is very easy to use" in Figure 8, which belong to one group, if "The image quality is beautiful" is selected as the opinion information indicating the representative opinion and "The new design is very easy to use" is not selected as the opinion information indicating the representative opinion, "TRUE" is identified as the representative opinion flag information for "The image quality is beautiful" and "FALSE" is identified as the representative opinion flag information for "The new design is very easy to use".

[0117] In the explanatory diagram of FIG. 5, the process here converts, for example, the opinion information 81 etc. into opinion information indicating the representative opinion 81A.

[0118] The method for selecting opinion information indicating a representative opinion is arbitrary, and for example, the following method may be used, or other methods may be used.

[0119] For example, each piece of opinion information belonging to each group is broken down into words, and the importance of each word (the degree of importance is determined according to the frequency of appearance of the word within the group, with the higher the frequency of appearance the higher the importance) is identified. Then, from among the words of all the opinion information belonging to the group, a predetermined number (e.g., two or three) of words with the highest importance are identified as group-side important words. Next, in each piece of opinion information belonging to the group, a predetermined number (e.g., two or three) of words with the highest importance are identified as opinion information-side important words. Next, similarities between the group-side important words and the opinion information-side important words are identified, and opinion information corresponding to a predetermined number (e.g., two or three) of opinion information-side important words with the highest similarities may be selected as opinion information indicating a representative opinion.

[0120] The similarity between the group-side important words and opinion-information-side important words may be determined, for example, by converting each word into numerical vector information using a predetermined model, and then determining the similarity based on the difference or distance between the numerical vector information.

[0121] == Variations == As a variation, for example, by focusing on the group likelihood information identified in SB4, a predetermined number of pieces of opinion information with the highest values ​​of the group likelihood information in each group may be selected as opinion information indicating a representative opinion. Alternatively, other methods may be applied.

[0122] ===SB6=== At SB6 in Fig. 6, the control unit 23 of the server device 2 acquires opinion information for which "negative" has been specified as emotion information at SA2 from the opinion information acquired at SA1 in Fig. 4. Then, at the following SB7 to SB10, processing is performed on this "negative" opinion information.

[0123] The processes of SB7 to SB10 are similar to those of SB2 to SB5, so only an outline will be explained.

[0124] ===SB7=== In SB7 of FIG. 6, the control unit 23 of the server device 2 identifies the feature amount of the "negative" opinion information acquired in SB6.

[0125] ===SB8=== In SB8 of FIG. 6, the control unit 23 of the server device 2 groups the "negative" opinion information acquired in SB6 based on the feature amount identified in SB7.

[0126] ===SB9=== At SB9 in FIG. 6, the control unit 23 of the server device 2 identifies group likelihood information for each piece of opinion information grouped at SB8.

[0127] ===SB10=== 6, the control unit 23 of the server device 2 identifies the representative opinion flag information for each piece of opinion information grouped in SB8, and then returns from the grouping-related processing in FIG.

[0128] ===SA4=== At SA4 in FIG. 4, the control unit 23 of the server device 2 generates group name information indicating the name of each group grouped at SA3 (specifically, SB3 and SB8 in FIG. 6).

[0129] Specifically, although the details are arbitrary, for example, the control unit 23 processes using a generative model stored outside the server device 2 (such as a GPT (Generative Pretrained Transformer) model that, when query information is input, outputs answer information indicating an answer corresponding to the query information).

[0130] ==Processing== In detail, the control unit 23 acquires, for each group, opinion information belonging to the group, and also acquires emotion information corresponding to each group, and then inputs query information including the acquired opinion information and instruction information including the acquired emotion information (information indicating an instruction to generate and output the name of the group to which the input opinion information belongs, taking into account the emotion indicated by the input emotion information) into the generative model by communicating with the outside.

[0131] Then, the generative model outputs group name information corresponding to the input opinion information and emotion information as response information corresponding to the inquiry information, and the control unit 23 acquires the group name information by communicating with the outside, and sets the acquired group name information as the group name information generated for the aforementioned group.

[0132] In addition to emotion information, the instruction information includes information indicating a predetermined naming rule for group names, i.e., information that has been prepared in advance through adjustments, etc., so that appropriate group name information can be obtained. Furthermore, because emotion information is included in the instruction information, it is possible to obtain appropriate group name information that takes emotion into consideration.

