Program, information processing system, and information processing method

Abstract

The purpose is to present the products under consideration. [Solution] A program to be executed by a computer equipped with a processor, which causes the processor to perform the following steps: an acquisition step to acquire review information about a product; a perspective identification step to identify the perspectives of the review information using a generating AI; a priority calculation step to calculate the priority of each product to be considered based on the review information for each perspective; a selection step to select products to be considered based on the priority of consideration; and a display processing step to display the products to be considered selected in the selection step.

Description

Technical Field

[0001] The present disclosure relates to a program, an information processing system, and an information processing method.

Background Art

[0002] In product development, efforts are being made to collect and analyze customer feedback and utilize it for product improvement and planning. In particular, a large number of product reviews (VoC) posted on e-commerce sites and the like have become important information sources for grasping product quality and customer satisfaction.

[0003] Patent Document Ⅰ discloses a technique for analyzing word-of-mouth data, extracting negative evaluations, classifying and organizing negative evaluations, and providing them to users.

Prior Art Documents

Patent Documents

[0004]

Patent Document Ⅰ

Summary of the Invention

Problems to be Solved by the Invention

[0005] However, when dealing with a large number of products, there has been a problem that it is difficult to determine which product should be preferentially considered for improvement or the like.

[0006] An object of the present disclosure is to present a product to be considered.

Means for Solving the Problems

[0007] To solve the above problems, a program according to one aspect of the present disclosure is a program to be executed by a computer equipped with a processor, which causes the processor to perform an acquisition step of acquiring review information relating to a product; a perspective identification step of identifying perspectives of the review information using a generating AI; a priority calculation step of calculating a priority for consideration of a product based on the review information for each perspective of each product; a selection step of selecting a product to be considered based on the consideration priority; and a display processing step of displaying the product to be considered selected in the selection step. [Effects of the Invention]

[0008] According to this disclosure, the products under consideration can be presented. [Brief explanation of the drawing]

[0009] [Figure 1] This is a block diagram showing an example of the overall system configuration. [Figure 2] This is a block diagram showing an example of a server device hardware configuration. [Figure 3] This is a block diagram showing the functional parts implemented by the control unit. [Figure 4] This is a block diagram showing an example of a terminal device hardware configuration. [Figure 5] This is a block diagram showing the functional section realized by the control unit of the terminal device. [Figure 6] This flowchart shows the process for displaying items under consideration. [Figure 7] This figure shows an example of the display screen for the product under consideration. [Modes for carrying out the invention]

[0010] The embodiments of this disclosure will be described below with reference to the drawings. In all the drawings illustrating the embodiments, common components are denoted by the same reference numerals, and repeated explanations are omitted. The following embodiments are not intended to unduly limit the content of this disclosure as described in the claims. Not all components shown in the embodiments are necessarily essential components of this disclosure. Also, each drawing is a schematic diagram and is not necessarily a strict illustration.

[0011] Furthermore, in the following description, "processor" refers to one or more processors. A processor may be expressed, for example, as processing circuitry. At least one processor is typically a microprocessor such as a CPU (Central Processing Unit), but may be other types of processors such as a GPU (Graphics Processing Unit). At least one processor may be single-core or multi-core. Also, at least one processor may be a general-purpose processor or a purpose-specific processor.

[0012] Furthermore, at least one processor may be a hardware circuit that performs some or all of the processing (for example, a broad-sense processor such as an FPGA (Field-Programmable Gate Array) or an ASIC (Application Specific Integrated Circuit)).

[0013] Furthermore, in the following explanation, the structure of each table is just an example; one table may be divided into two or more tables, or all or part of two or more tables may be a single table.

[0014] The program may be pre-installed on the information processing device described below, or, for example, the information processing device may be on a readable (e.g., non-temporary) recording medium and the program may be installed on the information processing device. Alternatively, the program may be sent from a program distribution server to the information processing device and installed there. Furthermore, in the following description, two or more programs may be implemented as a single program, or one program may be implemented as two or more programs.

