Content classification system, content classification method, and program

Abstract

To provide a content classification system, a content classification method, and a program that can reduce the cost required for content classification and improve content classification accuracy.SOLUTION: A content classification system includes: a storage unit that stores prompt information indicating a prompt used to classify a content into a corresponding category among multiple categories; an acquisition unit that acquires content information indicating details of the content to be classified; and a content classification unit that determines a category that corresponds to the content indicated by the content information on the basis of the acquired content information and the prompt information stored in the storage unit, and generates tag assignment information indicating information to be assigned as a tag to the content for the determined corresponding category.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a content classification system, a content classification method, and a program. [Background technology]

[0002] BACKGROUND ART Conventionally, there are various services that recommend content tailored to a user from among content provided to the user.

[0003] For example, Patent Document 1 below discloses a technology for realizing a service that recommends books tailored to a user. In this technology, each book is tagged in advance with a corresponding item from among a plurality of items indicating classification, and the recommended book is determined by comparing points indicating the degree of match between the user and each item with the tagged items of each book. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Patent No. 7343004 Summary of the Invention [Problem to be solved by the invention]

[0005] When recommending content using tagging of books (classification of content), as in the technology described in Patent Document 1, improving the accuracy of content classification is expected to improve the accuracy of content recommendations. In the technology described in Patent Document 1, content classification was performed manually by relevant parties. Therefore, in order to improve the accuracy of content classification, it is preferable to ask a person (an expert) with more specialized knowledge about each content and its classification to classify the content. However, having experts manually classify content incurs costs proportional to the number of pieces of content, so the more content a service handles, the higher the cost of content classification.

[0006] In view of the above-mentioned problems, an object of the present invention is to provide a content classification system, a content classification method, and a program that can reduce the cost of content classification and improve the accuracy of content classification. [Means for solving the problem]

[0007] In order to solve the above-mentioned problems, one aspect of the present invention provides a content classification system that includes: a memory unit that stores prompt information indicating a prompt used to classify content into a corresponding category among a plurality of categories; an acquisition unit that acquires content information indicating details of the content to be classified; and a content classification unit that determines, based on the acquired content information and the prompt information stored in the memory unit, a category that corresponds to the content indicated by the content information, and generates tagging information that indicates information to be assigned as a tag to the content for the category that is determined to correspond.

[0008] A content classification method according to one embodiment of the present invention is a content classification method executed by a computer, including: a storage process for storing prompt information in a storage unit that indicates a prompt used to classify content into an appropriate category from among a plurality of categories; an acquisition process for acquiring content information that indicates details of the content to be classified; and a content classification process for determining, based on the acquired content information and the prompt information stored in the storage unit, a category that corresponds to the content indicated by the content information, and generating tagging information that indicates information to be assigned as a tag to the content for the category determined to correspond.

[0009] A program according to one aspect of the present invention is a program for causing a computer to function as: a storage means for storing prompt information in a storage unit indicating a prompt used to classify content into an appropriate category from among a plurality of categories; an acquisition means for acquiring content information indicating details of the content to be classified; and a content classification means for determining, based on the acquired content information and the prompt information stored in the storage unit, a category that corresponds to the content indicated by the content information, and generating tagging information indicating information to be assigned as a tag to the content for the category determined to correspond. [Effects of the Invention]

[0010] According to the present invention, it is possible to reduce the cost of classifying content and improve the accuracy of content classification. [Brief explanation of the drawings]

[0011] [Figure 1] 1 is a diagram illustrating an example of a configuration of a content classification system according to an embodiment of the present invention. [Figure 2] FIG. 2 is a diagram showing an example of content classification categories according to the embodiment. [Figure 3] FIG. 2 is a block diagram showing an example of a functional configuration of a content classification server according to the present embodiment. [Figure 4] FIG. 10 is a diagram illustrating an example of prompt information according to the embodiment. [Figure 5] 10A and 10B are diagrams illustrating an example of product information and tagging information according to the embodiment. [Figure 6] FIG. 10 is a sequence diagram showing an example of a processing flow relating to generation of tagging information in the content classification system according to the present embodiment. [Figure 7] FIG. 10 is a sequence diagram showing an example of a processing flow regarding registration and publication of tagging information in the content classification system according to the present embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0012] Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

[0013] <1. Content classification system configuration> The configuration of a content classification system according to this embodiment will be described with reference to Figures 1 and 2. Figure 1 is a diagram showing an example of the configuration of a content classification system according to this embodiment.