[0133] Furthermore, with regard to the emotional information included in the instruction information, when generating group name information for a group grouped in SB3 of FIG. 6, the "positive" emotional information identified in SA2 is used, and when generating group name information for a group grouped in SB8 of FIG. 6, the "negative" emotional information identified in SA2 is used.

[0134] For example, for the group in the upper part of Figure 8, opinion information belonging to the group is obtained as "The image quality is beautiful" and "The new design is very easy to use," and "Positive" identified in SA2 of Figure 4 is obtained as emotion information corresponding to the group. Then, inquiry information including the obtained "The image quality is beautiful" and "The new design is very easy to use," etc., and instruction information including the obtained "Positive" is input into the generative model.

[0135] Then, for the response information corresponding to the inquiry information, the generative model outputs, for example, "The image quality is beautiful" and "The new design is very easy to use" as group name information corresponding to the input and "positive" comments, such as "The image quality is beautiful and desirable." In this case, the control unit 23 acquires the "image quality is beautiful and desirable" comment and uses the acquired "image quality is beautiful and desirable" as the group name information generated for the aforementioned group. That is, the control unit 23 generates "The image quality is beautiful and desirable" as the group name information for the aforementioned group.

[0136] Similar processing is also performed on the other groups in FIG. 8 to generate group name information for each group.

[0137] In the explanatory diagram of FIG. 5, the processing here generates group name information 61 and the like.

[0138] ===SA5=== Fig. 9 explains the selection of setting category information. At SA5 in Fig. 4, the control unit 23 of the server device 2 selects setting category information.

[0139] Specifically, although this is optional, it is assumed that, for example, a plurality of pieces of candidate category information (not shown in FIG. 1) are recorded in the recording unit 22. The "plurality of candidate category information" is information indicating a plurality of candidate categories that are candidates to be set as categories indicating a plurality of groups, and is information that is pre-recorded by, for example, a user of the information system 100 determining the candidate categories in accordance with his or her own analysis policy regarding the subject and inputting the information into the server device 2.

[0140] In this embodiment, a case will be described in which a plurality of candidate category information items such as "cost," "UI" (user interface), and "performance" shown in FIG. 9 are stored in the recording unit 22.

[0141] Specifically, for example, the control unit 23 performs processing using a generation model stored outside the server device 2 (such as a model that, when query information is input, outputs response information indicating an answer corresponding to the query information), as in the case of SA4 in Figure 3.

[0142] ==Processing== In detail, the control unit 23 acquires the group name information generated in SA4 for each group grouped in SA3 (more specifically, SB3 and SB8 in Figure 6), and also acquires opinion information for which "TRUE" has been specified as the representative opinion flag information in SB5 or SB10 in Figure 6 (i.e., opinion information indicating a representative opinion among the opinion information belonging to the group), and multiple candidate category information from the recording unit 22, and then inputs query information including the acquired multiple candidate category information, the acquired group name information, the acquired opinion information (opinion information indicating a representative opinion), and instruction information (information indicating an instruction to select and output one or more appropriate candidate category information corresponding to the input group name information and opinion information indicating a representative opinion from the input multiple candidate category information) into the generative model by communicating with the outside.

[0143] Then, the generative model selects and outputs one or more appropriate candidate category information corresponding to the input group name information and opinion information indicating the representative opinion from the multiple pieces of candidate category information input as response information corresponding to the inquiry information, and the control unit 23 acquires the one or more pieces of candidate category information by communicating with the outside, and selects the acquired candidate category information as the set category information.

[0144] It should be noted that the aforementioned instruction information includes information that has been prepared in advance through adjustments, etc., so that appropriate setting category information can be obtained. By configuring the information in this manner, it becomes possible to select appropriate candidate category information as setting category information.

[0145] For example, for the group in the upper part of Figure 9, "The image quality is good and pleasing" is obtained as group name information, and "The image quality is good" or the like is obtained as opinion information indicating a representative opinion. Furthermore, "Cost," "UI," and "Performance" or the like are obtained as multiple candidate category information, and then query information including these pieces of information and instruction information is input into the generation model.

[0146] Then, the generative model outputs "UI", which is one of the candidate category information, as response information corresponding to the inquiry information, and in this case, the control unit 23 selects "UI" as the set category information as shown in Figure 9.