[0015] Furthermore, while various types of identification information are used in the following description, the identification information only needs to be information that indicates a predetermined object, and the specific data is not limited to the embodiments. The identification information may be an identification number, or an identifier that includes letters or symbols.

[0016] Figure 1 is a block diagram showing an example of the overall configuration of System 1. System 1 makes suggestions for product improvements based on product reviews. As shown in Figure 1, System 1 includes, for example, a server device 10, terminal devices 20, and a generation AI system 30. The server device 10, terminal devices 20, and generation AI system 30 are connected by communication via, for example, a network 80. Although Figure 1 shows one terminal device 20, System 1 may have multiple terminal devices 20.

[0017] The server device 10 acquires information related to product reviews (review information) from, for example, the operating system of an e-commerce (EC) site or a company's system, identifies products that companies should consider improving based on the review information, and presents these to the companies. The server device 10 is an information processing device, and is implemented by a computer connected to the network 80.

[0018] The terminal device 20 is, for example, an information processing device operated by employees of a company that provides products, a raw material manufacturer of products, etc. The terminal device 20 is realized by, for example, a mobile terminal such as a smartphone or a tablet. The terminal device 20 may also be realized by, for example, a desktop PC (Personal Computer), a laptop PC, etc.

[0019] The generation AI system 30 provides the use of generative AI. In the present disclosure, the case where the generation AI system 30 provides the use of a large language model (LLM), which is one of the generative AIs, will be described as an example. A large language model is a natural language model designed to execute multiple tasks of natural language processing. A large language model is an example of a learned model and is a model learned using a large number (for example, several billion to several trillion) of parameters and high-level computing resources. A large language model is a computer program or algorithm designed to execute tasks of natural language processing. As an example, in natural language processing, by performing processing such as morphological analysis, syntactic analysis, information extraction, and text generation, the computer analyzes the language used by humans (that is, natural language) and enables predetermined processing to be performed. When a prompt (instruction sentence) is input, the large language model generates an output based on the text, image, etc. of the prompt. The prompt can be defined in natural language.

[0020] Examples of large language models include the GPT (registered trademark) series (Generative Pre-Trained Transformer) developed by OPEN AI, StableLM developed by Stability AI, Llama2 developed by Meta, Palm2 (registered trademark), LaMDA2 (registered trademark) developed by Google, etc. Note that not only large language models but other language models may also be used. The language model may be, for example, BERT (Bidirectional Encoder Representations from Transformers) developed by Google.

[0021] A prompt is mainly an instruction text input to a generative AI. The instruction text may include, for example, text, character strings, images, videos, audio, etc. Also, the prompt may include gestures. When the prompt is input to the generative AI, an answer is output by the generative AI.

[0022] FIG. 2 is a block diagram showing an example of the hardware configuration of the server device 10 shown in FIG. 1. As shown in FIG. 2, the server device 10 includes a control unit 110, a storage unit 120, a communication unit 130, and an input / output IF 140. The control unit 110 executes various processes by executing various programs stored in the storage unit 120. The control unit 110 is a processor such as a CPU, for example. A processor is hardware for executing an instruction set described in a program. A processor is composed of an arithmetic unit, registers, peripheral circuits, etc.

[0023] The storage unit 120 includes a main storage device and an auxiliary storage device. The storage unit 120 stores various programs and various information. The storage unit 120 stores, for example, an application program 121.

[0024] The application program 121 is application software for executing and managing the services provided by the server device 10. The application program 121 includes, for example, a programming language such as JavaScript (registered trademark) that is executed on a web browser application stored in the terminal device 20.

[0025] The communication unit 130 performs modulation and demodulation processing for the server device 10 to communicate with an external device (for example, a terminal device 20). The communication unit 130 performs transmission processing on the signal generated by the control unit 110 and transmits it to the external device. The communication unit 130 performs reception processing on the signal received from the external device and outputs it to the control unit 110. The input / output IF 140 functions as an interface to an input device for receiving input operations from the user and an output device for presenting information to the user.