[0014] The content classification system 1 shown in Figure 1 is a system for automatically classifying content to be classified into corresponding categories. A service provided using the content classification system 1 is also referred to as a "content classification service" below. A business that provides a content classification service is also referred to as a "service provider" below. In the content classification service, for example, a user is provided with a content classification function (first function) and a function of disclosing information obtained by content classification (second function). In the first function, upon receiving a request from a person (first user) who owns the content to be classified, the content classification system 1 classifies the first user's content. In the second function, in a service provided by the first user, information obtained by content classification about content handled in the service is disclosed to a person (second user) who uses the service. The information obtained by content classification is information indicating the category to which the content to be classified is determined to belong, and is information that is assigned (tagged) to the content. Therefore, the information obtained by content classification is also referred to as "tagged information" below.

[0015] In the following, this embodiment will be described by taking as an example a service provided by a first user that sells products (contents) to general consumers. Note that products are an example of content, and content may be something other than products. The first user is a customer (client) of the service provider, and is hereinafter also referred to as a "customer." The second user is a general consumer who uses the service provided by the first user, and is hereinafter also referred to as a "consumer." Furthermore, this embodiment will be described using an example in which content classification by the content classification system 1 is performed by determining whether a product is educational for children from multiple perspectives. That is, the content classification system 1 classifies products into at least one or more categories that indicate perspectives from among multiple perspectives from which the product to be classified has been determined to be educational for children.

[0016] Here, the categories of content classification according to this embodiment will be described with reference to Fig. 2. Fig. 2 is a diagram showing an example of categories of content classification according to this embodiment.

[0017] Figure 2 shows an example of how intelligence is classified into multiple categories for children using neuroscience. In the example shown in Figure 2, intelligence is classified into eight categories: "Language / Linguistics," "Logic / Mathematics," "Introspection," "Nature / Natural History," "Interpersonal," "Physical / Motor," "Visual / Spatial," and "Music / Rhythm." This classification is based on the eight intellectual domains of children in neuroscience.

[0018] "Language and Linguistics" refers to the ability to use language as a means of thinking and communicating. Specifically, this includes content related to word play, poetry, haiku, tongue twisters, etc.

[0019] "Logic and Mathematics" refers to the ability to use numbers and arithmetic effectively, as well as the ability to make causal connections and predictions. Specifically, this refers to content related to numbers, order, calculations, etc.

[0020] "Introspection" refers to morality, compassion, and the ability to grasp abstract concepts (things that cannot be seen). Specifically, it is the ability to relate to attachment, likes and dislikes, and what is important.

[0021] "Nature and Natural History" refers to the ability to pursue "why" through exploration and observation, and the ability to collect, organize, and classify information. Specifically, this skill applies to content related to nature, animals, observation, and exploration.

[0022] "Interpersonal" refers to the ability to understand other people's feelings and opinions, the ability to work together, etc. Specifically, this applies to content related to playing together, friends, family, etc.

[0023] "Physical and physical activity" refers to the ability to exercise and work with one's hands. Specifically, this applies to content related to exercise, nutrition, and food.

[0024] "Visual and spatial" refers to the ability to compose a solid from a flat surface, and the ability to distinguish colors and shapes. Specifically, this ability applies to content related to drawing, blocks, pop-up books, etc.

[0025] "Music and Rhythm" refers to the ability to understand pitch and rhythm, and the ability to read emotions from vocal intonation. Specifically, this applies to content related to singing, musical instruments, and performance.

[0026] The types and number of content classification categories are not limited to the example shown in FIG. 2, and any types and numbers may be set depending on the type of content, the target audience of the content, and the like.

[0027] As shown in FIG. 1, the content classification system 1 includes a customer terminal 10, a consumer terminal 20, and a content classification server 30. The network NW may be configured to transmit and receive information using, for example, a LAN (Local Area Network), a WAN (Wide Area Network), a telephone network (such as a mobile phone network or a fixed telephone network), a regional IP (Internet Protocol) network, or the Internet.

[0028] (1) Customer terminal 10 The customer terminal 10 is a terminal used by a customer. The customer terminal 10 is, for example, a mobile terminal such as a smartphone or a tablet terminal, or a PC (Personal Computer). The customer terminal 10 is communicably connected to the content classification server 30 via a network NW.

[0029] A customer operates the customer terminal 10 to classify products to be sold to consumers. Before starting to classify products, the customer must first prepare information necessary for classifying the products. The information necessary for classifying products includes, for example, prompt information and content information.