[0147] Furthermore, similar processing is performed for the other groups in FIG. 9 to select the setting category information for each group.

[0148] In the explanatory diagram of FIG. 5, the processing here sets the setting category information 51 and the like.

[0149] ===SA6=== In SA6 of FIG. 4, the control unit 23 of the server device 2 performs management processing on the opinion information acquired in SA1 based on the processing results of the above-mentioned processes (SA1 to SA5).

[0150] The "management process" is a process for managing opinion information, and is a concept that includes, for example, a process for storing information related to opinion information, and a process for outputting (display output or audio output) information related to the opinion information.

[0151] In this embodiment, a process for storing opinion-related information in FIG. 3 will be described as a process for storing information related to opinion information, and a process for displaying and outputting information related to opinion information based on the opinion-related information will also be described.

[0152] Storing Opinion-Related Information = Product name information, date information, and post content information = The control unit 23 stores, in the opinion-related information of FIG. 3, information with the same names as those in the post-related information of FIG. 2 as product name information, date information, and post content information.

[0153] =Opinion information, emotion information, emotion likelihood information= The control unit 23 stores the opinion information acquired in SA1 and the emotion information and emotion likelihood information identified in SA2 in FIG. 4 as the opinion information, emotion information, and emotion likelihood information in FIG.

[0154] =Group name information, group likelihood information= The control unit 23 stores the group name information generated at SA4 in FIG. 4 and the group likelihood information identified at SB4 or SB9 in FIG. 6 as the group name information and group likelihood information in FIG.

[0155] In SA4 to SA5 in FIG. 4, only opinion information with emotional information of “positive” and “negative” is processed (see SB1 and SB6 in FIG. 6), and opinion information with emotional information of “neutral” is not processed. Therefore, for opinion information with emotional information of “neutral”, group name information and group likelihood information are not stored (the same applies to representative opinion flag information and set category information), as shown in the third row of FIG. 3.

[0156] =Representative opinion flag information, setting category information= The control unit 23 stores the representative opinion flag information specified at SB5 and SB10 in FIG. 6 and the setting category information selected at SA5 in FIG. 4 as the representative opinion flag information and setting category information in FIG.

[0157] In this way, the opinion-related information in FIG. 3 is stored.

[0158] ==Display information about opinion information== The control unit 23 generates screen information for displaying information related to the opinion information based on the opinion-related information in Figure 3 and transmits it to the terminal device 1, thereby making it possible to display the information related to the opinion information in a predetermined display mode on the display 13 of the terminal device 1.

[0159] =Part 1= Specifically, although this is optional, for example, it may be configured to display information indicating the number of pieces of opinion information belonging to a group indicated by each group name information on the date on which a specific product was posted (date information) based on the product name information, date information, opinion information, and group name information in Fig. 3. In other words, it may be configured to display the change over time in the number of pieces of opinion information belonging to a group.

[0160] For example, it may display information showing the change over time in the number of pieces of opinion information associated with "product name information" = "AAA" and "group name information" = "good image quality" in Fig. 3. In this case, it may be displayed as a line graph with the horizontal axis representing the date and the vertical axis representing the number of pieces of opinion information, or may be displayed in another display format (similar to "Part 2" below).

[0161] In this case, when the user selects and inputs specific information (for example, "positive" or "UI") as the emotion information or set category information in FIG. 3, the number of pieces of opinion information associated with the specific information may be displayed, taking into consideration the emotion information or set category information.

[0162] =Part 2= Furthermore, for example, it may be configured to display information indicating the number of pieces of opinion information associated with each emotion indicated by the emotion information on the date (date information) when a post was made about a specific product, based on the product name information, date information, opinion information, and emotion information in Fig. 3. In other words, it may be configured to display the change over time in the number of pieces of opinion information associated with each emotion.

[0163] For example, information showing the change over time in the number of pieces of opinion information associated with "product name information"="AAA" and "emotion information"="positive" in FIG. 3 may be displayed.

[0164] In this case, the number and percentage of opinion information associated with each product name information and each emotion information may be displayed in a predetermined format (e.g., bar graph, pie chart, radar chart, etc.) without taking date information into consideration.

[0165] In this case, when the user selects and inputs specific information (for example, "clear and desirable image quality" or "UI") as the group name information or setting category information in Figure 3, the number of pieces of opinion information associated with the specific information may be displayed, taking into consideration the group name information or setting category information.