[0026] Figure 3 is a block diagram showing the functional units implemented by the control unit 110. The control unit 110 includes, as functional units, an acquisition unit 111, a viewpoint identification unit 112, a classification unit 113, an examination priority calculation unit 114, an examination target selection unit 115, and a display processing unit 116. Specifically, the control unit 110 implements each functional unit by reading a program (including an application program 121) stored in the storage unit 120 and executing the instructions contained in the program. Each functional unit will be described later with reference to Figure 6.

[0027] Figure 4 is a block diagram showing an example of the hardware configuration of the terminal device 20 shown in Figure 1. As shown in Figure 4, the terminal device 20 comprises a control unit 210, a storage unit 220, a communication unit 230, an input unit 240, and an output unit 250. Each part included in the terminal device 20 is electrically connected, for example, by a bus.

[0028] The control unit 210, storage unit 220, and communication unit 230 are the same as the control unit 110, storage unit 120, and communication unit 130 of the server device 10, respectively. The control unit 210 performs various processes by executing various programs stored in the storage unit 220.

[0029] The storage unit 220 stores, for example, an application program 221. The application program 221 includes, for example, a programming language that runs on a web browser application (not shown) stored in the storage unit 220. The communication unit 230 performs processing such as modulation and demodulation for the terminal device 20 to communicate with an external device (for example, a server device 10).

[0030] The input unit 240 receives information input from the user. The input unit 240 may be implemented by, for example, a touch-sensitive device that receives instructions or other information by touching an operating surface. If the terminal device 20 is a PC or the like, the input unit 240 may be implemented by a reader, keyboard, mouse, or the like. The input unit 240 converts the instructions or other information input by the user into electrical signals and outputs them to the control unit 210. The input unit 240 may also include, for example, a receiving port that receives electrical signals input from an external input device. The input unit 240 may also include a microphone that receives audio input.

[0031] The output unit 250 outputs various information. The output unit 250 includes, for example, a display unit. The display unit displays various information according to the control of the control unit 210. The output unit 250 may also include a speaker for outputting sound. The output unit 250 may also include an output port for outputting electrical signals to an external output device.

[0032] Figure 5 is a block diagram showing the functional units implemented by the control unit 210. The control unit 210 comprises a reception unit 211, a communication processing unit 212, and a presentation unit 213 as functional units. Specifically, the control unit 210 implements each functional unit by reading the application program 221 stored in the storage unit 220 and executing the instructions contained in the application program 221.

[0033] The reception unit 211 receives instructions or information input from the input unit 240. The communication processing unit 212 transmits and receives data with external devices via the network 80. The presentation unit 213 presents various information to the output unit 250.

[0034] Figure 6 is a flowchart showing the process of displaying items for consideration by the server device 10. The process of displaying items for consideration may be performed periodically, for example, once a month, or it may be performed in response to a request from a user.

[0035] In step S100, the acquisition unit 111 acquires review information and concept information of the target product posted during the target period. Here, the target period and target product are specified by the user. The target period is, for example, a period of one year, one month, etc. The target product may be all of the target company's products, or it may be, for example, multiple products (product group) manufactured from the same material. Thus, the target product can be any product group and is not limited to the embodiment. As another example, a product group using materials provided by a predetermined material supplier may be specified.

[0036] Review information includes, for example, the review text written in natural language, the review title, the date and time of posting, the poster's attributes (age, gender, etc.), the product rating (for example, the number of stars indicating a 5-point rating from 1 to 5), the product name, product code, product category (e.g., scissors, cutter, ruler), and specific information about the purchased product (color, size, etc.).

[0037] The acquisition unit 111 further acquires concept information from an external device via the communication unit 130. Concept information includes, for example, the product's development intent, target customer base, and selling points (strengths). Concept information can be obtained from, for example, product plans, marketing materials, etc. The information acquired by the acquisition unit 111 is stored in the storage unit 120.

[0038] The acquisition unit 111 may, for example, use web scraping technology to acquire review information and concept information from a specific e-commerce site. Alternatively, the acquisition unit 111 may use crawling technology to acquire review information from review sites, social media, etc.