[0030] The prompt information is information indicating a prompt used to classify content into at least one of a plurality of categories, and includes, for example, summary information, category information, and output specification information. The summary information is information that indicates an overview of content classification. The category information is information that indicates details of multiple categories. For example, the category information is information that indicates, as category details, the type of tag that indicates the category, the definition of each tag type, and specific examples that correspond to each definition. The output specification information is information that indicates the specification of the output of tagged information. In the following, this embodiment will be described taking as an example a case where the prompt information is prepared by the service provider, not by the customer. In this case, the customer prepares the information necessary to create the prompt information and provides it to the service provider. The service provider creates the prompt information based on the information provided by the customer and uploads it to the content classification system 1.

[0031] In this embodiment, content classification is performed using AI (Artificial Intelligence). In this embodiment, since tagging information is output as text, AI capable of generating text is used. The AI can perform inference similar to humans by, for example, generating text using LLMs (Large Language Models). Prompt information is used as information to instruct the AI on content classification. The AI classifies content according to the information indicated by the prompt information. Therefore, the AI performs classification with higher accuracy as the amount of prompt information increases, but the classification direction is somewhat narrowed (i.e., classification with low flexibility). On the other hand, the AI may perform classification with lower accuracy as the amount of prompt information decreases, but the classification direction is not very narrowed (i.e., classification with high flexibility). As a result, a customer can adjust the accuracy or flexibility of classification by AI by adjusting the amount of prompt information.

[0032] The content information is information indicating details of the content to be classified, for example, information indicating the title and description of the classification object. In this embodiment, the content information is product information, that is, information indicating product details. In this case, the title of the classification object is the product name, and the description of the classification object is detailed product information (for example, a product description). The customer prepares the product information as data in, for example, a CSV (Comma Separated Values) file format.

[0033] (2) Consumer terminal 20 The consumer terminal 20 is a terminal used by a consumer. The consumer terminal 20 is, for example, a mobile terminal such as a smartphone or a tablet terminal, or a PC. The consumer terminal 20 is communicably connected to the content classification server 30 via a network NW.

[0034] A consumer uses a service provided by a customer by operating the consumer terminal 20. When the consumer selects a product for sale through the service, the consumer can view the product information and tagging information of the product.

[0035] (3) Content classification server 30 The content classification server 30 is a server that executes a process for automatically classifying products to be classified into corresponding categories, and is an example of a content classification device. The content classification server 30 is configured with one or more servers (e.g., cloud servers). The content classification server 30 is connected to the customer terminal 10 and the consumer terminal 20 via a network NW so as to be able to communicate with each other. The content classification server 30 is equipped with an AI function that can store prompt information prepared by a service provider in advance and generate tagging information in text format.

[0036] An overview of the content classification service will now be described with reference to Fig. 1. It is assumed that the prompt information has already been stored in the content classification server 30. First, a customer operates the customer terminal 10 to upload product information about the product to be classified to the content classification server 30. The content classification server 30 uses AI to classify the products and generate tagging information based on the product information uploaded by the customer and pre-stored prompt information. After generation, the content classification server 30 proposes the generated tagging information to the customer. This allows the customer to easily classify products simply by uploading product information, and because classification is performed by AI, highly accurate classification results can be obtained. The customer checks the tagging information proposed by the content classification server 30 via the customer terminal 10. If the tagging information needs to be changed, the customer operates the customer terminal 10 to request the content classification server 30 to reclassify the product. If the tagging information does not need to be changed, the customer registers the generated tagging information in the content classification server 30. After registering the tagging information, the customer makes the product with the tagging information available to consumers. A consumer operates the consumer terminal 20 to use a service provided by a customer. While using the service, the consumer selects, for example, a product for which the consumer wishes to check details. If tagging information has been assigned to the selected product, the content classification server 30 displays the product information of the selected product and its tagging information on the screen of the consumer terminal 20. This allows the consumer to view the product information of the selected product and its tagging information.

[0037] Various screens are displayed on the customer terminal 10 and the consumer terminal 20 by an application (hereinafter also referred to as "app") for using the content classification system 1. For example, a screen related to product classification is displayed on the customer terminal 10. A customer can classify products by operating the screen displayed on the customer terminal 10 by the app. A screen on the consumer terminal 20 on which product information and its tagging information can be confirmed is displayed. A consumer can view product information and its tagging information by operating the screen displayed on the consumer terminal 20 by the app. The app functions may be provided by installing the app on each device (i.e., native app), or by a web system (i.e., web app). In the case of web apps, each app is managed by a server, and its functions are provided via a web browser.

