Search system, search method and search program

Abstract

We provide technology that supports product searches in digital catalogs. [Solution] The search system 100 includes a digital catalog system 120 capable of posting catalogs of multiple products, a UI 802 that sets one or more areas in each of the multiple product catalogs and enables specification of one or more pieces of attribute information included in the one or more areas, a data extraction unit 804 that extracts one or more pieces of attribute information from each of the multiple product catalogs as structured data 700, a product DB 850 that stores the structured data 700 for each of the multiple products, and an interaction IF 110 that interacts with a generation AI 130. The interaction IF 110 inputs a question received from a user to the generation AI 130, obtains from the generation AI 130 one or more keywords related to the one or more pieces of attribute information included in the question, searches the product DB 850 using a search key including the one or more keywords, and outputs information on one or more products included in the search results.

Description

[Technical Field]

[0001] The present disclosure relates to techniques for searching digital catalogs. [Background technology]

[0002] In recent years, the use of digital catalogs that can be viewed online has become widespread. Digital catalogs are sometimes provided as web applications. In such cases, users access the URL (Uniform Resource Locator) of the digital catalog service using a browser or other device. Digital catalogs offer various functions not available in paper media, such as product search functions.

[0003] Regarding digital catalogs, for example, Japanese Patent Laid-Open Publication No. 2022-172694 (Patent Document 1) discloses a catalog data distribution device that "links product catalog data with customer information and stores it, acquires customer information for identifying a customer, identifies a customer from the acquired customer information, acquires catalog data extraction information related to the identified customer, selects distribution catalog data corresponding to the acquired catalog data extraction information from the stored distribution catalog data, extracts catalog data corresponding to the selected distribution catalog data from the catalog data, identifies distribution destination information based on the acquired customer information, and distributes the extracted catalog data to the destination of the identified distribution destination information" (see [Abstract]). Furthermore, other technologies related to digital catalogs are disclosed, for example, in Patent Document 2. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Publication No. 2022-172694 [Patent Document 2] Japanese Patent Application Laid-Open No. 2000-090096 Summary of the Invention [Problem to be solved by the invention]

[0005] Users who purchase products such as building materials and machine parts select products based on various criteria. However, digital catalogs often do not systematically organize and display information about each product. As a result, users are unable to easily find desired products from digital catalogs. The technologies disclosed in Patent Documents 1 and 2 are not based on this background and do not have a function for easily searching for desired products, such as alternative products, from digital catalogs based on various criteria.

[0006] The present disclosure has been made in view of the above-described background, and an object of one aspect is to provide a technology for supporting product searches in digital catalogs. [Means for solving the problem]

[0007] According to one embodiment, a search system is provided. The search system includes a digital catalog system capable of listing catalogs of multiple products, a UI (User Interface) that sets one or more areas in each of the multiple product catalogs and enables specification of one or more pieces of attribute information included in the one or more areas, a data extraction unit that extracts one or more pieces of attribute information from each of the multiple product catalogs as structured data, a product DB (Database) that stores the structured data for each of the multiple products, and an interaction IF for interacting with a generation AI (Artificial Intelligence). The interaction IF inputs a question received from a user to the generation AI, obtains one or more keywords from the generation AI that are related to the one or more pieces of attribute information included in the question, searches the product DB using a search key that includes the one or more keywords, and outputs information on one or more products included in the search results.

[0008] In one aspect, searching the product database includes comparing, for each structured data item in a catalog of multiple products, each of one or more keywords with each of one or more pieces of attribute information corresponding to each of the one or more keywords, and determining, based on the results of the comparison, whether each of the multiple products satisfies the user's needs.

[0009] In one aspect, the dialogue IF is configured to obtain, based on the fact that the question contains text requesting a search for alternative products to a certain product, one or more keywords containing one or more attribute information of the certain product from the generation AI, and to search the product database using a search key containing the one or more attribute information of the certain product.

[0010] In one aspect, at least one of the digital catalog system and the dialogue IF is configured to input a first product number of a first product and a second product number of a second product compatible with the first product into the product DB to register relationship information between the first product and the second product, and the digital catalog system is configured to output first catalog information linked to the first product number and second catalog information linked to the second product number to the dialogue IF. The dialogue IF is configured to input the first catalog information and the second catalog information to the generation AI, obtain from the generation AI relationship information estimated based on the first catalog information and the second catalog information, and register the relationship information in the product DB.

[0011] In one aspect, the relationship information includes a product number for each product, at least a portion of the structured data, and the relationship between the products. At least a portion of the structured data is configured to include at least one or more of attribute information of the color, size, or specifications of each product.

[0012] In one aspect, the relationship between products is configured to be able to include at least one of a new / old relationship, an inclusion relationship, a different color, a different size, and a different specification.

[0013] In one aspect, searching a product database using a search key including one or more keywords includes searching for one or more alternative products that are linked to a product based on the relationships between the products and that satisfy conditions related to one or more attribute information included in the query.

[0014] According to another embodiment, there is provided a search method executed by a search system of a digital catalog system capable of publishing catalogs of multiple products. The method includes: setting one or more areas in each of the multiple product catalogs and specifying one or more attribute information included in the one or more areas; extracting the one or more attribute information from each of the multiple product catalogs as structured data; storing the structured data for each of the multiple products in a product database; inputting a question received from a user to a generation AI; obtaining from the generation AI one or more keywords related to the one or more attribute information included in the question; searching the product database using a search key including the one or more keywords; and outputting information on one or more products included in the search result.

[0015] According to another embodiment, there is provided a search program executed by a search system of a digital catalog system capable of publishing catalogs of multiple products. The search program causes the search system to perform the following operations: set one or more areas in each of the multiple product catalogs and specify one or more pieces of attribute information included in the one or more areas; extract the one or more pieces of attribute information from each of the multiple product catalogs as structured data; store the structured data for each of the multiple products in a product database; input a question received from a user to a generation AI; obtain from the generation AI one or more keywords related to the one or more pieces of attribute information included in the question; search the product database using a search key including the one or more keywords; and output information on one or more products included in the search results. [Effects of the Invention]

[0016] According to one embodiment, a digital catalog may support product searches. The above and other objects, features, aspects and advantages of the present disclosure will become apparent from the following detailed description of the disclosure taken in conjunction with the accompanying drawings. [Brief explanation of the drawings]

