Information processing system, information processing method, and program

Abstract

To consolidate commodity masters in a plurality of sales channels, thereby efficiently updating information and ensuring consistency.SOLUTION: A CPU 11 of a management server 1 constituting an information processing system includes: an information management unit 32 which stores commercial material information regarding commercial materials in a first format, and stores the commercial material information in a second style; an update detection unit 33 which detects that the commercial material information has been updated; and a conversion unit 39 which converts, when the update of the commercial material information stored in the first format is detected, the commercial material information from the first format to the second format.SELECTED DRAWING: Figure 3

Description

[Technical Field]

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

[0002] BACKGROUND ART Techniques for supporting management of information relating to products provided to customers are known (for example, Patent Document 1). [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Publication No. 2019-067048 Summary of the Invention [Problem to be solved by the invention]

[0004] However, for companies that offer products through multiple sales channels (e.g., e-commerce sites, physical stores, etc.), product information is managed separately for each channel, which requires a significant amount of effort to maintain consistency and update the information.

[0005] The present invention has been made in view of the above circumstances, and aims to realize the unified management of product master data across multiple sales channels, thereby improving the efficiency of information updates and ensuring consistency. [Means for solving the problem]

[0006] In order to achieve the above object, one aspect of the present invention is to The information processing system comprises a first format storage means for storing product information relating to a product in a first format, a second format storage means for storing the product information in a second format, an update detection means for detecting that the product information has been updated, and a format conversion means for converting the product information from the first format to the second format when an update of the product information stored in the first format is detected.

[0007] An information processing device, an information processing method, and a program corresponding to the information processing system according to one aspect of the present invention are also provided as an information processing device, an information processing method, and a program corresponding to the information processing system according to one aspect of the present invention. [Effects of the Invention]

[0008] According to the present invention, it is possible to realize the centralized management of product master data across multiple sales channels, thereby improving the efficiency of information updates and ensuring consistency. [Brief explanation of the drawings]

[0009] [Figure 1] 1 is a diagram illustrating an example of the overall configuration of an information processing system according to an embodiment of the present invention. [Figure 2] 2 is a block diagram showing an example of a hardware configuration of a main management server that constitutes the information processing system of FIG. 1. FIG. [Figure 3] FIG. 2 is a functional block diagram showing an example of the functional configuration of a main management server. [Figure 4] 10 is a flowchart showing an example of a processing flow of a main management server. [Figure 5] 1. FIG. 4 is a diagram showing a specific example of a setting screen displayed on a user terminal constituting the information processing system of FIG. DETAILED DESCRIPTION OF THE INVENTION

[0010] Hereinafter, this embodiment will be described with reference to the drawings. <Information Processing System S> FIG. 1 is a diagram showing an example of the overall configuration of an information processing system S according to one embodiment of the present invention. The information processing system S is an information processing system configured to include a main management server 1, an inventory management server 2, a purchasing management server 3, external systems 4-1 to 4-n (n is an integer value of 1 or greater), and a user terminal 5, all connected via a network N. In the following, unless it is necessary to explain each of the external systems 4-1 to 4-n individually, they will be collectively referred to as the "external system 4." The external system 4 includes, for example, a server that provides a generation AI.

[0011] Network N is a communications network connecting the main management server 1, inventory management server 2, purchasing management server 3, external system 4, and user terminal 5. This component is the infrastructure that enables communication between all systems centered around the main management server 1 and serves as the basis for data exchange and collaborative processing. Network N integrates various communications technologies, such as the Internet, dedicated local area networks (LANs), virtual private networks (VPNs), and mobile communications networks, to provide optimal communications paths tailored to the characteristics and requirements of the connected systems. A variety of standard protocols, such as HTTPS (REST API, SOAP API, etc.), SFTP, MQTT, and WebSocket, are supported for communications with external systems 4, allowing for flexible adaptation to the technical specifications of each system. To ensure communications security, multi-layered defense mechanisms are implemented, including data encryption (e.g., TLS / SSL), mutual authentication (e.g., client certificates), IP address restrictions, and firewalls. Furthermore, optimization technologies such as load balancing, redundant configurations, caching, and compressed transfer are also applied to enhance communication stability and efficiency. Network N is also equipped with a communication status monitoring function, constantly monitoring indicators such as bandwidth utilization, latency, and packet loss to maintain communication quality. It also has an anomaly detection function, which automatically issues alerts and implements workarounds if signs of communication failure or unauthorized access are detected. These functions enable Network N to achieve safe and efficient data distribution between the main management server 1 and each external system, and function as an important infrastructure that supports the reliability of information synchronization and collaborative processing.