[0038] <2. Functional configuration of content classification server> The configuration of the content classification system 1 according to this embodiment has been described above. Next, the functional configuration of the content classification server 30 according to this embodiment will be described with reference to Figures 3 to 5. Figure 3 is a diagram showing an example of the functional configuration of the content classification server 30 according to this embodiment. As shown in FIG. 3, the content classification server 30 includes a communication unit 310, a storage unit 320, and a control unit 330.

[0039] (1) Communications unit 310 The communication unit 310 has a function of transmitting and receiving various information. The communication unit 310 is communicably connected to the customer terminal 10 and the consumer terminal 20 via the network NW, and transmits and receives various information to and from each terminal.

[0040] (2) Storage section 320 The storage unit 320 has a function of storing various information. The storage unit 320 is a storage medium provided as hardware in the content classification server 30, such as a hard disk drive (HDD). Drive), SSD (Solid State Drive), flash memory, EEPROM (Electrically Erasable Programmable Read Only Memory) It may be composed of Read Only Memory (RAM), Random Access Read / Write Memory (RAM), Read Only Memory (ROM), or any combination of these storage media. As shown in FIG. 2, the storage unit 320 includes a prompt information storage unit 321, a content information storage unit 322, and a tagging information storage unit 323.

[0041] (2-1) Prompt Information Storage Unit 321 The prompt information storage unit 321 has a function of storing prompt information. The prompt information storage unit 321 stores prompt information that is prepared and uploaded by a service provider, for example.

[0042] A specific example of prompt information according to this embodiment will now be described with reference to Fig. 4. Fig. 4 is a diagram showing an example of prompt information according to this embodiment.

[0043] As shown in Fig. 4, the prompt information is information that includes summary information, category information, and output specification information. Fig. 4 shows specific examples of summary information, category information, and output specification information.

[0044] The summary information not only provides an overview of the content classification, but also provides the AI with an overview of what it will do next (its role). As an example, Figure 4 shows the summary information as follows: "You are an expert in product classification. From the product name and detailed information, you can determine whether the product is educational for children and classify it based on the following eight criteria." Based on this summary information, the AI recognizes that it will now classify the product.

[0045] Category information is information that provides details of multiple categories (classification perspectives) and also provides AI with know-how for classifying products. As an example, Figure 4 shows the tag types as "A. Language and Linguistics," "B. Logic and Mathematics," "C. Introspection," "D. Nature and Natural History," "E. Interpersonal," "F. Body and Movement," "G. Vision and Space," and "H. Music and Rhythm." These tag types are the same as the categories explained with reference to Figure 2. Based on these tag types, the AI recognizes perspectives for classifying products.

[0046] The definition of "A. Language and Linguistics" is "the ability to use language as a means of thinking and communicating," and specific examples include "word games, poetry, haiku, and tongue twisters." If the product information for a product to be classified contains elements related to the definition or specific examples of "A. Language and Linguistics," the AI will classify the product into the "A. Language and Linguistics" category.

[0047] The definition of "B. Logic and Mathematics" is "the ability to use numbers and arithmetic effectively, and the ability to make causal connections and predictions," and specific examples include "numbers, sequences, and calculations." If the product information for the product to be classified contains elements related to the definition or specific examples of "B. Logic and Mathematics," the AI will classify the product into the "B. Logic and Mathematics" category.

[0048] The definition of "C. Introspection" is "the power of morality and compassion, the ability to grasp abstract concepts (things that cannot be seen)," and specific examples include "attachment, likes and dislikes, importance." If the product information of the product to be classified contains elements related to the definition or specific examples of "C. Introspection," the AI will classify the product into the "C. Introspection" category.

[0049] The definition of "D. Nature and Natural History" is "the ability to pursue 'why' through exploration and observation, and the ability to collect, organize, and classify information," and specific examples include "nature, animals, observation, exploration." If the product information for the product to be classified contains elements related to the definition or specific examples of "D. Nature and Natural History," the AI will classify the product into the "D. Nature and Natural History" category.

[0050] The definition of "E. Interpersonal" is "the ability to understand other people's feelings and opinions, and the ability to work together," and specific examples include "playing together, friends, family." If the product information for the product to be classified contains elements related to the definition or specific examples of "E. Interpersonal," the AI will classify the product into the "E. Interpersonal" category.