[0017] [Figure 1] FIG. 1 shows an example of a configuration of a search system 100 according to the present embodiment. [Figure 2] FIG. 10 is a diagram showing an example of a registration process of an alternative product in the search system 100. [Figure 3] FIG. 10 is a diagram illustrating an example of a search process for an alternative product in the search system 100. [Figure 4] FIG. 4 is a diagram showing an example of a device 400 used as the search system 100. [Figure 5] 10 is a diagram showing an example of a communication sequence of a search process for a substitute product in the search system 100. FIG. [Figure 6] FIG. 1 is a diagram illustrating an example of a product catalog. [Figure 7] FIG. 7 is a diagram showing an example of the configuration of structured data 700. [Figure 8] FIG. 2 is a diagram showing an example of the configuration of a search system 100 related to the registration process of structured data 700. [Figure 9] FIG. 8 is a diagram showing an example of a UI 802 provided by the digital catalog system 120. [Figure 10] FIG. 6 is a diagram showing an example of a result of setting a context in a product catalog 600. [Figure 11] FIG. 10 is a diagram showing an example of an extraction screen for an area 1080. [Figure 12] FIG. 10 is a diagram showing an example of a context setting in an area 1080. [Figure 13] FIG. 12 is a diagram showing an example of an extraction screen for an area 1290. [Figure 14] FIG. 12 is a diagram showing an example of a context setting in an area 1290. [Figure 15] FIG. 9 is a diagram showing an example of a screen 900 after a context is set. DETAILED DESCRIPTION OF THE INVENTION

[0018] Hereinafter, embodiments of the technical concept of the present disclosure will be described with reference to the drawings. In the following description, identical components are assigned the same reference numerals. Their names and functions are also the same. Therefore, detailed descriptions thereof will not be repeated. Furthermore, each embodiment, each modification, each software or program configuration, each hardware configuration, each function, each process, etc. may be selectively combined as appropriate.

[0019] <Search system functions> FIG. 1 is a diagram illustrating an example of the configuration of a search system 100 according to the present embodiment. The search system 100 provides a function for searching for substitute products for a given product from among products listed in a digital catalog. The products here may include building materials, mechanical parts, electronic parts, and any other products. Substitute products include any products that are compatible with a given product. In one aspect, a substitute product may be a successor to a given product. In another aspect, a substitute product may be a product with the same standards, shape, or specifications as a given product. Furthermore, in another aspect, a substitute product may include any product that can be used in place of a given product in a user's application. The search system 100 may be implemented as one or more devices. Therefore, the search system 100 may be interpreted as a search device.

[0020] The search system 100 includes an interactive IF 110, a digital catalog system 120, and a generation AI 130. The search system 100 provides a function for searching for alternative products to a given product by linking the generation AI 130 and the digital catalog system 120.

[0021] The interactive IF 110 has two main functions. The first function is an interactive function with the user terminal 140. The interactive IF 110 accepts questions from the user and provides answers to the questions to the user. As an example, the interactive IF 110 may interact with the user terminal 140 via a chat screen, a command line, or the like. In one aspect, the interactive IF 110 may provide the interactive function to the user terminal 140 as a web application. In another aspect, the user terminal 140 may interact with the interactive IF 110 via an installed interactive application.

[0022] The second function is to link the digital catalog system 120 and the generation AI 130. As an example, the dialogue IF 110 communicates with both the digital catalog system 120 and the generation AI 130 during the search process for alternative products. When executing the search process, the dialogue IF 110 inputs a question received from the user terminal 140 to the generation AI 130. The generation AI 130 outputs information to be used in the search process of the digital catalog system 120, such as a product number and keywords, to the dialogue IF 110 as a response. The dialogue IF 110 inputs the received information to the digital catalog system 120 as a search key and obtains the search results from the digital catalog system 120. Alternatively, the dialogue IF 110 may use the received information as a search key to search the alternative product DB 220 (see FIG. 2 , described later). The dialogue IF 110 may obtain the product number of the alternative product as the search result. The dialogue IF 110 also obtains catalog information (text, images, etc.) based on the obtained product number of the alternative product from the digital catalog system 120. Then, the dialogue IF 110 inputs the question and catalog information of the alternative product to the generation AI 130 and obtains a response in natural language from the generation AI 130. The dialogue IF 110 outputs the obtained response to the user terminal 140. Through this series of processes, the user can search for alternative products of a certain product from the digital catalog system 120 via a question in natural language, just as if they were conversing with a human.

[0023] The digital catalog system 120 provides a digital catalog service. The digital catalog system 120 has functions such as a product information registration function, a product information viewing function, and a product search function using keywords or product numbers. The digital catalog system 120 may be implemented as a web application. The digital catalog system 120 may also have an API (Application Programming Interface) for linking with other services or applications. The digital catalog system 120 is also configured to be able to communicate with an alternative product DB 220 that stores related information about compatible products. In some aspects, the alternative product DB 220 may be included in the digital catalog system 120.

[0024] The generation AI 130 has a function to extract product numbers, keywords, or a combination thereof from a question input by a user to the dialogue IF 110. The extracted information is used by the dialogue IF 110 to search the digital catalog system 120 or the alternative product DB 220. The generation AI 130 also has a function to generate an answer to the user based on the question and catalog information of alternative products input from the dialogue IF 110. The generation AI 130 also has a function to infer relationships between mutually compatible products. The inferred relationships between products are registered in the alternative product DB 220.

[0025] In one aspect, the generation AI 130 may be included in the search system 100. In another aspect, the generation AI 130 may exist outside the search system 100. In this case, the dialogue IF 110 is configured to be able to communicate with the external generation AI 130. In another aspect, the generation AI 130 may learn sentences and images registered in the digital catalog system 120 using RAG (Retrieval Augmented Generation).

[0026] As described with reference to FIG. 1 , the search system 100 includes a digital catalog system 120 capable of listing multiple products, an alternative product DB 220 storing relationship information between the multiple products and one or more alternative products compatible with each of the multiple products, and a dialogue IF 110 for dialogue with the generation AI 130. Upon receiving a question about an alternative product for a given product, the dialogue IF 110 inputs the question to the generation AI 130 and acquires keywords related to the given product included in the question from the generation AI 130. The dialogue IF 110 then searches the alternative product DB 220 using the acquired keywords as search keys to acquire product numbers for each of the one or more alternative products compatible with the given product from the alternative product DB 220. The dialogue IF 110 then inputs the acquired product numbers for each of the one or more alternative products into the digital catalog system 120 and acquires catalog information for each of the one or more alternative products from the digital catalog system 120. Furthermore, the interaction IF 110 is configured to output the question and catalog information of each of the one or more alternative products to the generation AI 130, and to output the answer obtained from the generation AI 130 to the user terminal 140.