[0012] The information processing system S is a system that allows users to receive a service (hereinafter referred to as "this service") that centrally manages and automatically synchronizes information about products ("product information") across multiple sales channels, such as e-commerce sites and brick-and-mortar stores. A user company that receives this service manages a product database 51 in a first standard format through a main management server 1, enabling it to automatically distribute product information optimized for each sales channel. Specifically, this service allows product information to be automatically converted and transmitted to external systems 4, such as malls / delivery platforms and brick-and-mortar store POS systems, in the formats required by each of these systems.

[0013] In addition, by linking with the generation AI, this service can provide users with advanced information processing services such as correcting inconsistencies in expressions in product information (hereinafter referred to as "variations in notation"), automatically generating appealing statements, resizing images, tagging, etc. Users can utilize the generation AI to optimize product information simply by defining conversion rules through the settings screen.

[0014] In the information processing system S, real-time inventory information and purchasing information are associated with product information through cooperation between the main management server 1, inventory management server 2, purchasing management server 3, and external system 4. This enables inventory management according to sales status and comprehensive product analysis including cost information. New product information from the purchasing management server 3 is automatically imported into the main management server 1 and distributed in a form optimized for each sales channel.

[0015] That is, the information processing system S enables two-way information sharing. For example, the information processing system S enables feedback (e.g., sales information, customer reviews, etc.) from the external systems 4-1 to 4-n to the main management server 1 to be used to update and analyze the product DB 51. Based on this integrated information, users can search and analyze product information using queries in natural language. The generation AI converts this into appropriate command statements and processes them.

[0016] This service provided by information processing system S allows users to significantly reduce the burden of managing products across multiple channels while providing consistent information and performing advanced data analysis.

[0017] [Main management server 1] The main management server 1 constituting the information processing system S is an information processing device that serves as a server for managing the entire information processing system S. The main management server 1 may be constructed in a cloud environment. The main management server 1 can be implemented as a single physical server, or it may be distributed across multiple virtual machines in a cloud environment to ensure scalability and availability. The main management server 1 is managed, for example, by an administrator of the information processing system S. The main management server 1 is capable of executing predetermined application programs that enable the information processing system S to be used. The main management server 1 transmits various types of information to each of the inventory management server 2, the purchasing management server 3, the external system 4, and the user terminal 5, and enables them to execute various processes. The main management server 1 also acquires various types of information transmitted from each of the inventory management server 2, the purchasing management server 3, the external system 4, and the user terminal 5, and enables them to execute various processes.

[0018] The main management server 1 realizes the centralized management of product information across multiple sales channels and automatic synchronization with an external system 4. The main management server 1 is a core information processing device for realizing the centralized management of product information. The main management server 1 comprehensively manages data linkage between the inventory management server 2, the purchasing management server 3, and the external system 4. The main management server 1 is capable of both batch processing and real-time processing, and ensures the consistency of parallel processing and transaction processing for efficiently processing large amounts of product information. Furthermore, in terms of security, the main management server 1 protects confidential data such as product information and customer information through encrypted communication and authentication processing.

[0019] [Inventory management server 2] The inventory management server 2 is a server that manages the inventory status of products in real time. The inventory management server 2 has the function of sending and receiving master update and transaction information to and from the main management server 1, and accumulating inventory fluctuation information in a data lake. The main functions of the inventory management server 2 include inventory management (e.g., recording arrivals, shipments, returns, etc.), inventory quantity management (e.g., status management such as actual inventory, reserved inventory, and on-order inventory), inventory allocation management (e.g., location management by store, warehouse, shelf position, etc.), expiration management (e.g., tracking best-before dates, expiration dates, and disposal schedules), and inventory management (e.g., supporting regular physical inventory).

[0020] In cooperation with the main management server 1, the inventory management server 2 mainly communicates with the main management server 1 to receive product information (for example, registering product information for new products, updating product information for existing products, etc.) and to send inventory status information (for example, updating inventory quantities, inventory receipt and delivery history, etc.). This ensures consistency between product information and inventory information, and guarantees the accuracy of inventory display and sales availability judgments in each sales channel. Inventory fluctuation information is also accumulated in a data lake and is used for advanced analysis such as correlation analysis with sales data and demand forecasting.