[0051] The definition of "F. Body and Exercise" is "the ability to exercise or work with hands," and specific examples include "gymnastics, dietary education, and food." If the product information for a product to be classified contains elements related to the definition or specific examples of "F. Body and Exercise," the AI will classify the product into the "F. Body and Exercise" category.

[0052] The definition of "G. Visual and Spatial" is "the ability to create three-dimensional objects from flat surfaces, and the ability to distinguish colors and shapes," and specific examples include "drawing, blocks, and pop-up books." If the product information for the product to be classified contains elements related to the definition or specific examples of "G. Visual and Spatial," the AI will classify the product into the "G. Visual and Spatial" category.

[0053] The definition of "H. Music and Rhythm" is "the ability to understand pitch and rhythm, and the ability to read emotions from vocal intonation," and specific examples include "singing, musical instruments, and performance." If the product information for the product to be classified contains elements related to the definition or specific examples of "H. Music and Rhythm," the AI will classify the product into the "H. Music and Rhythm" category.

[0054] The output specification information is information that indicates the specification of the output of tagged information, and is also information that informs the AI of the content and format of the information to be generated as tagged information. As an example, Figure 4 shows the output specification information as follows: "Given the 'product name' and 'detailed information,' consider the order in which of these eight perspectives you think are most likely to apply, and suggest the top three. When doing so, please include a reason for your suggestion in 200 characters or less. The output text should be 'Top 1 perspective: reason for suggestion,' 'Top 2 perspective: reason for suggestion,' and 'Top 3 perspective: reason for suggestion.'" Based on this output specification information, the AI recognizes how to generate and output tagging information.

[0055] (2-2) Content Information Storage Unit 322 The content information storage unit 322 has a function of storing product information. The content information storage unit 322 stores, for example, product information (an example of content information) uploaded from the customer terminal 10 when the customer classifies products.

[0056] (2-3) Tagging information storage unit 323 The tagging information storage unit 323 has a function of storing tagging information. The tagging information storage unit 323 stores tagging information generated by AI. The tagging information storage unit 323 stores the tagging information in association with the product information of the corresponding product among the product information stored in the content information storage unit 322. In this way, a tag is assigned (tagged) to the product.

[0057] (3) Control unit 330 The control unit 330 has a function of controlling the overall operation of the content classification server 30. The control unit 330 is realized, for example, by causing a CPU (Central Processing Unit) or a GPU (Graphics Processing Unit) that the content classification server 30 has as hardware to execute a program. As shown in FIG. 2, the control unit 330 includes an acquisition unit 331, a content classification unit 332, and an output control unit 333.

[0058] (3-1) Acquisition part 331 The acquisition unit 331 has a function of acquiring various types of information, such as prompt information and product information. The acquisition unit 331 acquires prompt information received from a terminal (not shown) of a service provider by the communication unit 310. The acquisition unit 331 stores the acquired prompt information in the prompt information storage unit 321. As a result, the prompt information is registered in the content classification server 30. Furthermore, the acquisition unit 331 acquires product information received by the communication unit 310 from the customer terminal 10. The acquisition unit 331 stores the acquired product information in the content information storage unit 322. As a result, the product information is registered in the content classification server 30. Note that the product information acquired by the acquisition unit 331 is product information on the product to be classified. Therefore, the acquisition unit 331 outputs the acquired product information to the content classification unit 332.

[0059] (3-2) Content Classification Unit 332 The content classification unit 332 has a function of classifying content. First, the content classification unit 332 determines a category that corresponds to the product indicated by the product information, based on the product information acquired by the acquisition unit 331 and the product information stored in the prompt information storage unit 321 (storage unit 320). At this time, the content classification unit 332 performs product classification indicated by the summary information included in the prompt information. In this product classification, the content classification unit 332 determines to which of multiple categories (tag types) indicated by the category information included in the prompt information the product indicated by the product information corresponds. Note that the number of tag types that correspond to one product may be either one or multiple.