[0027] With the above configuration and series of processes, the search system 100 can search for alternative products to a certain product from a product lineup with a complex product number system. Also, a user can interact with the search system 100 using natural language.

[0028] <Registration process for replacement products> FIG. 2 illustrates an example of a registration process for substitute products in the search system 100. The search system 100 can extract and store relationship information between compatible products by linking the generation AI 130 and the digital catalog system 120. Hereinafter, the relationship information between compatible products will be referred to as "relationship information between products" or "relationship information." The relationship information between products includes the product number and attributes of each product and the relationship between the products. The attributes may include at least one or more of any product-related information, such as the color, size, or specifications of each product. The relationship between products may include any relationship based on the product number system estimated from the catalog information. As an example, the relationship between products may include at least one or more of any relationships, such as a new-old relationship, an inclusion relationship, a different color, a different size, or different specifications. As a specific example, assume that product B is the successor to product A. In this case, a new-old relationship exists between products A and B. Furthermore, assume that products C1, C2, and C3 exist that are different sizes. In this case, a relationship of different sizes exists between products C1, C2, and C3.

[0029] The search system 100 can easily search for alternative products to a given product by referencing the relationship information between products. Also, the search system 100 can search for alternative products that are compatible with a given product and match the user's needs (color, specifications, etc.) by referencing the attributes and the relationships between products.

[0030] 1, the search system 100 includes a batch processing unit 210 and an alternative product DB 220 as components related to the registration process of alternative products. The batch processing unit 210 cooperates with the digital catalog system 120 and the interactive IF 110 to acquire related information 230 and register the related information 230 in the alternative product DB 220.

[0031] The substitute product DB 220 stores the inter-product relationship information 230 described above. The substitute product DB 220 can be realized as a database in any format. As one example, the substitute product DB 220 may be a graph database. As another example, the substitute product DB 220 may be a database in any format, such as a document database or a relational database. The substitute product DB 220 may also be included in the digital catalog system 120. The substitute product DB 220 may also be a database management system. Furthermore, the substitute product DB 220 has functions for creating, referencing, updating, and deleting data.

[0032] Next, as an example, a specific procedure for registering relationship information between products will be described using the process of registering relationship information between a first product and a second product that are compatible in the substitute product DB 220. The following description is an example, and the search system 100 can register relationships between any number of products in the substitute product DB 220.

[0033] In step S210, the digital catalog system 120 receives an input of a list of alternative products from the terminal 225 of a manufacturer's representative. The list of alternative products may be expressed in any data format. The list of alternative products includes part numbers of multiple compatible products. In this example, the list of alternative products includes part numbers of a first compatible product and a second compatible product.

[0034] In step S220, digital catalog system 120 outputs the full catalog text and images for each product included in the alternative product list to batch processor 210. In this example, digital catalog system 120 outputs the full catalog text and images for each of the first product and the second product to batch processor 210.

[0035] In step S230, the batch processing unit 210 outputs the full text and images of the catalog for each product included in the obtained substitute product list to the dialogue IF 110. In step S240, the dialogue IF 110 inputs the full text and images of the catalog for each product included in the obtained substitute product list to the generation AI 130.

[0036] In step S250, the generation AI 130 extracts relationship information between products from the full-text catalog text and images of each product included in the input alternative product list. For example, the generation AI 130 extracts attributes such as the color, size, or specifications of each product from the input data. The generation AI 130 also extracts relationships between products, such as new-old relationships, inclusion relationships, different colors, different sizes, or different specifications, from the input data. In this example, the generation AI 130 may extract information from the input data, such as that a first product and a second product have a new-old relationship and that the second product is a successor to the first product. The generation AI 130 outputs the extracted relationship information between products to the interaction IF 110.

[0037] In step S260, the interactive IF 110 outputs the obtained relationship information between products to the batch processing unit 210. In step S270, the batch processing unit 210 registers the obtained relationship information between products in the substitute product DB 220. In this example, the relationship information between the first product and the second product is registered in the substitute product DB 220. The search system 100 can execute the registration process of such relationship information between products at any time. As a result, the relationship information 230 is accumulated in the substitute product DB 220.

[0038] In one aspect, the dialogue IF 110 may accept input of the alternative product list instead of the digital catalog system 120. In this case, the dialogue IF 110 obtains the full text and images of the catalog of each product included in the alternative product list from the digital catalog system 120. In another aspect, the dialogue IF 110 may include a batch processing unit 210. Furthermore, in another aspect, the generation AI 130 may be fine-tuned in advance via the administrator's terminal 240 in order to extract relationship information between products with high accuracy.

[0039] As described with reference to FIG. 2 , at least one of the digital catalog system 120 and the interaction IF 110 is configured to input a first product number of the first product and a second product number of the second product when registering relationship information between a first product and a second product compatible with the first product in the substitute product DB 220. The digital catalog system 120 is also configured to output first catalog information linked to the first product number and second catalog information linked to the second product number to the interaction IF 110. The interaction IF 110 is configured to input the first catalog information and the second catalog information to the generation AI 130, obtain relationship information between the first product and the second product estimated based on the first catalog information and the second catalog information from the generation AI 130, and register the relationship information in the substitute product DB 220. The interaction IF 110 may communicate directly with the substitute product DB 220. The interaction IF 110 may also communicate with the alternative product DB 220 via the batch processing unit 210 .

[0040] The relationship information may also include product numbers and attributes for each product, as well as relationships between products. The attributes can include at least one of the color, size, and specifications for each product. The relationships between products include relationships based on product number systems estimated from catalog information.

[0041] Furthermore, the relationship based on the product number system is configured to be able to include at least one of a new / old relationship, an inclusion relationship, a color difference, a size difference, and a specification difference.

[0042] In this way, the search system 100 uses the generation AI 130 to extract and store relationship information between products from the information in the digital catalog. This enables the search system 100 to realize the search process shown in Fig. 3. Furthermore, the search system 100 not only associates the part numbers of compatible products but also stores relationship information such as the relationship between old and new products and different colors, thereby enabling the search system 100 to search for alternative products suitable for the user's needs.

[0043] <Search process for alternative products> Fig. 3 is a diagram showing an example of a search process for an alternative product in the search system 100. With reference to Fig. 3, a process will be described in which the search system 100 searches for a second product that is an alternative product to a first product and presents information about the second product to the user. The following description is an example, and the search system 100 may present information about two or more products as alternative products to a certain product.