[0166] =Part 3= Furthermore, for example, based on the product name information, opinion information, and setting category information of Figure 3, information indicating the number or percentage of opinion information associated with each product name information and each setting category information may be configured to be displayed in a predetermined format (e.g., bar graph, pie chart, radar chart, etc.).

[0167] In this case, when the user selects and inputs specific information (for example, "the image quality is good and desirable" or "positive") as the group name information or emotion information in Figure 3, the number of pieces of opinion information associated with the specific information may be displayed.

[0168] =Part 4= In the above, it was explained that the number of pieces of opinion information, etc. are displayed, but it may also be configured to display the opinion information itself related to each piece of displayed information (i.e., "The image quality is beautiful," etc.), the number of pieces of posted content information, or the posted content information itself, when a specified operation from the user is received, or automatically.

[0169] In addition, all or part of the opinion-related information in FIG. 3 may be configured to be displayed in a predetermined format (such as a graph or chart based on an arbitrary item, or a table showing all of the information).

[0170] Furthermore, for example, the information may be configured to be able to be filtered or sorted based on group name information, emotion information, set category information, or other information.

[0171] =Part 5= Note that the above explanation regarding the display of each piece of information included in the opinion-related information in Figure 3 is an example, and this opinion-related information may be configured to be displayed in any other manner as long as it is possible to provide useful information regarding the target (a specified application program).

[0172] (Effects of this embodiment) According to this embodiment, by dividing a plurality of pieces of opinion information into a plurality of groups based on a predetermined criterion and managing the plurality of pieces of opinion information based on the plurality of groups, it is possible to manage the opinion information on a group basis, for example, and therefore to provide useful information regarding the target (a predetermined application program).

[0173] Furthermore, since each of the multiple pieces of opinion information is a combination of first word information corresponding to a subject and second word information corresponding to a predicate (a predicate corresponding to the subject), for example, appropriate information can be used as opinion information, making it possible to provide useful information regarding the subject.

[0174] Furthermore, by dividing a plurality of pieces of opinion information into a plurality of groups based on predetermined criteria and emotional information, it is possible to appropriately group the pieces of opinion information taking into consideration, for example, emotions, and therefore to provide useful information about the subject.

[0175] Furthermore, by managing multiple pieces of opinion information based on multiple groups and the group name information (name information) of those groups, it is possible to manage opinion information on a group basis, for example, taking into account the group name information, thereby making it possible to provide useful information regarding the subject.

[0176] Furthermore, by managing multiple pieces of opinion information based on multiple groups, the group name information of the groups, and selected candidate category information, it is possible to manage opinion information on a group basis, for example, taking into account group name information and candidate category information, thereby making it possible to provide useful information regarding the subject.

[0177] [III] Modifications to the embodiment Although the embodiments of the present invention have been described above, the specific configurations and means of the present invention can be modified and improved as desired within the scope of the technical ideas of the inventions set forth in the claims. Such modifications will be described below.

[0178] (About the problem to be solved and the effects of the invention) First, the problems that the invention aims to solve and the effects of the invention are not limited to those described above, and may vary depending on the implementation environment of the invention and the details of the configuration, and may solve only some of the problems described above or achieve only some of the effects described above.

[0179] (Regarding decentralization and integration) Furthermore, the electrical components described above are conceptual functional components and do not necessarily have to be physically configured as shown in the drawings. In other words, the specific form of distribution or integration of each part is not limited to that shown in the drawings, and all or part of them can be functionally or physically distributed or integrated in any unit depending on various loads, usage conditions, etc.

[0180] (shape, numbers, structure, time series) The components illustrated in the embodiments and drawings may be modified and improved as desired within the scope of the technical concept of the present invention with respect to the shapes, values, paragraphs, or structures or chronological relationships of multiple components. For example, the information items included in the post-related information in Figure 2 and the information items included in the opinion-related information in Figure 3 are examples, and are not limited to these items. Some items may be omitted, or other items may be added.

[0181] (Emotional Information (Part 1)) Furthermore, in the above embodiment, the case where the processing from SA3 in FIG. 4 (SB1 to SB10 in FIG. 6) onwards is executed for emotion information that is "positive" or "negative" has been described. However, the present invention is not limited to this, and for example, the processing from SA3 in FIG. 4 onwards may also be configured to be executed for emotion information that is "neutral".