[0060] In determining the type of tag that corresponds to the content, the content classification unit 332 determines the type of tag that corresponds to the content based on at least one of the definition and specific example indicated by the category information. For example, the content classification unit 332 determines the type of tag to which the content belongs based only on the definition indicated by the category information. In this case, the content classification unit 332 is able to perform classification with a higher degree of freedom because the classification direction is not as narrowed as when specific examples are used. Alternatively, the content classification unit 332 may determine the type of tag to which the content corresponds based only on the specific example indicated by the category information. In this case, the content classification unit 332 narrows the classification direction to a certain extent compared to when using definitions, resulting in classification with a lower degree of freedom, but in return, more accurate classification can be achieved. Furthermore, the content classification unit 332 may determine the type of tag to which the content corresponds based on both the definition and specific examples indicated in the category information. In this case, the content classification unit 332 narrows the classification direction further than when using only either the definition or specific examples, and the degree of freedom in classification is reduced, but the content classification unit 332 can perform classification with higher accuracy.

[0061] The content classification unit 332 generates tagging information to be assigned as a tag to a product for at least one category determined to be relevant in the product classification. At this time, the content classification unit 332 generates and outputs the tagging information in accordance with the designation of the output designation information included in the prompt information. When one product corresponds to one category, the content classification unit 332 generates one piece of tagging information. On the other hand, when one product corresponds to multiple categories, the content classification unit 332 generates multiple pieces of tagging information. The content classification unit 332 proposes to the customer the details of information to be attached to the content as tags based on the generated tagging information. If the customer confirms the proposed details and finds that changes are necessary, the content classification unit 332 regenerates the tagging information. If the customer confirms the proposed details and finds that changes are not necessary, the content classification unit 332 stores the tagging information in the tagging information storage unit 323. This causes the tagging information to be registered in the content classification server 30. Note that if the content classification unit 332 proposes multiple tags based on multiple pieces of tagging information, the customer may be able to select the required tag from the multiple tags. This can also omit the process of the content classification unit 332 regenerating the tagging information. The number of tags to be assigned to content may be specified by the customer. In this case, the content classification unit 332 generates tagging information for one product in the number specified by the customer. The customer specifies the number of tags to be assigned to content, for example, using output specification information. In the example of output specification information shown in FIG. 4, it is specified that the top three perspectives be proposed out of eight perspectives. That is, in the example of output specification information shown in FIG. 4, the customer specifies that three tags be assigned to the content. In this case, the content classification unit 332 generates tagging information for three categories determined to be relevant in the product classification.

[0062] In this embodiment, the content classification unit 332 classifies products using AI. In this case, the content classification unit 332 inputs product information of the products to be classified to the AI. The AI classifies the products based on the prompt information using the product information as input, and generates and outputs tagging information. The content classification unit 332 obtains the tagging information output from the AI as the classification result.

[0063] Here, a specific example of product information and tagging information according to this embodiment will be described with reference to Fig. 5. Fig. 5 is a diagram showing an example of product information and tagging information according to this embodiment.

[0064] Figure 5 shows specific examples of product information, which is information input to the AI, and tagging information, which is information output from the AI.

[0065] The product information input to the AI consists of the product name and detailed information. As an example, in Figure 5, the product name is "Find the ***!", and the detailed information is "This is a hidden object series filled with *** characters. It can be enjoyed by children from around 3 years old, but it's also fun for the whole family." Based on this product information, the AI classifies the product indicated by the product information.

[0066] The tagging information output by the AI is based on the output specification information included in the prompt information. For example, Figure 5 shows the tagging information as follows: "1st: Visual / Spatial: 'Find the ***!' is a series of hidden object games, and the goal is for children to find the *** character. This activity helps develop spatial awareness and visual exploration skills. 2nd: Introspection: This picture book allows children to express their interests and preferences, as it can be enjoyed by the whole family. The process of searching for the character also provides children with an opportunity to reflect on their own knowledge and memories. 3rd: Language / Language: Sharing information about the *** character helps children develop their language skills. Enjoying picture books with their families also improves their communication skills." In this example, the AI follows the output specification information and outputs the top three categories to which the product falls and the reasons for their recommendation (up to 200 characters) in the format of "1st Perspective: Reason for Recommendation," "2nd Perspective: Reason for Recommendation," and "3rd Perspective: Reason for Recommendation."

[0067] (3-3) Output control unit 333 The output control unit 333 has a function of controlling the output of various information. For example, the output control unit 333 transmits tagging information obtained by product classification by the content classification unit 332 to the customer terminal 10 via the communication unit 310, and causes it to be displayed. After registering the tagging information, the output control unit 333 transmits product information for viewing by consumers and the tagging information to the consumer terminal 20 via the communication unit 310, and causes it to be displayed.