[0044] In step S305, the dialogue IF 110 accepts an input of a question from the user terminal 140. In this example, the dialogue IF 110 accepts an input of a question including the product number of the first product, "AA2023." In one aspect, the dialogue IF 110 may accept a question in natural language via the chat screen 300. In another aspect, the dialogue IF 110 may accept an input of only the product number as a question. In yet another aspect, the dialogue IF 110 may accept an input of the product number and one or more conditions as a question. The conditions may include attributes of the substitute product, such as color, size, and specifications. The conditions may also be interpreted as attributes. As one example, a user may search for a substitute product that is compatible with a certain building material and has a color of "red." As another example, a user may search for a substitute product that is compatible with a certain electronic component and has a withstand voltage of "200 V or higher."

[0045] In step S310, the dialogue IF 110 inputs the acquired question to the generation AI 130. The generation AI 130 extracts keywords from the question. The keywords include at least the product part number. The keywords may also include conditions the user desires for the replacement product. The conditions may include any attributes such as color, size, price, etc. In step S315, the dialogue IF 110 acquires keywords from the generation AI 130.

[0046] In step S320, the dialogue IF 110 searches the alternative product DB 220 using the acquired keyword as a search key. More specifically, the dialogue IF 110 inputs a search request including the keyword to the alternative product DB 220. Based on the search request, the alternative product DB 220 refers to the relationship information 230 and searches for an alternative product that meets the search criteria. In this example, the alternative product DB 220 searches the relationship information 230 for an alternative product to the first product.

[0047] In step S325, the interactive IF 110 obtains search results from the alternative product DB 220. If an alternative product is found, the search results include at least the product number of the alternative product. If an alternative product is not found, the search results include information indicating that no alternative product exists. In this example, the alternative product DB 220 outputs search results including the second product's product number "AA2024" to the interactive IF 110.

[0048] In step S330, the interaction IF 110 uses the acquired product number of the alternative product as a search key to search the digital catalog system 120. More specifically, the interaction IF 110 inputs a search request including the product number of the alternative product to the digital catalog system 120.

[0049] In step S335, the interactive IF 110 obtains search results from the digital catalog system 120. The search results include catalog information such as text information and images of the alternative products. In one aspect, the search results may include the entire text of the catalog page of the alternative product. In another aspect, the search results may include images of the alternative product. In another aspect, the search results may include the URL of the catalog page of the alternative product. In yet another aspect, the search results may include any combination of this information.

[0050] In step S340, the dialogue IF 110 inputs the question and the search results to the generation AI 130. The generation AI 130 creates an answer in natural language based on the question and the search results. In step S345, the dialogue IF 110 obtains the answer from the generation AI 130. In step S350, the dialogue IF 110 transmits the obtained answer to the user terminal 140. In this example, the dialogue IF 110 transmits the part number "AA2024" and URL of the second product to the user terminal 140. In one aspect, the dialogue IF 110 may include at least one piece of information in the answer, such as a list of alternative products, summaries of the alternative products, and thumbnails of catalog pages of the alternative products.

[0051] As described with reference to FIG. 3, obtaining keywords related to a certain product included in a question from generation AI 130 may include extracting the product number of the certain product from the question, extracting attribute information required for an alternative product to the certain product, and including the product number of the certain product and the attribute information required for the alternative product in the keywords.

[0052] In addition, searching the alternative product DB 220 using a keyword as a search key may include searching for one or more alternative products that are linked to a certain product based on the relationship between the products and that include the required attribute information.

[0053] In this way, the search system 100 can search for alternative products for a given product by referencing the relationship information between products. Furthermore, the search system 100 can search for alternative products suitable for the user's needs by narrowing down the alternative products by combining product numbers and conditions.

[0054] <Hardware configuration of the search system> FIG. 4 is a diagram showing an example of a device 400 used as search system 100. In one aspect, search system 100 may be realized by one device 400. In another aspect, search system 100 may be realized by a plurality of devices 400. In another aspect, each component such as interactive IF 110 and digital catalog system 140 may be realized by an individual device 400. Furthermore, search system 100 may be built on a cloud environment made up of a plurality of devices 400. Each component shown in FIGS. 1 to 3 can be realized by executing a program on the hardware shown in FIG. 4.

[0055] The device 400 includes a processor 401, a memory 402, a storage 403, an external device IF 404, an input IF 405, an output IF 406, and a communication IF 407. The device 400 also includes a bus 408 that connects these components. In some aspects, the device 400 may include two or more of each component, or may not include some of the components. The device 400 may also include any component not shown in FIG. 4 as necessary.

[0056] The processor 401 may execute programs for implementing various functions of the device 400. The processor 401 may be configured, for example, with at least one integrated circuit. According to an embodiment, the integrated circuit may include at least one central processing unit (CPU), at least one graphics processing unit (GPU), at least one field programmable gate array (FPGA), at least one application specific integrated circuit (ASIC), at least one AI chip, or a combination thereof.

[0057] The memory 402 functions as a workspace for the processor 401. The memory 402 stores programs executed by the processor 401 and data referenced by the processor 401. In one aspect, the memory 402 can be realized by a dynamic random access memory (DRAM), a static random access memory (SRAM), or the like.

[0058] Storage 403 is a non-volatile memory that stores programs executed by processor 401 and data referenced by processor 401. Processor 401 executes programs read from storage 403 to memory 402 and references data read from storage 403 to memory 402. In one aspect, storage 403 can be realized by a hard disk drive (HDD), a solid state drive (SSD), an erasable programmable read only memory (EPROM), an electrically erasable programmable read only memory (EEPROM), a flash memory, or the like.

[0059] The external device IF 404 can be connected to any external device such as a printer, a scanner, an external HDD, etc. In one aspect, the external device IF 404 can be realized by a USB (Universal Serial Bus) terminal or the like.

[0060] The input IF 405 can be connected to any input device such as a keyboard, a mouse, a touchpad, a gamepad, etc. In one aspect, the input IF 405 can be realized by a USB terminal, a PS / 2 terminal, a Bluetooth (registered trademark) module, etc.

[0061] The output IF 406 can be connected to any output device such as a cathode ray tube display, a liquid crystal display, an organic electroluminescence (EL) display, etc. In one aspect, the output IF 406 can be realized by a USB terminal, a D-sub terminal, a DVI (Digital Visual Interface) terminal, an HDMI (High-Definition Multimedia Interface) terminal, a DisplayPort terminal, etc.

[0062] The communication IF 407 is connected to other devices via a wired network or a wireless network. In one aspect, the communication IF 407 may be implemented by a wired local area network (LAN) port, a Wi-Fi (registered trademark) (Wireless Fidelity) module, or the like. In another aspect, the communication IF 407 may transmit and receive data using a communication protocol such as TCP / IP (Transmission Control Protocol / Internet Protocol) or UDP (User Datagram Protocol).