[0039] Next, in step S102, the perspective identification unit 112 uses the generation AI of the generation AI system 30 to identify the perspective of each review piece of information. A perspective is information that indicates the content of the review, such as "poor cutting performance," "easy to hold," or "large handle." The perspective identification unit 112 further assigns a sentiment evaluation score (e.g., positive, negative, neutral) to each perspective.

[0040] Specifically, the perspective identification unit 112 sends a prompt to the generating AI system 30 that includes review information (text data) written in natural language, a perspective identification instruction that identifies a perspective common to multiple reviews, and an emotion rating point identification instruction. The perspective identification instruction identifies "poor cutting performance" as a single perspective from "poor cutting performance" and "hard and difficult to use." The emotion rating point identification instruction identifies one of three emotion ratings for the perspective: positive, negative, or neutral.

[0041] The generation AI system 30 uses the generation AI to identify the perspective and sentiment rating of each review piece of information based on the received prompt, and transmits the perspective and sentiment rating of each review piece of information to the server device 10.

[0042] The perspective identification unit 112 further identifies perspectives from the concept information by inputting a prompt to the generating AI that includes concept information and a perspective identification instruction that identifies a perspective from the concept information.

[0043] Next, in step S104, the classification unit 113 classifies the perspectives into multiple semantic categories. The classification unit 113 uses the generation AI of the generation AI system 30 to classify the perspectives of each review information into multiple semantic categories. In this embodiment, the classification unit 113 classifies them into five categories: "size," "material," "functionality," "design / color," and "price."

[0044] "Size" is a category related to the size of the product. "Materials" is a category related to the materials used. "Functionality" is a category related to the functionality of the product, such as sharpness and ease of use. "Design / Color" is a category related to the appearance, shape, and color of the product. "Price" is a category related to the price, discounts, and cost performance of the product. For example, "good sharpness" falls under the meaning category of "Functionality."

[0045] Furthermore, semantic categories are classified into higher-level broader categories. For example, semantic categories are broadly divided into two main categories: "semantic categories related to materials" and "semantic categories related to products." "Semantic categories related to materials" include "materials and ease of use" and "functionality." "Semantic categories related to products" include "size," "design and color," and "price," etc. The broader categories are assumed to be predetermined.

[0046] Specifically, the classification unit 113 identifies the semantic category to which each viewpoint belongs by sending a prompt to the generating AI system 30 that includes a classification instruction to classify each viewpoint into a semantic category. For example, suppose there is a review that says, "These scissors cut well, but the handle is a little small." In this case, the viewpoint "cut well" is identified, a positive sentiment rating is obtained, and it is classified under "materials / ease of use." Furthermore, the viewpoint "too big" is identified, a negative sentiment rating is obtained, and it is classified under "size."

[0047] Next, in step S106, the consideration priority calculation unit 114 calculates the consideration priority for each product based on the number of reviews for each semantic category. The consideration priority is calculated as the sum of the negative / positive score, the time series score, and the spike score.

[0048] Here, the negative / positive score is calculated based on the number of negative and positive reviews for the same viewpoint. In this embodiment, the higher the ratio of negative reviews to positive reviews (negative review rate), the higher the negative / positive score calculated. Note that the negative / positive score can be any value set based on the number of negative and positive reviews.

[0049] Furthermore, a threshold is set for the negative review rate, and if the negative review rate is above the threshold, a larger value may be set for the negative / positive score than if it were below the threshold.

[0050] Thus, when the negative review rate is the first percentage, the negative / positive score is the first value, and when the negative review rate is the second percentage, which is higher than the first percentage, the negative / positive score is the second value, which is higher than the first value.

[0051] Another example is that the negative / positive score may be calculated based on the difference between the number of negative reviews and the number of positive reviews.

[0052] As a result, products with a higher number of negative reviews compared to positive reviews will have a higher consideration priority. The consideration priority calculation unit 114 calculates a positive / negative score for each of several aspects of a single target product, and the sum of these scores is used as the positive / negative score for the target product.