[0021] The inventory management server 2 is equipped with a function for optimizing inventory, minimizing both excess inventory and the risk of stockouts through functions such as calculating the optimum inventory level for each product, managing reorder points, and automatically suggesting orders. Furthermore, the inventory management server 2 is equipped with a function for linking with the logistics system, optimizing the entire supply chain by issuing picking instructions, managing deliveries, and ensuring traceability. With these functions, the inventory management server 2, together with the main management server 1, efficiently manages the physical distribution of merchandise, thereby achieving both the accuracy of merchandise information and inventory optimization.

[0022] [Purchase management server 3] The purchasing management server 3 is a server that manages information related to product purchasing. The purchasing management server 3 sends catalog update information to the main management server 1 and provides transaction information related to purchasing, including cost details, to the main management server 1. The purchasing management server 3 also functions as an ERP / purchasing system, linking with the information processing system S. Furthermore, the purchasing management server 3 also has the function of scanning product proposals and sending the digitized information to the main management server 1.

[0023] The main functions of the purchasing management server 3 include order management (e.g., order planning, purchase order issuance, order progress management, etc.), supplier management (e.g., business partner information, terms of trade, evaluation information, etc.), price management (e.g., purchase price negotiation, price history management, etc.), arrival management (e.g., arrival schedule, inspection, return processing, etc.), and payment management (e.g., invoice matching, payment schedule management, etc.).

[0024] In cooperation with the main management server 1, the purchasing management server 3 mainly communicates with the main management server 1 to send information on new products and updated product catalogs, share purchasing costs and terms of trade, and notify of expected arrivals. This enables the rapid registration of new products, accurate reflection of price revisions, and the linking of inventory plans and purchasing plans. In addition, unstructured document data such as product plans and specifications is scanned and digitized through OCR and AI analysis, and is used to enrich product information.

[0025] The purchasing management server 3 has a cost analysis function. This enables the purchasing management server 3 to support decision-making, such as analyzing purchase cost trends, optimizing orders, and evaluating suppliers. Furthermore, the purchasing management server 3 has a predictive ordering function. This enables the purchasing management server 3 to make automatic ordering proposals based on demand forecasts linked to sales data, and to support the development of advance ordering plans that take seasonality into account. With these functions, the purchasing management server 3 can support the development of plans to improve the efficiency of the upstream product supply chain. With these functions, the purchasing management server 3 efficiently manages the upstream product supply chain, secures necessary products at appropriate costs and quality, and provides basic product information to the main management server 1.

[0026] [External System 4] External systems 4 are external systems that can link with the main management server 1 via network N, each with different product information formats and constraints. External systems 4 include a variety of systems, such as EPR / sales systems, e-commerce malls / delivery platforms, POS systems, and search engines, each with its own unique roles and characteristics. These external systems 4 are the main data exchange destinations for the main management server 1, and product information managed in a standard format (first format) is converted into an external format (second format) to suit the requirements of each external system 4 and distributed.

[0027] The external system 4 has its own data structure, communication protocol, and constraints (e.g., character limit, format requirements for product images, etc.), and details of these are recorded in the information management unit 32. The external system 4 acquires product information optimized by the main management server 1. In addition, feedback from the external system 4 (e.g., sales performance, inventory updates, customer reviews, etc.) is received by the main management server 1, used to update the product DB 51, and stored in the data lake. The external system 4 and the main management server 1 are linked bidirectionally, and the timing of information transmission and reception is optimized according to the characteristics of the external system 4. For example, a flexible linking structure is adopted, such as immediate synchronization for systems with high real-time requirements (e.g., POS systems) and periodic synchronization for systems based on batch processing (e.g., some e-commerce malls). With these functions, the external system 4 functions as an important component of a product information ecosystem centered on the main management server 1, achieving consistency and optimization of product information across multiple channels.

[0028] Of the external systems 4, the EPR / Sales system is a system that manages a company's management resources and sales activities in an integrated manner. The EPR / Sales system receives master updates from the main management server 1 and sends transaction information (e.g., sales details) to the main management server 1. The EPR / Sales system has an Enterprise Resource Planning (ERP) function that comprehensively manages a company's core business operations, as well as management functions specialized for sales activities. The main functions of the EPR / Sales system include accounting management (e.g., financial accounting, management accounting, etc.), sales management (e.g., order processing, shipping management, billing management, etc.), inventory management (e.g., inventory quantity, valuation value management, etc.), human resources management (e.g., employee information, payroll calculation, etc.), and purchasing management (e.g., order management, procurement management, etc.).