[0063] <Search system communication sequence> FIG. 5 is a diagram showing an example of a communication sequence of a search process for alternative products in search system 100. In one aspect, each component shown in FIG. 5 may be a program module or the like. In this case, each component does not need to be physically separated and may be built on the same device or in a cloud environment. In another aspect, processor 401 may load a program for performing the process of FIG. 5 from storage 403 into memory 402 and execute the program. In another aspect, some or all of the process may be realized as a combination of circuit elements configured to perform the process. Furthermore, in another aspect, the order of the following steps may be reversed.

[0064] In step S505, the user terminal 140 transmits a question to the interactive IF 110 via the chat screen 300. The question may be in natural language or may be a keyword. For example, if a user wants to search for an alternative product to a certain product, the user includes the product number of the product in the question. Furthermore, the user may include requirements for the alternative product in the question. The requirements may include attribute information required for the alternative product, such as color, size, and specifications. In some aspects, the user terminal 140 may transmit a question to the interactive IF 110 via any screen other than the chat screen 300. The screen on which the question is input may be provided by the search system 100 or may be a screen of an application installed on the user terminal 140.

[0065] In step S510, the dialogue IF 110 receives a question from the user terminal 140 via the chat screen 300. In step S515, the dialogue IF 110 inputs the received question to the generation AI 130. In step S520, the generation AI 130 breaks down the question and extracts keywords. As an example, the generation AI 130 extracts a product number from the question. Furthermore, the generation AI 130 may extract conditions required for an alternative product from the question. In step S525, the generation AI 130 outputs the extracted keywords to the dialogue IF 110.

[0066] In step S530, the interaction IF 110 inputs a search request including keywords to the alternative product DB 220. In step S535, the alternative product DB 220 searches the relationship information 230 for alternative products that meet the search criteria. The search criteria may include at least the product number of a certain product. The alternative product DB 220 searches the relationship information 230 for products that are compatible with the certain product as alternative products. Furthermore, the search criteria may include, in addition to the product number of the certain product, conditions required for the alternative product. In this case, the alternative product DB 220 searches the relationship information 230 for products that are compatible with the certain product and that meet the conditions. In step S540, the alternative product DB 220 outputs the part numbers of one or more alternative products as search results to the interaction IF 110. If no alternative products exist, the alternative product DB 220 outputs a search result indicating that no alternative products exist to the interaction IF 110.

[0067] In step S545, the interactive IF 110 inputs the product numbers of the one or more acquired alternative products to the digital catalog system 120. In step S550, the digital catalog system 120 outputs the catalog ID (Identifier), page ID, and catalog URL of each of the one or more input alternative products to the interactive IF 110.

[0068] In step S555, the interaction IF 110 inputs at least one of the acquired catalog ID, page ID, and catalog URL to the digital catalog system 120. That is, the interaction IF 110 requests catalog information of an alternative product identified by at least one of the acquired catalog ID, page ID, and catalog URL from the digital catalog system 120. In step S560, the digital catalog system 120 extracts the full text of the catalog information of the alternative product identified by at least one of the catalog ID, page ID, and catalog URL. Furthermore, the digital catalog system 120 may also extract images included in the catalog information. In step S565, the digital catalog system 120 outputs the extracted full text to the interaction IF 110. Furthermore, the digital catalog system 120 may output images in addition to the full text to the interaction IF 110. In some aspects, the full text may be data represented in HyperText Markup Language (HTML), Extensible Markup Language (XML), or any other format. In another aspect, based on the product number acquired in step S545, the digital catalog system 120 may output the full text of the product catalog identified by the product number to the interactive IF 110. In this case, the processes of steps S550 and S555 may not be executed.

[0069] In step S570, the dialogue IF 110 inputs the question acquired in step S510, the full text acquired in step S565, and linking information between the product number included in the question and the product number listed in the full text to the generation AI 130. The dialogue IF 110 may extract the product number included in the question, search the full text for a page containing the product number, and create linking information for the product number. The linking information for the product number may include linking information between the product number included in the question, the product number included in the full text, and the catalog page number or location of the product number. The dialogue IF 110 may use the generation AI 130 to extract the product number included in the question. In some aspects, the dialogue IF 110 may further input any information, such as a catalog URL, to the generation AI 130. In step S575, the generation AI 130 generates an answer based on the input information. The answer is expressed in natural language and may include any information about alternative products, such as a summary, a product number, a catalog URL, or the like. In step S580, the generation AI 130 outputs the generated answer to the dialogue IF 110.

[0070] In step S585, the dialogue IF 110 outputs the acquired answer and catalog URL to the user terminal 140 via the chat screen 300. In one aspect, the catalog URL may be included in the generated answer. In another aspect, the dialogue IF 110 may output the catalog URL acquired in step S550 to the user terminal 140. In step S590, the user terminal 140 receives the answer and catalog URL from the dialogue IF 110 via the chat screen 300.

[0071] 5, the search system 100 can execute a series of processes from steps S505 to S590 by executing a program on hardware. That is, the search system 100 can execute the following operations: accept an input of a question about an alternative product of a certain product listed in the digital catalog system 120, which can list multiple products; input the question to the generation AI 130; acquire keywords related to the certain product included in the question from the generation AI 130; search for relationship information defining relationships between the multiple products using the keywords as search keys; acquire product numbers of one or more alternative products that are compatible with the certain product; input the product numbers of the one or more alternative products to the digital catalog system 120; acquire catalog information for each of the one or more alternative products from the digital catalog system 120; output the question and the catalog information for each of the one or more alternative products to the generation AI 130; and output the answer acquired from the generation AI 130 to the user terminal 140.

[0072] <Using structured data> 1 to 5, the process of registering and searching for alternative products for a given product by search system 100 has been described. Search system 100 has not only the function of registering and searching for alternative products, but also the function of converting data contained in a product catalog into structured data and storing it. By referring to the structured data, search system 100 can interpret the user's intended use and present products suitable for that intended use.

[0073] FIG. 6 shows an example of a product catalog. Product catalog 600 is an example of a product catalog displayed by digital catalog system 120. Product catalog formats vary depending on the company, product type, etc. Therefore, conventional search systems cannot easily analyze the content (hereinafter referred to as "context") described in a product catalog, even if they refer to a product catalog database. Context encompasses any information described in a product catalog, such as product photos, model numbers, types, and product descriptions.