[0053] Furthermore, if the target perspective and its corresponding perspective are identified from the product's concept information, the negative / positive score may be weighted to a higher value. For example, if the product concept is "sharpness," but the negative / positive score for the perspective "poor sharpness" is high, it corresponds to a greater need for consideration compared to a product where "sharpness" is not a concept. In this case, the negative / positive score should be increased.

[0054] Furthermore, when calculating the negative / positive score, the product evaluation score obtained for the product in question may also be taken into consideration. For example, when the product evaluation score is high, there tends to be fewer negative reviews compared to when the product evaluation score is low. Therefore, even if the negative review rate is the same, a higher negative / positive score may be assigned when the product evaluation score is high compared to when the product evaluation score is low.

[0055] Let's explain time-series scores. A time-series score is a value determined by how much the average product rating has fallen from the average product rating at a certain point in the past. The greater the decline, the higher the time-series score. Examples of past points in time include the same season of the previous year or the previous quarter (the previous three months). As a result, products with worsening reviews will be considered more likely to be evaluated than products with improving reviews.

[0056] The time-series score may also be a value determined by how much the negative / positive score has decreased from the negative / positive score at a certain point in the past. Alternatively, the time-series score may include two aspects: the time-series change in the average product rating score and the time-series change in the negative / positive score.

[0057] Let's explain the spike score. The consideration priority calculation unit 114 ranks the negative aspects of the target product in descending order of the number of reviews. It then calculates the ratio of the number of reviews for the top-ranked aspect to the number of reviews for the second-ranked semantic category. If this ratio exceeds a preset threshold (for example, 3.0 times), the consideration priority calculation unit 114 determines that a "spike has been detected" and increases the "spike score" of that product.

[0058] A situation where dissatisfaction is unusually concentrated on a single specific flaw is likely to indicate a problem common to many buyers. For example, if negative reviews for a product are "poor cutting performance" (595 reviews), "small handle" (46 reviews), and "limited color variations" (30 reviews), the ratio of the top two complaints is a high spike of 12.9 times. This suggests that the product is not simply poorly rated overall, but that there is a specific design issue, "poor cutting performance." The priority calculation unit 114 may further increase the spike score in stages according to this ratio. For example, a ratio of 3.0 times or higher could be assigned 1 point, 5.0 times or higher 2 points, 10.0 times or higher 3 points, and so on, with higher ratios resulting in higher spike scores. This allows for the prioritization of detecting complaints held by a large number of buyers rather than a variety of complaints that depend on individual buyers.

[0059] In this embodiment, the spike score is calculated from the ratio between first and second place. However, the spike score may also be calculated based on the number of negative reviews between other ranks, such as between second and third place.

[0060] Furthermore, a larger ratio between first and second place may be assigned a higher spike score. Thus, when the ratio between first and second place is the first ratio, the spike score is the first value, and when this ratio is higher than the first ratio (the second ratio), the spike score is the second value, which is higher than the first value.

[0061] Next, in step S108, the product selection unit 115 selects products to be considered based on the consideration priority. Specifically, the product selection unit 115 selects products whose consideration priority is equal to or higher than a threshold as products to be considered. Alternatively, the product selection unit 115 may select a predetermined number of products as products to be considered in order of decreasing consideration priority.

[0062] Next, in step S110, the display processing unit 116 displays the products under consideration on the display unit of the terminal device 20. Figure 7 shows an example of the display screen for products under consideration. The display screen 400 shows the number of products, common issues and common strengths of the products. Furthermore, products with a consideration priority above a threshold are classified as requiring action, and their product names, the reasons for selecting them as products to be considered, and countermeasures are displayed. Multiple products classified as requiring action are displayed in order of consideration priority. This allows the user to easily understand which product has a higher priority.

[0063] Furthermore, products are classified into four stages according to their priority for consideration: requiring action, caution, monitoring, and safe, in descending order of priority. In addition, the basis for each product and countermeasures are obtained when the control unit 110 inputs a prompt to the generation AI, which includes an instruction to generate these based on the review information of each product.