[0029] In the EPR / Sales System, the main communication content in linking with the Main Management Server 1 is the reception of product information and the transmission of sales transaction information (for example, order information, sales results, etc.). This allows the latest product information to be used consistently within the ERP, and transaction data such as sales results and inventory fluctuations is aggregated in the data lake via the Main Management Server 1, strengthening the analysis platform.

[0030] The EPR / Sales System also has a management analysis function, enabling multifaceted sales analysis by product, channel, region, etc., profitability evaluation, and budget / actual management. The EPR / Sales System also has a function to support budget formulation, enabling budget setting based on past performance data and future forecasts, and sales plan formulation support. With these functions, the EPR / Sales System not only plays an important role as a company's management foundation, but also enables the use of the latest and most accurate product information and the aggregation and analysis of detailed sales data through collaboration with the main management server 1.

[0031] Among the external systems 4, the e-commerce mall / delivery platform is an online shopping mall. Each external system 4 has its own product information format and constraints (e.g., image size, aspect ratio, etc.), and receives product information converted into an appropriate format by the main management server 1. Unlike a company's own e-commerce site, an e-commerce mall / delivery platform is characterized as a huge online marketplace with multiple merchants. Therefore, to ensure consistency and searchability across the platform, strict constraints are placed on the format and presentation of product information. Specific constraints include, for example, limits on the number of characters for product names and descriptions, a fixed system for categorizing products, uniform product image sizes and formats, and standardized input formats for product attribute information.

[0032] In cooperation with the main management server 1, the EC mall / delivery platform obtains product information converted to conform to these constraints, ensuring a system in which the information is displayed accurately and effectively on each platform. In addition, feedback such as sales performance, customer reviews, and inventory fluctuation information is provided to the main management server 1 from each platform, and is used for comprehensive sales analysis and inventory management.

[0033] E-commerce malls / delivery platforms have their own promotional functions. For this reason, the main management server 1 optimizes product information to match the campaigns and advertising functions of each platform. These functions make the e-commerce mall / delivery platform an important part of a company's sales channel, and by linking with the main management server 1, they realize the centralized management and optimal display of product information across multiple platforms.

[0034] Among the external systems 4, the POS system is a system that records and manages sales data at stores. The POS system has the function of sending POS data to the data lake of the main management server 1. The POS system serves as a contact point for collecting detailed transaction data at the point of sale in a physical store, and is therefore essential for integrating sales data with online channels. Key functions of the POS system include sales registration (e.g., product scanning, price inquiry, discount processing, etc.), payment processing (e.g., cash, credit card, electronic money, etc.), receipt issuance, sales aggregation, and inventory linkage.

[0035] In cooperation with the main management server 1, the POS system mainly communicates with the main management server 1 by receiving merchandise information and sending detailed sales data (for example, sales by product, sales by time period, customer demographic information, etc.). The POS data sent from the POS system includes all transaction details, such as the date and time of the sale, store information, register number, operator information, product code, quantity, unit price, discount information, and payment method, and this data is stored in the data lake of the main management server 1.

[0036] Furthermore, the correspondence between product barcodes and product codes and product information is also managed through the main management server 1, and a system has been established in which the registration of new products and updates to product information are immediately reflected in the POS systems of all stores. The POS system has a function for checking in-store inventory, allowing store staff and customers to check product inventory status (for example, in-store inventory, inventory at other stores, warehouse inventory, etc.) in real time. With these functions, the POS system plays a central role in collecting sales data from physical store channels, and, through cooperation with the main management server 1, realizes integrated management of product information and sales data across online and offline sales channels.

[0037] Of the external systems 4, the generation AI is an artificial intelligence system that supports the conversion and analysis of product information. The generation AI operates in conjunction with the main management server 1 and has the function of supporting the processing of a high-performance database that can handle 1 billion records and 500 columns. The generation AI is an AI system that combines advanced natural language processing capabilities based on a large-scale language model (LLM) with image recognition and image generation capabilities, and plays a central role in the advanced processing and analysis of product information.

[0038] The main functions of generative AI include text generation (e.g., product descriptions, appealing statements, etc.), text conversion (e.g., translation, correction of spelling variations, summarization, etc.), image analysis (e.g., feature extraction from product images, category estimation, etc.), image processing (e.g., resizing, cropping, style adjustment, etc.), and data analysis support (e.g., pattern detection, anomaly detection, predictive model building, etc.).

[0039] In cooperation with the main management server 1, the generation AI receives instructions (hereinafter referred to as "prompts") generated by the main management server 1, converts and analyzes product information based on the contents of the prompts, and returns the results to the main management server 1. The generation AI also cooperates with the main management server 1 to support the preprocessing and abstraction of large-scale data, enabling efficient knowledge extraction from massive amounts of product information and transaction data.