[0074] The search system 100 analyzes the context of each catalog registered in the digital catalog system 120 and provides a function for creating structured data from attribute information included in the context. The attribute information includes each piece of information included in the structured data 700 (category, item, unit, order code, etc.). By referencing the structured data, the search system 100 can search for a product that meets a user's needs from among multiple products registered in the digital catalog system 120. Hereinafter, the procedure by which the search system 100 creates structured data will be described using the product catalog 600 as an example.

[0075] 7 is a diagram showing an example of the configuration of structured data 700. The structured data 700 has a hierarchical structure. The structured data 700 includes one or more hierarchical categories 710, one or more items 730, and one or more units 750.

[0076] The categories 710 indicate hierarchical categories of items. For example, the category "Handles" may include smaller categories "Backstop Handles" and "Surface-Mounted Handles." While two categories are shown in the example of FIG. 7, this is merely an example. The structured data 700 may include any number of categories 710. The categories 710 may be represented by keywords or IDs that indicate the categories.

[0077] The items 730 correspond to products listed in the catalog. As an example, the product catalog 600 lists four products (items) with product model numbers "A001," "B001," "C001," and "D001." In this case, the search system 100 may extract four items 730 from the product catalog 600. The items 730 may be expressed by keywords or IDs that identify the items. Each item 730 may include detailed item information 740. The detailed item information 740 may include any information related to the item, such as a series ID, series name, catchphrase, and accessories for the item.

[0078] The units 750 correspond to the different specifications of each product. As an example, the product model number "A001" listed in the product catalog 600 includes three units "AA-XX001, AA-XX002, and AA-XX003." These units differ in dimensions, etc. In this case, the search system 100 can extract three units 750 related to the item 730 with the product model number "A001" from the product catalog 600. Each unit 750 can include unit detail information 760. The unit detail information 760 can include any information related to the unit, such as an order code, product number, material, finish, etc. The units 750 can be represented by a keyword or ID that identifies the unit.

[0079] The search system 100 creates structured data 700 by analyzing multiple product catalogs registered in the digital catalog system 120. The structured data 700 can be used for product search processing via the interactive IF 110. The structured data 700 can be expressed in any data format.

[0080] Fig. 8 is a diagram showing an example of the configuration of the search system 100 related to the registration process of structured data 700. Although omitted in Fig. 8, the search system 100 may have all of the components shown in Fig. 2. The digital catalog system 120 extracts or generates various information 810 from the registered product catalog 600. The various information 810 includes images, text information, coordinate information for each character, character direction and block information, etc., published in the product catalog 600.

[0081] The search system 100 extracts structured data 700 from various information 810. To accomplish this, the search system 100 includes a UI 802, a data extraction unit 804, and a product DB 850. Some or all of the UI 802, the data extraction unit 804, and the product DB 850 may be included in the digital catalog system 120.

[0082] The UI 802 is a UI for editing the product catalog 600 in order to extract structured data from the various information 810. More specifically, the user may divide the product catalog 600 into multiple areas via the UI 802 and set a context to be included in each area. Furthermore, the user may set one or more pieces of attribute information to be included in each area within the context. For example, assume that a certain area includes a context of information about a unit. The context of the information about the unit may include one or more pieces of attribute information about the unit. Details of the editing process of the product catalog 600 using the UI 802 will be described later with reference to FIGS. 9 to 15.

[0083] The data extraction unit 804 extracts one or more pieces of attribute information as structured data from the product catalog 600 edited by the UI 802. The data extraction unit 804 stores the extracted structured data in the product DB 850. The data extraction unit 804 inputs a prompt 820 containing the context of the information located at each coordinate in the product catalog to the generation AI 130. That is, the data extraction unit 804 inputs the information set in the product catalog 600 by the UI 802 to the generation AI 130. The search system 100 also inputs various pieces of information 810 related to the product catalog 600 to the generation AI 130. The data extraction unit 804 may input the various pieces of information 810 to the generation AI 130.

[0084] The generation AI 130 outputs structured data 700 based on various information 810 and prompt 820. The data extraction unit 804 acquires the structured data 700 from the generation AI 130 and stores it in the product DB 850. The product DB 850 may include information from the alternative product DB 220. The structured data 700 may also include relationship information 230 for alternative products. The batch processing unit 210 may cause the search system 100 to perform a series of operations from extraction to storage of the structured data 700 based on an input operation on the UI 802.

[0085] When the search system 100 stores the structured data 700, it can provide users with more advanced product search functions. By referencing the structured data 700 for each product, the search system 100 can search the entire product catalog for products that meet various user needs. For example, the dialogue IF 110 inputs a question received from a user into the generation AI 130. The dialogue IF 110 then acquires, from the generation AI 130, one or more keywords related to one or more attribute information items included in the question. The attribute information items may include any information included in the product catalog 600, such as dimensions, finish, and color. The dialogue IF 110 searches the product DB 850 using a search key including one or more keywords and outputs information about one or more products included in the search results. That is, the dialogue IF 110 outputs information about one or more products that meet the product conditions included in the question.

[0086] 7 and 8, the search system 100 includes a digital catalog system 120 capable of listing catalogs of multiple products, a UI 802 capable of setting one or more areas in each of the multiple product catalogs and specifying one or more pieces of attribute information included in the one or more areas, a data extraction unit 804 for extracting one or more pieces of attribute information from each of the multiple product catalogs as structured data 700, a product DB 850 for storing the structured data 700 for each of the multiple products, and an interaction IF 110 for interacting with the generation AI 130. The interaction IF 110 is configured to input a question received from a user to the generation AI 130, obtain from the generation AI 130 one or more keywords related to the one or more pieces of attribute information included in the question, search the product DB 850 using a search key including the one or more keywords, and output information on one or more products included in the search results.

[0087] Furthermore, searching product DB 850 includes comparing each of the one or more keywords with each of the one or more pieces of attribute information corresponding to each of the one or more keywords, for each of structured data 700 of the catalogs of the multiple products. Furthermore, searching product DB 850 includes determining whether each of the multiple products satisfies the user's needs based on the result of the comparison.

[0088] Furthermore, based on the fact that the question contains text requesting a search for an alternative product to a certain product, the dialogue IF 110 may be configured to obtain one or more keywords including one or more attribute information of the certain product from the generation AI 130, and to search the product DB 850 using a search key including the one or more attribute information of the certain product. In this way, the dialogue IF 110 can search the product DB 850 for products that have attribute information common to the attribute information of the certain product.