[0064] As described above, in the system of this embodiment, the priority of consideration is calculated based on review information for each viewpoint, and the products to be considered are selected and displayed based on the priority of consideration. Therefore, the products to be considered can be identified.

[0065] As a first variation, in calculating the priority of consideration, only perspectives from predetermined semantic categories may be referenced. For example, "size" is easily perceived as a buyer's mistake, and its influence on the product is more limited than that of other semantic categories. Therefore, the priority of consideration calculation unit 114 may calculate the priority of consideration using perspectives other than those belonging to the semantic category of "size."

[0066] As a second variation, the consideration priority may be calculated for each major category. For example, the consideration priority may be calculated only from perspectives belonging to the "semantic category group attributable to materials," and the products to be considered may be selected and displayed. This would allow, for example, materials providers to display the products they should consider. On the other hand, the consideration priority may be calculated only from perspectives belonging to the "semantic category group attributable to products," and the products to be considered may be selected and displayed. This would allow products providers to display the products they should consider.

[0067] The types of materials and products are not limited to the embodiments described. Other examples include semiconductor materials and electronic components, metal materials and industrial products, materials and clothing, raw materials and processed foods, etc.

[0068] Furthermore, the major categories are not limited to two; there may be three or more. For example, in order to make proposals to material providers, product providers, and design developers, the priority of consideration may be calculated based only on the perspectives belonging to each of the three major categories: "semantic categories related to materials," "semantic categories related to product design," and "semantic categories related to product planning," and the products to be considered may be selected and displayed accordingly.

[0069] A third variation is that, in calculating the priority of consideration, perspectives to which positive sentiment ratings have been assigned may be taken into account. For example, if a perspective corresponding to a perspective to which a positive sentiment rating has been assigned is not included in the concept information, it is considered that there is room for improvement in the marketing method. Therefore, if a perspective to which a positive sentiment rating has been assigned is not included in the concept information, the priority of consideration may be set to a higher value than when such a perspective is included in the concept information. This makes it possible to obtain a priority of consideration that reflects the perspective to which a positive sentiment rating has been assigned.

[0070] A fourth variation is to display the actions to be taken along with the products to be considered. Specifically, the control unit 210 inputs a prompt to the generating AI that includes instructions to present the actions to be taken based on the perspectives obtained for the products to be considered and the consideration priority, thereby acquiring and displaying the actions to be taken. This allows the user to confirm the actions according to the consideration priority. The perspectives and priorities to be included in the prompt may be limited to the major categories mentioned above. This allows the control unit 210 to acquire and display actions according to the major categories.

[0071] In the above embodiments, the case in which the functional unit is implemented by a processor has been described, but it is not limited to this. The functional unit may be any hardware known to perform processing.

[0072] Furthermore, although the server device 10 is configured as a single unit in the above embodiment, it may also be configured as a collection of multiple devices. The method of distributing the multiple functions required to realize the server device 10 among multiple devices may be determined according to the processing capacity of each device, the specifications required for the server device 10, etc. Some or all of the functions provided by the server device 10 in this embodiment may be provided in the terminal device 20 instead of the server device 10.

[0073] Although several embodiments of this disclosure have been described above, these embodiments can be implemented in a variety of other forms, and various omissions, substitutions, and modifications can be made without departing from the spirit of the invention. For example, configurations and processes in one embodiment may be combined with configurations and processes in another embodiment, or a modification of one embodiment may be applied to another embodiment. These embodiments and their variations are included in the scope and spirit of the invention, as well as in the claims and their equivalents.

[0074] (Note) The details described in each of the above embodiments are noted below.

[0075] (Note 1) A program to be executed on a computer equipped with a processor, The aforementioned processor, Steps for obtaining product review information, A perspective identification step involves using a generating AI to identify the perspectives of the review information, A priority calculation step that calculates the priority level at which a product should be considered based on review information from each product's perspective, Based on the aforementioned priority of consideration, there is a selection step to select the products to be considered, A display processing step that displays the product under consideration selected in the selection step, A program to execute.