[0182] Also, for example, processing may be performed without taking emotion information into consideration. In this case, for example, SA2 in Fig. 4 may be omitted, and the opinion information acquired in SA1 may be subjected to the processes SB2 to SB5 in Fig. 6, and processing similar to each process may be performed in SA4 to SA6 in Fig. 4 without using emotion information.

[0183] (Emotional Information (Part 2)) Furthermore, in the above embodiment, a case has been described in which three types of information, "positive," "negative," and "neutral," are used as emotion information, but this is not limiting. For example, any two of the above types of information may be used, or a configuration may be made in which four or more types of information including other information are used. Furthermore, any information related to emotion other than "positive," "negative," or "neutral" (for example, information indicating the intensity of the emotion, the event, type, or direction to which the emotion is directed, etc.) may be used as emotion information.

[0184] In addition, hierarchical information may be configured to be used as emotional information. For example, information such as "direction of opinion," "self-expression," "instructions," and "advice" may be set as information belonging to a level below "neutral," and this level of information may also be used as emotional information. Note that "positive" and "negative" may also be configured in a similar manner.

[0185] (Regarding subject-related information) Furthermore, in the above embodiment, the case where the opinion information in FIG. 3 is the target-related information has been described, but this is not limiting. For example, each process may be performed assuming that the posted content information in FIG. 2 is the target-related information. In this case, the posted content information may be acquired at SA1 in FIG. 4, and from SA2 onward, the process of the embodiment may be performed by replacing the opinion information with the posted content information. In this case, the process content may be changed as appropriate so that the purpose of each process can be achieved.

[0186] In the above embodiment, the case where opinion information is acquired in SA1 of Fig. 4 based on the posted content information of Fig. 3 has been described, but this is not limiting. For example, an administrator or a user may input opinion information to the server device 2, and the opinion information may be stored in advance in the recording unit 22. SA1 of Fig. 4 may acquire this opinion information and execute each process from SA2 onwards.

[0187] In addition, in the above embodiment, we have described the case where ``subject-related information'' is text information that indicates the subject's subjective opinion regarding the subject, but this is not limited to this, and the subject-related information may also be information that indicates facts regarding the subject.

[0188] For example, the posting content information in Fig. 2 may include information corresponding to an error message (such as "Function A does not work") output by a computer regarding the product "AAA" as information indicating facts about the subject, and the processing of the above embodiment may be performed including this information. In this case, the information corresponding to this error message is also acquired and processed as opinion information.

[0189] In this case, the names of information such as "posted content information" and "opinion information" may be replaced with appropriate names such as "posted content information" and "opinion information." In this case, as described above, the system may be configured to process "neutral" or process without considering emotional information.

[0190] Furthermore, in the above embodiment, the case where the object-related information is text information has been described, but this is not limiting, and any type of information (for example, image information, etc.) may be configured to be used as the object-related information.

[0191] (About the model) Furthermore, the processing performed using the trained model (the model recorded in the information system 100) described in the above embodiment may be realized using a generative model recorded externally, and the processing performed using the generative model may be realized using a trained model.

[0192] (About setting category information) Furthermore, in the above embodiment, the case where one or more pieces of candidate category information are selected from all of the plurality of pieces of candidate category information has been described, but the present invention is not limited to this.

[0193] For example, a first group including a plurality of pieces of candidate category information and a second group including a plurality of other pieces of candidate category information may be defined in advance, and by performing processing similar to SA5 in FIG. 4, one or more pieces of candidate category information may be selected as set category information from the plurality of pieces of candidate category information in the first group, and one or more pieces of candidate category information may be selected as other set category information from the plurality of pieces of candidate category information in the second group.

[0194] Furthermore, for example, it is also possible to predetermine three or more groups each containing a plurality of candidate category information, not limited to the first group and the second group, and perform processing similar to SA5 in Figure 4 to select one or more candidate category information from the plurality of candidate category information in each group as the set category information.

[0195] (Regarding opinion information) 4, the case where information that combines first word information corresponding to a subject and second word information corresponding to a predicate is acquired as opinion information has been described, but this is not limitative. For example, as long as it is information that indicates the subjective opinion of a subject regarding an object or information that indicates a fact regarding an object, it may be configured to acquire information that has any component, not limited to components of a sentence such as a subject or a predicate, as opinion information.