[0089] Furthermore, at least one of digital catalog system 120 and interaction IF 110 may be configured to input a first product number of a first product and a second product number of a second product compatible with the first product into product DB 850, thereby registering relationship information 230 between the first product and the second product. In this case, digital catalog system 120 is configured to output first catalog information linked to the first product number and second catalog information linked to the second product number to interaction IF 110. Interaction IF 110 is configured to input the first catalog information and the second catalog information to generation AI 130, obtain relationship information 230 estimated based on the first catalog information and the second catalog information from generation AI 130, and register the relationship information in product DB 850.

[0090] Furthermore, relationship information 230 includes the product number for each product, at least a portion of structured data 700, and the relationship between the products. At least a portion of structured data 700 is configured to be able to include at least one or more of attribute information of color, size, or specifications for each product.

[0091] Furthermore, searching the product DB 850 using a search key including one or more keywords includes searching for one or more alternative products that are linked to a certain product based on the relationship between the products and that satisfy conditions related to one or more attribute information included in the question. For example, suppose that a user inputs a question about an alternative product to a certain product into the dialogue IF 110. In this case, the search system 100 may search for the alternative product by referring to the relationship information 230. Furthermore, if the question includes conditions such as dimensions or color, the search system 100 may further refer to the structured data 700 to determine whether the alternative product included in the relationship information 230 satisfies the conditions included in the question. Furthermore, the search system 100 may search the structured data 700 in the product DB 850 using one or more conditions included in the question as a search key. In this case, the search system 100 does not need to refer to the relationship information 230.

[0092] 6 to 8 by executing a program on hardware. That is, the search system 100 can set one or more areas in each of a plurality of product catalogs and specify one or more pieces of attribute information included in the one or more areas, extract one or more pieces of attribute information from each of the plurality of product catalogs as structured data 700, store the structured data 700 for each of the plurality of products in the product DB 850, input a question received from a user to the generation AI 130, obtain from the generation AI 130 one or more keywords related to the one or more pieces of attribute information included in the question, search the product DB 850 using a search key including the one or more keywords, and output information on one or more products included in the search results.

[0093] <Structured data registration screen> 9 is a diagram showing an example of a UI 802 provided by the digital catalog system 120. A screen 900 is an example of a screen displayed by the digital catalog system 120. A context setting button 910 is located at the top of the screen 900. In the example of FIG. 9, the product catalog 600 is displayed on the screen 900.

[0094] The user may transition screen 900 to a context setting mode by pressing context setting button 910. In the context setting mode, the user may select a desired area 920 on product catalog 600. Based on the selection of area 920, screen 900 displays a dialog 930 for inputting a context. Dialog 930 includes a context input item 932 and a scope setting item 934.

[0095] The user may input what context is included in the selected area 920 in the context input field 932. That is, the user may set an area in the product catalog 600 and specify the context included in the area. As an example, the user may set individual attribute information for some or all of any area selected from the product catalog via the context input field 932. Hereinafter, the operation input for selecting an area from the product catalog 600 and the operation input for specifying the context of the area will be collectively referred to as the "setting input." The effective range setting field 934 is configured to allow input of the range to which the setting input is to be reflected. For example, the user may select the setting reflection range in the effective range setting field 934 to automatically reflect the same setting in other catalogs of the same format.

[0096] The user may repeatedly enter setting inputs for each part of the product catalog 600. By repeatedly receiving the setting inputs, the digital catalog system 120 may obtain information to be included in the prompt 820. The context set for each area may include one or more pieces of attribute information. The search system 100 extracts the one or more pieces of attribute information included in the context set for each area and includes the one or more pieces of attribute information in the structured data 700.

[0097] FIG. 10 is a diagram showing an example of the result of setting the context in the product catalog 600. The product catalog 600 shown in FIG. 10 is a cut-out portion of the screen 900. The product catalog 600 shown in FIG. 10 also shows a state where a set of setting inputs has been received. The context of areas 1010 and 1020 is the page number. The context of areas 1030 and 1040 is the classification (first level). The context of areas 1050 and 1060 is the classification (second level). The context of area 1070 is annotations for the catalog or the entire page. The context of areas 1080, 1085, 1090, and 1095 is product information.

[0098] FIG. 11 is a diagram showing an example of an extraction screen for area 1080. UI 802 provides a function for selecting areas hierarchically and setting a context for each hierarchical area. In the example of FIG. 11, area 1080 showing product information is displayed in an enlarged form. UI 802 may also enlarge the selected area based on the selection of an arbitrary area.

[0099] FIG. 12 is a diagram showing an example of context settings in area 1080. Area 1080 shown in FIG. 12 shows a state in which a set of setting inputs has been received. The context of area 1210 is a logo. The context of area 1220 is a series name. The context of area 1230 is a series model number. The context of area 1240 is a pictogram indicating materials, etc. The context of area 1250 is an image (bird's-eye view). The context of area 1260 is a catchphrase. The context of area 1270 is a product description. The context of area 1280 is an image (dimension drawing). The context of area 1290 is unit information.

[0100] Fig. 13 is a diagram showing an example of an extraction screen for area 1290. In the example of Fig. 13, area 1290 showing unit information is displayed in an enlarged manner. The UI 802 can further enlarge an area set within the area.

[0101] FIG. 14 is a diagram showing an example of context settings in area 1290. Area 1290 shown in FIG. 14 shows a state in which a set of setting inputs has been received. The context of area 1405 is information on compliance with the RoHS (Restriction of Hazardous Substances) Directive. The context of area 1410 is the presence or absence of CAD (Computer Aided Design) data. The context of area 1415 is the order code. The context of area 1420 is the product number. The context of area 1425 is the material. The context of area 1430 is the finish. The context of area 1435 is size L. The context of area 1440 is size P. The context of area 1445 is size H. The context of area 1450 is weight. The context of area 1455 is quantity (boxes). The context of area 1460 is quantity (cartons). These contexts may be set as attribute information of each area.

[0102] 9 to 14, the UI 802 can divide the product catalog registered in the digital catalog system 120 into multiple areas. The UI 802 can also perform division processing of the product catalog hierarchically. Furthermore, the UI 802 is configured to be able to set the context included in each area.

[0103] A context includes one or more pieces of attribute information extracted as part of structured data 700. As an example, the context of area 1420 includes the attribute information "product number." That is, the context of area 1420 includes one piece of attribute information. Similarly, each of areas 1405 to 1460 includes one piece of attribute information. Furthermore, the context of area 1290 includes all of the attributes included in areas 1405 to 1460. That is, the context of area 1420 includes 12 pieces of attribute information.