[0068] <3. Processing flow> The functional configuration of the content classification server 30 according to this embodiment has been described above. Next, the flow of processing in the content classification system 1 according to this embodiment will be described with reference to Figs.

[0069] (1) Process flow for generating tagging information First, the flow of processing related to generation of tagging information in the content classification system 1 according to this embodiment will be described with reference to Fig. 6. Fig. 6 is a sequence diagram showing an example of the flow of processing related to generation of tagging information in the content classification system 1 according to this embodiment. It is assumed that the prompt information has already been registered in the prompt information storage unit 321 of the storage unit 320 of the content classification server 30.

[0070] 6, first, the customer performs an operation to create product information on the customer terminal 10 (step S101). Through this operation, the customer creates product information on the products to be classified as data in a CSV file format. The customer terminal 10 uploads the product information created by the customer to the content classification server 30 (step S102). The acquisition unit 331 of the content classification server 30 acquires the product information received by the communication unit 310 from the customer terminal 10 (step S103). The acquisition unit 331 stores and registers the acquired product information in the content information storage unit 322 of the storage unit 320 (step S104).

[0071] After registering the product information, the customer performs an operation to classify the products on the customer terminal 10 (step S105). In response to an operation by the customer, the customer terminal 10 transmits a product classification execution request to the content classification server 30 (step S106). When the communication unit 310 receives the product classification execution request from the customer terminal 10, the content classification unit 332 of the content classification server 30 executes product classification processing for the product information previously received and registered (step S107). After the product classification process, the output control unit 333 of the content classification server 30 transmits the tagging information generated by the content classification unit 332 to the customer terminal 10 via the communication unit 310 (step S108).

[0072] The customer terminal 10 displays the tagging information received from the content classification server 30 (step S109). The customer checks the tagging information displayed on the customer terminal 10 (step S110). In checking the tagging information, the customer determines whether or not the content of the tagging information needs to be changed (step S111). If a change is necessary (step S111 / YES), the process proceeds to step S112. On the other hand, if a change is not necessary (step S111 / NO), the process proceeds from connector A to step S201 in FIG. 7.

[0073] When the process proceeds to step S112, the customer performs an operation on the customer terminal 10 to request a re-proposal of tagging information (step S112). In response to the operation by the customer, the customer terminal 10 transmits a re-proposal request to the content classification server 30 (step S113). When the communication unit 310 receives the re-proposal request from the customer terminal 10, the content classification unit 332 of the content classification server 30 executes a re-proposal process (step S114). In the re-proposal process, the content classification server 30 repeats the process from step S107.

[0074] (2) Processing flow for registering and publishing tagging information Next, a process flow relating to the registration and publication of tagged information in the content classification system 1 according to this embodiment will be described with reference to Fig. 7. Fig. 7 is a sequence diagram showing an example of a process flow relating to the registration and publication of tagged information in the content classification system 1 according to this embodiment.

[0075] As shown in FIG. 7, when the process proceeds to step S201, the customer performs an operation to register tagging information on the customer terminal 10 (step S201). In response to an operation by the customer, the customer terminal 10 transmits a tagging information registration request to the content classification server 30 (step S202). When the communication unit 310 receives the tagging information registration request from the customer terminal 10, the content classification unit 332 of the content classification server 30 stores and registers the tagging information generated in step S107 in the tagging information storage unit 323 (step S203). After registration, the output control unit 333 of the content classification server 30 makes the product public (step S204).

[0076] The consumer operates the consumer terminal 20 to select the product for which the consumer wishes to check details (step S205). In response to the operation by the consumer, the consumer terminal 20 displays the product information of the selected product and its tagging information (step S206). The consumer views the product information and tag information of the product displayed on the consumer terminal 20 (step S207).

[0077] As described above, the content classification system 1 according to this embodiment includes a memory unit 320 that stores prompt information indicating a prompt used to classify content into a corresponding category among a plurality of categories; an acquisition unit 331 that acquires content information indicating details of the content to be classified; and a content classification unit 332 that determines the category that corresponds to the content indicated by the content information based on the acquired content information and the prompt information stored in the memory unit 320, and generates tagging information indicating information to be assigned as a tag to the content for the category that is determined to correspond.

[0078] With this configuration, customers can easily classify products simply by uploading product information to the content classification system 1, and can obtain more accurate classification results than if they were to classify products manually. Therefore, the content classification system 1 according to this embodiment can reduce the cost required for classifying content and improve the accuracy of content classification.