[0196] (Selecting setting category information) Furthermore, in SA5 of Fig. 4, the case where opinion information for which "TRUE" is specified as the representative opinion flag information in SB5 or SB10 of Fig. 6 (i.e., opinion information indicating a representative opinion among the opinion information belonging to the group) is acquired and processed has been described, but this is not limited to this. For example, a configuration may be adopted in which attention is focused on the group likelihood information specified in SB4 or SB9 of Fig. 6, a predetermined number of pieces of opinion information in descending order of the value of the group likelihood information is acquired, and processing is performed using the opinion information. In other words, a configuration may be adopted in which a predetermined number of pieces of opinion information in descending order of the value of the group likelihood information is used, instead of opinion information indicating a representative opinion.

[0197] (Regarding omission of processing) Furthermore, in the above embodiment, some processes may be omitted, for example, SA4 or SA5 in Fig. 4 may be omitted. Note that if SA4 is omitted, group name information is not generated, but since the opinion information is grouped in SA3, it becomes possible to manage the opinion information by displaying it in group units in SA6.

[0198] (Regarding changes to processing) In the above-described embodiments, any other method may be applied as long as it achieves the purpose of each process.

[0199] (About combinations) Furthermore, the techniques of the embodiments and modifications may be combined in any manner.

[0200] (Addendum) The management system of Appendix 1 comprises an acquisition means for acquiring a plurality of pieces of object-related information that indicate a subject's subjective opinion regarding an object or facts regarding the object, a grouping means for dividing the plurality of pieces of object-related information acquired by the acquisition means into a plurality of groups based on predetermined criteria, and a management means for managing the plurality of pieces of object-related information acquired by the acquisition means based on the plurality of groups divided by the grouping means.

[0201] The management system of Appendix 2 is the management system described in Appendix 1, wherein each of the plurality of object-related information acquired by the acquisition means is information that combines first word information corresponding to a subject and second word information corresponding to a predicate.

[0202] The management system of Appendix 3 is the management system described in Appendix 1, wherein the plurality of pieces of object-related information acquired by the acquisition means indicate at least the subjective opinion of the subject regarding the object, and the management system further includes an identification means for identifying emotional information indicating the subject's emotion regarding each of the plurality of pieces of object-related information acquired by the acquisition means, and the grouping means divides the plurality of pieces of object-related information acquired by the acquisition means into a plurality of groups based on the predetermined criterion and the emotional information identified by the identification means.

[0203] The management system of Appendix 4 is the management system of Appendix 1, further comprising a generation means for generating name information indicating the name of each of the plurality of groups divided by the grouping means based on the object-related information belonging to each of the plurality of groups divided by the grouping means, and the management means manages the plurality of object-related information acquired by the acquisition means based on the plurality of groups divided by the grouping means and the name information generated by the generation means.

[0204] The management system of Appendix 5 is the management system described in Appendix 4, further comprising a selection means for selecting one or more pieces of candidate category information from a plurality of pieces of candidate category information based on the plurality of groups divided by the grouping means and the name information generated by the generation means, wherein the plurality of pieces of candidate category information are information indicating a plurality of candidate categories that are candidates to be set as categories indicating each of the plurality of groups divided by the grouping means, and the management means manages the plurality of pieces of target-related information acquired by the acquisition means based on the plurality of groups divided by the grouping means, the name information generated by the generation means, and the one or more pieces of candidate category information selected by the selection means.

[0205] The management program of Appendix 6 causes a computer to function as an acquisition means for acquiring a plurality of pieces of object-related information that indicate a subject's subjective opinion regarding an object or facts regarding the object, a grouping means for dividing the plurality of pieces of object-related information acquired by the acquisition means into a plurality of groups based on predetermined criteria, and a management means for managing the plurality of pieces of object-related information acquired by the acquisition means based on the plurality of groups divided by the grouping means.

[0206] (Effect of supplementary notes) According to the management system described in Appendix 1 and the management program described in Appendix 6, multiple pieces of target-related information are divided into multiple groups based on predetermined criteria, and the multiple pieces of target-related information are managed based on the multiple groups.This makes it possible to manage the target-related information on a group basis, for example, and therefore to provide useful information about the target.