[0104] FIG. 15 is a diagram showing an example of a screen 900 after context setting. The screen 900 further includes a structured data extraction button 1500. After completing the setting input for the product catalog 600, the user can extract structured data 700 from the product catalog 600 by pressing the structured data extraction button 1500. The search system 100 can extract the structured data 700 from the product catalog 600 using the generation AI 130 based on the pressing of the structured data extraction button 1500. In this case, the search system 100 can generate a prompt 820 from the setting input for the product catalog 600 and input the prompt 820 to the generation AI 130. The prompt 820 can include context or attribute information set for each area and an instruction to cause the generation AI 130 to extract one or more attribute information from each area. Furthermore, the search system 100 can register the extracted structured data 700 in the product DB 850.

[0105] Furthermore, search system 100 may display dialog 1510 based on pressing structured data extraction button 1500. The user may select and download part or all of structured data 700 from dialog 1510. The user may register the downloaded structured data 700 in product DB 850.

[0106] Furthermore, the search system 100 may extract information selected in the dialog 1510 as structured data 700. As an example, if item information is selected, the search system 100 may extract attribute information of an item (product) from the product catalog 600. As another example, if unit information is selected, the search system 100 may extract attribute information of a unit from the product catalog 600.

[0107] <Summary> As described above, the search system 100 according to this embodiment has the function of estimating and registering relationships between products, and the function of interpreting user questions and registering alternative products, by linking the digital catalog system 120 and the generation AI 130. This enables the search system 100 to search for alternative products to a given product from a product lineup with a complex product number system.

[0108] Furthermore, the search system 100 provides a function for extracting structured data 700 from a product catalog. The search system 100 can easily search for products that meet various user needs by interpreting a user's query and searching the structured data 700.

[0109] The embodiments disclosed herein should be considered to be illustrative in all respects and not restrictive. The scope of the present disclosure is defined by the claims, not by the above description, and is intended to include all modifications within the meaning and scope equivalent to the claims. Furthermore, the disclosures described in the embodiments and each modification are intended to be implemented, as far as possible, either alone or in combination. [Explanation of symbols]

[0110] 100 Search system, 110 Dialogue IF, 120 Digital catalog system, 130 Generative AI, 140 User terminal, 210 Batch processing unit, 220 Substitute product DB, 225, 240 Terminal, 230 Relationship information, 300 Chat screen, 400 Device, 401 Processor, 402 Memory, 403 Storage, 404 External device IF, 405 Input IF, 406 Output IF, 407 Communication IF, 408 Bus, 600 Product catalog, 700 Structured data, 710 Category, 730 Item, 740 Item details, 750 Unit, 760 Unit details, 804 Data extraction unit, 810 Various information, 820 Prompt, 850 Product DB, 900 Screen, 910 Context setting button, 920,1010,1020,1030,1040,1050,1060,1070,1080,1085,1090,1095,1210,1220,1230,1240,1250,1260,1270,1280,1290,1405,1410,1415,1420,1425,1430,1435,1440,1445,1450,1455,1460 area, 930,1510 dialog, 932 context input field, 934 validity range setting field, 1500 structured data extraction button.

Claims

1. A digital catalog system that can display catalogs of multiple products, a UI (User Interface) that sets one or more areas in each of the catalogs of the plurality of products and allows specification of one or more pieces of attribute information included in the one or more areas; a data extraction unit for extracting the one or more pieces of attribute information as structured data from each of the catalogs of the plurality of products; a product database (DB) that stores the structured data for each of the plurality of products; An interaction IF for interacting with the generating AI (Artificial Intelligence), The interaction IF is A question received from a user is input to the generating AI, obtaining, from the generating AI, one or more keywords related to the one or more pieces of attribute information included in the question; searching the product DB using a search key including the one or more keywords; A search system that outputs information about one or more products included in search results.

2. The searching of the product DB includes: comparing, for each of the structured data of each of the catalogs of the plurality of products, each of the one or more keywords with each of the one or more pieces of attribute information corresponding to each of the one or more keywords; The search system according to claim 1 , further comprising: determining whether each of the plurality of products satisfies the user's needs based on a result of the comparison.

3. The interaction IF is Based on the fact that the question includes text requesting a search for an alternative product to a certain product, the one or more keywords including the one or more attribute information of the certain product are acquired from the generation AI; The search system according to claim 1 or 2, configured to search the product DB using the search key including the one or more pieces of attribute information of the certain product.

4. At least one of the digital catalog system and the interaction IF is a first product number of a first product and a second product number of a second product that is compatible with the first product are input into the product DB, thereby enabling relationship information between the first product and the second product to be registered; the digital catalog system is configured to output first catalog information linked to the first product number and second catalog information linked to the second product number to the interactive IF; The interaction IF is inputting the first catalog information and the second catalog information into the generating AI; obtaining, from the generating AI, the relationship information estimated based on the first catalog information and the second catalog information; The search system according to claim 1 or 2, configured to register the relationship information in the product DB.

5. the relationship information includes a product number for each product, at least a part of the structured data, and a relationship between products; The search system according to claim 4 , wherein at least a portion of the structured data is configured to be capable of including at least one of attribute information of color, size, and specifications for each product.

6. The search system according to claim 5 , wherein the relationship between the products is configured to be able to include at least one of a new / old relationship, an inclusion relationship, a different color, a different size, and a different specification.

7. The search system of claim 3, wherein searching the product DB using the search key including the one or more keywords includes searching for one or more alternative products that are linked to the certain product based on the relationship between the products and that satisfy conditions related to the one or more attribute information included in the question.

8. A search method executed by a search system of a digital catalog system capable of listing catalogs of multiple products, setting one or more areas in each of the catalogs of the plurality of products and specifying one or more pieces of attribute information to be included in the one or more areas; extracting the one or more pieces of attribute information as structured data from each of the catalogs of the plurality of products; storing the structured data for each of the plurality of products in a product DB; Inputting a question received from a user into the generation AI; Obtaining, from the generating AI, one or more keywords related to the one or more attribute information included in the question; searching the product DB using a search key including the one or more keywords; and outputting information about one or more products included in the search results.

9. A search program executed by a search system of a digital catalog system capable of listing catalogs of multiple products, setting one or more areas in each of the catalogs of the plurality of products and specifying one or more pieces of attribute information to be included in the one or more areas; extracting the one or more pieces of attribute information as structured data from each of the catalogs of the plurality of products; storing the structured data for each of the plurality of products in a product DB; Inputting a question received from a user into the generation AI; Obtaining, from the generating AI, one or more keywords related to the one or more attribute information included in the question; searching the product DB using a search key including the one or more keywords; and outputting information on one or more products included in the search results.

Images