[0079] <4. Modifications> The above describes the embodiments. Next, modifications of the above-described embodiments will be described. Note that each modification described below may be applied to the embodiments alone or in combination with each other. Furthermore, each modification may be applied in place of the configuration described in the embodiments, or may be applied in addition to the configuration described in the embodiments.

[0080] In the above-described embodiment, an example has been described in which the service provider creates prompt information based on information provided by a customer and uploads it to the content classification server 30. However, the present invention is not limited to such an example. For example, the customer may create prompt information and upload it from the customer terminal 10 to the content classification server 30. The prompt information may be changeable after uploading, so that if there is a problem with the accuracy of the tagging information generated by the content classification server 30, the customer can adjust the accuracy of the tagging information by changing the prompt information.

[0081] The above describes a modified embodiment of the present invention. Note that some or all of the functions of the content classification system 1, the customer terminal 10, the consumer terminal 20, and the content classification server 30 in the above-described embodiment may be implemented by a computer. In this case, a program for implementing these functions may be recorded on a computer-readable recording medium, and the program may be loaded into a computer system and executed. Note that the term "computer system" as used herein includes hardware such as an operating system and peripheral devices. Furthermore, the term "computer-readable recording medium" refers to portable media such as flexible disks, optical magnetic disks, ROMs, and CD-ROMs, as well as storage devices such as hard disks built into computer systems. Furthermore, the term "computer-readable recording medium" may also include media that dynamically store programs for a short period of time, such as communication lines used when transmitting programs via networks such as the Internet or telephone lines, or media that store programs for a fixed period of time, such as volatile memory within the server or client computer system. Furthermore, the above program may be one that realizes part of the above-mentioned functions, or may be one that can realize the above-mentioned functions in combination with a program already recorded in a computer system, or may be one that is realized using a programmable logic device such as an FPGA (Field Programmable Gate Array).

[0082] The embodiments of the present invention have been described in detail above with reference to the drawings, but the specific configuration is not limited to that described above, and various design changes can be made within the scope of the gist of the present invention. [Explanation of symbols]

[0083] 1...content classification system, 10...customer terminal, 20...consumer terminal, 30...content classification server, 310...communication unit, 320...storage unit, 321...prompt information storage unit, 322...content information storage unit, 323...tagging information storage unit, 330...control unit, 331...acquisition unit, 332...content classification unit, 333...output control unit, NW...network

Claims

1. a storage unit that stores prompt information indicating a prompt used to classify content into a corresponding category among a plurality of categories; an acquisition unit that acquires content information indicating details of content to be classified; a content classification unit that determines a category that corresponds to the content indicated by the content information based on the acquired content information and the prompt information stored in the storage unit, and generates tagging information that indicates information to be assigned as a tag to the content for the category that is determined to correspond; A content classification system comprising:

2. the prompt information includes summary information indicating an overview of content classification, category information indicating details of the plurality of categories, and output designation information indicating designation of output of the tagging information, the content classification unit executes content classification indicated by the summary information, determines in the content classification to which of a plurality of categories indicated by the category information the content indicated by the content information falls, and generates and outputs the tagging information in accordance with the specification of the output specification information. The content classification system of claim 1 .

3. The category information is information indicating types of tags indicating the categories, definitions of the tag types, and specific examples corresponding to the definitions, the content classification unit determines the type of the tag to which the content corresponds based on at least one of the definition and the specific example. The content classification system of claim 2 .

4. The content classification unit suggests to the user the content of information to be tagged as a tag based on the generated tagging information, and regenerates the tagging information if the suggested content needs to be changed. The content classification system of claim 1 .

5. a storing step of storing prompt information in a storage unit, the prompt information indicating a prompt used to classify content into a corresponding category among a plurality of categories; an acquisition step of acquiring content information indicating details of the content to be classified; a content classification process for determining a category corresponding to the content indicated by the content information based on the acquired content information and the prompt information stored in the storage unit, and generating tagging information indicating information to be assigned as a tag to the content for the category determined to be the corresponding category; 1. A computer-implemented method for content classification, comprising:

6. Computer, a storage unit configured to store prompt information in a storage unit, the prompt information indicating a prompt used to classify content into a corresponding category among a plurality of categories; an acquisition means for acquiring content information indicating details of content to be classified; a content classification means for determining a category corresponding to the content indicated by the content information based on the acquired content information and the prompt information stored in the storage unit, and generating tagging information indicating information to be assigned as a tag to the content for the category determined to be the corresponding category; A program to function as a

Images