[0040] The Generative AI has a continuous learning function, and through feedback of processing results, it improves its accuracy while adapting to the industry- and company-specific expression styles and product characteristics. Furthermore, the Generative AI has multimodal processing capabilities, and by handling different types of data, such as text data, image data, and numerical data, it achieves more advanced product information processing. These functions allow the Generative AI to function as the core of advanced information conversion and analysis in the Main Management Server 1, efficiently achieving large-scale data processing and complex optimization that would be difficult for humans to achieve.

[0041] [User terminal 5] The user terminal 5 is an information processing device operated by a user of this service. The user terminal 5 is configured, for example, as a personal computer, a tablet terminal, a smartphone, etc. The user terminal 5 is capable of executing a predetermined application program that enables use of the information processing system S. The user terminal 5 is capable of executing various processes based on various information transmitted from each of the main management server 1, the inventory management server 2, the purchasing management server 3, and the external system 4, as well as various information input by the user. The user terminal 5 is also capable of transmitting various information to each of the main management server 1, the inventory management server 2, the purchasing management server 3, and the external system 4.

[0042] The above-described processes performed by the main management server 1, inventory management server 2, purchase management server 3, external system 4, and user terminal 5 constituting the information processing system S are merely examples. In other words, it is sufficient for the information processing system S as a whole to have the functions to realize the above-described processes, and some or all of the functions to realize the above-described processes may be shared or cooperated within the information processing system S.

[0043] For example, some or all of the functions of the main management server 1 may be functions of other devices in the information processing system S. Also, some or all of the functions of other devices in the information processing system S may be functions of the main management server 1. Furthermore, some or all of the functions of the main management server 1 may be transferred to other servers (not shown). This promotes processing in the information processing system S as a whole and also makes it possible for processes to complement each other.

[0044] <Hardware configuration> [Main management server 1 hardware configuration] FIG. 2 is a block diagram showing an example of the hardware configuration of the main management server 1 that constitutes the information processing system S of FIG. The main management server 1 includes a CPU (Central Processing Unit) 11, a ROM (Read Only Memory) 12, a RAM (Random Access Memory) 13, a bus 14, an input / output interface 15, an output unit 16, an input unit 17, a memory unit 18, a communication unit 19, and a drive 20.

[0045] The CPU 11 executes various processes in accordance with programs recorded in the ROM 12 or programs loaded from the storage unit 18 into the RAM 13. The RAM 13 also stores data and the like required for the CPU 11 to execute various processes. The CPU 11, the ROM 12, and the RAM 13 are interconnected via a bus 14. An input / output interface 15 is also connected to this bus 14.

[0046] The input / output interface 15 is connected to an output unit 16, an input unit 17, a storage unit 18, a communication unit 19, and a drive 20. The output unit 16 is composed of a display, a speaker, etc., and outputs various types of information as images, sounds, etc. The input unit 17 is composed of a keyboard, a mouse, a touch panel, etc., and accepts input of various types of information. The storage unit 18 is composed of a hard disk, a DRAM (Dynamic Random Access Memory), etc., and stores various types of data. The communication unit 19 communicates with other devices via the above-mentioned network N, which is composed of the Internet, etc.

[0047] Removable media 21, such as a magnetic disk, optical disk, magneto-optical disk, or semiconductor memory, is appropriately attached to the drive 20. Programs read from the removable media 21 by the drive 20 are installed in the storage unit 18 as needed. The removable media 21 can also store various data stored in the storage unit 18 in the same way as the storage unit 18.

[0048] [Hardware configuration of inventory management server 2, purchase management server 3, and user terminal 5] The inventory management server 2, purchasing management server 3, and user terminal 5 each have the same hardware configuration as the main management server 1 shown in Fig. 2. That is, the inventory management server 2, purchasing management server 3, and user terminal 5 each have a CPU, ROM, RAM, bus, input / output interface, output unit, input unit, memory unit, communication unit, drive, and removable media (not shown), which correspond to the CPU 11, ROM 12, RAM 13, bus 14, input / output interface 15, output unit 16, input unit 17, memory unit 18, communication unit 19, drive 20, and removable media 21 shown in Fig. 2, respectively.