[0207] According to the management system described in Appendix 2, each of the multiple object-related information is a combination of first word information corresponding to a subject and second word information corresponding to a predicate, so that, for example, appropriate information can be used as object-related information, thereby making it possible to provide useful information regarding the object.

[0208] According to the management system described in Appendix 3, by dividing a plurality of pieces of object-related information into a plurality of groups based on predetermined criteria and emotional information, it is possible to appropriately group the information taking into account, for example, emotions, and thereby provide useful information about the object.

[0209] According to the management system described in Appendix 4, by managing multiple object-related information based on multiple groups and the name information of the groups, it is possible to manage object-related information on a group basis, for example, taking name information into consideration, thereby making it possible to provide useful information about the object.

[0210] According to the management system described in Appendix 5, by managing multiple target-related information based on multiple groups, the name information of the groups, and selected candidate category information, it is possible to manage target-related information on a group basis, taking into account, for example, name information and candidate category information, thereby making it possible to provide useful information about the target. [Explanation of symbols]

[0211] 1. Terminal equipment 2. Server device 11 Communications Department 12 Touchpad 13. Display 14 Recording section 15 Control Unit 21 Communications Department 22 Recording section 23 Control Unit 51 Setting Category Information 61 Group Name Information 71 Emotional information 72 Emotional information 73 Emotional information 81 Opinion information 81A Representative opinion 82 Opinion information 83 Opinion information 91 Post content information 100 Information Systems

Claims

1. 1. A management system comprising: an acquisition means for acquiring a plurality of pieces of subject-related information indicating a subject's subjective opinion regarding the subject or facts regarding the subject; a grouping means for dividing the plurality of pieces of object-related information acquired by the acquisition means into a plurality of groups based on a predetermined criterion; a generating means for generating name information indicating a name of each of the plurality of groups divided by the grouping means, based on the object-related information belonging to each of the plurality of groups divided by the grouping means; a management means for managing the plurality of object-related information acquired by the acquisition means based on the plurality of groups divided by the grouping means and the name information generated by the generation means, the management system further comprises a selection means for selecting one or more pieces of candidate category information from a plurality of pieces of candidate category information based on the plurality of groups divided by the grouping means and the name information generated by the generation means; the plurality of candidate category information is information indicating a plurality of candidate categories that are candidates to be set as categories indicating each of the plurality of groups divided by the grouping means, the management means manages the plurality of pieces of object-related information acquired by the acquisition means based on the plurality of groups divided by the grouping means, the name information generated by the generation means, and one or more pieces of candidate category information selected by the selection means; Management system.

2. Each of the plurality of pieces of object-related information acquired by the acquisition means is information that combines first word information corresponding to a subject and second word information corresponding to a predicate. The management system according to claim 1 .

3. The plurality of pieces of object-related information acquired by the acquisition means indicate at least the subjective opinion of the subject regarding the object, The management system includes: The apparatus further includes a specifying unit that specifies emotion information indicating an emotion of the subject for each of the plurality of pieces of subject-related information acquired by the acquiring unit, the grouping means divides the plurality of pieces of object-related information acquired by the acquisition means into a plurality of groups based on the predetermined criterion and the emotion information identified by the identification means. The management system according to claim 1 .

4. 1. A management program comprising: Computer, an acquisition means for acquiring a plurality of pieces of subject-related information indicating a subject's subjective opinion regarding the subject or facts regarding the subject; a grouping means for dividing the plurality of pieces of object-related information acquired by the acquisition means into a plurality of groups based on a predetermined criterion; a generating means for generating name information indicating a name of each of the plurality of groups divided by the grouping means, based on the object-related information belonging to each of the plurality of groups divided by the grouping means; a management unit that manages the plurality of object-related information acquired by the acquisition unit based on the plurality of groups divided by the grouping unit and the name information generated by the generation unit, The management program causes the computer to: a selecting means for selecting one or more pieces of candidate category information from a plurality of pieces of candidate category information based on the plurality of groups divided by the grouping means and the name information generated by the generating means, the plurality of candidate category information is information indicating a plurality of candidate categories that are candidates to be set as categories indicating each of the plurality of groups divided by the grouping means, the management means manages the plurality of pieces of object-related information acquired by the acquisition means based on the plurality of groups divided by the grouping means, the name information generated by the generation means, and one or more pieces of candidate category information selected by the selection means; Management program.

Images