[0076] This configuration allows users to easily identify which products they should consider by calculating consideration priority based on review information for each perspective and selecting and displaying products based on that priority.

[0077] (Note 2) The program described in Appendix 1 displays a plurality of the products under consideration in order of their consideration priority in the display processing step.

[0078] With this configuration, multiple products under consideration are displayed in order of priority, allowing users to quickly see which product should be prioritized.

[0079] (Note 3) The aforementioned perspective is classified into multiple semantic category groups, including semantic category groups attributable to materials and semantic category groups attributable to products. The program described in Appendix 1 or 2 calculates the priority for consideration for each of the semantic category groups in the priority calculation step.

[0080] This configuration allows for the calculation of consideration priority for each semantic category group, enabling each department to appropriately identify the products they should consider.

[0081] (Note 4) The program described in any one of Appendix 1 to 3 further displays the actions to be performed on the product under consideration, which are generated by the generating AI based on the perspective and consideration priority of the product under consideration, in the display processing step.

[0082] This configuration presents specific actions to be taken along with the products to be considered, allowing the person in charge to quickly implement improvement measures and countermeasures.

[0083] (Note 5) The aforementioned priority for consideration is calculated based on the number of negative and positive reviews for the same perspective, and is one of the programs described in any one of the appendices 1 to 4.

[0084] This configuration calculates the priority of consideration by taking into account not only the number of negative reviews but also the number of positive reviews, allowing for a more accurate determination of the need for review based on the actual evaluation status of the product.

[0085] (Note 6) The aforementioned priority for consideration is calculated based on the number of negative and positive reviews for the same viewpoint, and whether or not the viewpoint corresponds to the viewpoint expressed in the product concept, as described in Appendix 5.

[0086] This configuration allows for a more strategic review of products, as product concept information is taken into account when calculating the priority for consideration.

[0087] (Note 7) The aforementioned priority for consideration is calculated based on the number of negative and positive reviews for the same perspective, and the product rating obtained for the product, as described in Appendix 5.

[0088] This configuration allows for the calculation of consideration priority by taking into account an overall evaluation metric, namely the product evaluation score, thus enabling the acquisition of appropriate priorities from a comprehensive perspective.

[0089] (Note 8) The aforementioned priority for consideration is calculated based on the program described in Appendix 5, which determines the change from the corresponding value over a predetermined period in the past, based on the number of negative and positive reviews for the same viewpoint.

[0090] According to this configuration, the degree of deterioration in evaluation over time is taken into account when calculating the priority for consideration, so products whose evaluation has rapidly declined can be identified early and considered for consideration.

[0091] (Note 9) The aforementioned priority for consideration is calculated based on the number of negative reviews for each aspect, and is one of the programs listed in any one of Appendix 1 to 8.

[0092] This configuration allows for the display of products to be considered based on the number of negative reviews.

[0093] (Note 10) In the acquisition step described above, concept information of the product is further acquired, The aforementioned priority for consideration is calculated based on the reviews of the positive meaning category of the product and the concept information of the product, according to one of the programs described in any one of Appendix 1 to 9.

[0094] This configuration allows for the appropriate presentation of products, taking concept information into consideration.

[0095] (Note 11) An information processing method performed by one or more processors, The aforementioned one or more processors include an acquisition step of acquiring review information about a product, The one or more processors perform a perspective identification step in which they identify the perspectives of the review information using generating AI, The aforementioned one or more processors perform a priority calculation step, which calculates the priority of each product that should be considered based on review information for each product's perspective, The one or more processors perform a selection step of selecting products to be considered based on the consideration priority, The one or more processors include a display processing step that displays the product under consideration selected in the selection step, Information processing methods, including those mentioned above.

[0096] (Note 12) An information processing system comprising one or more processors, The one or more processors described above We obtain product review information, The generation AI identifies the perspective of the aforementioned review information, Based on review information from each product's perspective, we calculate the priority level for which products should be considered. Based on the aforementioned priority of consideration, select the products to be considered. An information processing system that displays the selected product under consideration. [Explanation of symbols]

[0097] 1 System 10 Server devices 20 Terminal devices 30 Generative AI Systems 80 Networks 110 Control Unit 111 Acquisition Department 112 Perspective Identification Section 113 Classification Department 114. Priority Calculation Unit 115 Selection of Subjects for Consideration 116 Display Processing Unit 120 Storage section 121 Application Programs 130 Communications Department 140 Input / Output Interfaces 210 Control Unit 220 Storage section 221 Application Programs 230 Communications Department 240 Input section 250 Output section

Claims

1. A program to be executed on a computer equipped with a processor, The aforementioned processor, Steps for obtaining product review information, A perspective identification step involves inputting the aforementioned review information and a prompt including a perspective identification instruction to identify a perspective common to multiple review pieces of information into a generating AI, and obtaining the perspective from the generating AI to identify the perspective of the review information. A priority calculation step calculates the priority of a product that should be considered for improvement based on the number of negative and positive reviews for each product, identified from the review information for each perspective. Based on the aforementioned priority of consideration, there is a selection step to select the products to be considered, The system executes a display processing step that displays the product under consideration selected in the selection step, The aforementioned review priority increases as the ratio of negative review information to positive review information increases.

2. The program according to claim 1, wherein the display processing step displays a plurality of products under consideration in order of their consideration priority.

3. The aforementioned perspective is classified into multiple semantic category groups, including semantic category groups attributable to materials and semantic category groups attributable to products. The program according to claim 1, wherein the priority calculation step calculates the priority for consideration for each of the semantic category groups.

4. The program according to claim 1, wherein the display processing step further displays actions to be performed on the product under consideration, which are generated by the generating AI based on the viewpoint and the consideration priority of the product under consideration.

5. The program according to claim 1, wherein the priority of consideration is calculated based on whether the viewpoint from which the ratio was calculated corresponds to a viewpoint shown in the product concept.

6. The program according to claim 1, wherein the priority of consideration is calculated based on the product evaluation score obtained for the product, and when the ratios are equal, the higher the product evaluation score, the higher the priority of consideration.

7. The program according to claim 1, wherein the priority of consideration is calculated based on the change in the ratio over a predetermined period in the past.

8. The program according to claim 1, wherein the priority of consideration is calculated based on the ratio of the number of review information to negatives for each viewpoint.

9. In the acquisition step described above, concept information of the product is further acquired, The program according to claim 1, wherein the priority of consideration is calculated based on reviews of the positive meaning category of the product and concept information of the product.

10. An information processing method performed by one or more processors, The one or more processors mentioned above perform an acquisition step of acquiring review information about a product, The one or more processors input the review information and a prompt including a perspective identification instruction to the generating AI to identify a perspective common to the multiple review pieces of information, and obtain the perspective from the generating AI to identify the perspective of the review information, in a perspective identification step. The one or more processors perform a priority calculation step, which calculates the priority of a product to be considered for improvement based on the number of negative and positive reviews for the same perspective identified from the review information for each perspective of each product. The one or more processors perform a selection step of selecting products to be considered based on the consideration priority, The one or more processors include a display processing step that displays the product under consideration selected in the selection step, The aforementioned priority for consideration increases as the ratio of the number of negative review information to the number of positive review information increases.

11. An information processing system comprising one or more processors, The one or more processors described above We obtain product review information, The AI ​​generates the review information and a prompt that includes an instruction to identify a viewpoint common to multiple review pieces of information, and obtains the viewpoint from the AI ​​to identify the viewpoint of the review information. Based on the number of negative and positive reviews for each product, identified from the review information for each perspective, the priority level for considering improvements for each product is calculated. Based on the aforementioned priority of consideration, select the products to be considered. The selected product under consideration is displayed, An information processing system in which the priority of consideration increases as the ratio of the number of negative review information to the number of positive review information increases.

Images