[0049] <Functional configuration of main management server 1> FIG. 3 is a functional block diagram showing an example of the functional configuration of the main management server 1. As shown in FIG. In the CPU 11 of the main management server 1, an information acquisition unit 31 as feedback acquisition means, and an information management unit 32 as first format storage means, second format storage means, update history recording means, synchronization result storage means, priority storage means, specification storage means, and update history storage means function during operation. In addition, in the CPU 11, an update detection unit 33 as update detection means, an external connection unit 34 as external connection means, an external synchronization unit 35 as external synchronization means and synchronization identification means, and a display control unit 36 ​​as display control means function during operation. In addition, in the CPU 11, a prompt generation unit 37 as prompt generation means, an abstraction unit 38 as abstraction means, a conversion unit 39 as format conversion means and command statement conversion means, and a version management unit 40 as version management means function during operation. In addition, in the CPU 11, a query generation unit 41 as query generation means, a consistency verification unit 42 as consistency verification means, an attribute correction unit 43 as attribute correction means, and a transmission control unit 44 function during operation.

[0050] Furthermore, various databases are provided in the storage unit 18 of the main management server 1. For example, databases such as a product DB 51 storing product information, a priority DB 52 storing priority information, and a specification DB 53 storing specification information are provided.

[0051] The information acquisition unit 31 acquires various types of information. For example, the information acquisition unit 31 acquires various types of information transmitted to the main management server 1 from each of the inventory management server 2, the purchase management server 3, the external system 4, and the user terminal 5. For example, the information acquisition unit 31 acquires feedback of product information from the external system 4.

[0052] Specifically, the information acquisition unit 31 detects updates and status changes related to product information that occur in each external system 4 and provides a function that serves as an entry point for reflecting the updates in the product information stored in the product DB 51. This allows information that occurs only in the external system 4 (e.g., sales performance, customer reviews, inventory fluctuations, etc.) to be acquired, improving the comprehensiveness and freshness of the product information. The method for acquiring feedback is not particularly limited. For example, when actively acquiring information, new information is acquired by periodically polling the API of the external system 4 or checking the file sharing area. On the other hand, when passively acquiring information, notification is received in real time when a change occurs via callbacks, webhooks, message queues, etc. from the external system 4.

[0053] The information acquisition unit 31 normalizes the format of the acquired feedback. Specifically, the information acquisition unit 31 converts the format specific to the external system 4 into a standard format. Next, the information acquisition unit 31 verifies the feedback. Specifically, the information acquisition unit 31 checks the consistency and validity of the feedback. If the verification results in no problems, the processing is delegated to the information management unit 32, and the results are reflected in the product information.

[0054] The information acquisition unit 31 has an arbitration function when conflicting feedback about the same product is received from multiple external systems 4. This makes it possible to select optimal information based on the priority of the external systems 4, the freshness and reliability of the information, etc. The information acquisition unit 31 also has the function of recording and analyzing the acquisition history. This allows tracking of the frequency and content of feedback from which external systems 4, which can be used to optimize the acquisition process.

[0055] The information acquisition unit 31 also has a function for detecting anomalies. For example, if a feedback pattern that is significantly different from normal (such as a sudden price fluctuation or a large inventory decrease) is detected, the user is notified. These functions of the information acquisition unit 31 enable two-way information sharing between the external system 4 and the product information stored in the product DB 51, ensuring consistency and freshness of information throughout the system.

[0056] The information management unit 32 stores and manages various types of information in the database of the storage unit 18. For example, the information management unit 32 stores and manages product information in the product DB 51 in a first format or a second format. The specification information includes a wide range of information such as data structure definitions (e.g., field names, data types, length restrictions, etc.), communication protocol information (e.g., API specifications, endpoint URLs, etc.), authentication information (e.g., API keys, authentication methods, etc.), constraints (e.g., character count restrictions, prohibited characters, image sizes, etc.), and business rules (e.g., category systems, tax rate settings, etc.). The specification information is referenced by the conversion unit 39 and consistency verification unit 42, which will be described later, and is used for conversion into a format compatible with an external system and verification processing.

[0057] The information management unit 32 has a function for managing specification changes (e.g., API version upgrades) of the external system 4, and analyzes and manages the history of specification updates of the external system 4 and the scope of impact of the changes. The information management unit 32 also manages information on the operating status and processing limits (e.g., rate limits, batch size limits, etc.) of the external system 4. Such information is used to control optimal synchronization processing. Furthermore, the information management unit 32 manages processing patterns unique to the external system 4 (e.g., special error handling, unique conversion logic, etc.). Such information is used for customization processing for each external system 4.

[0058] The information management unit 32 has a function to automatically collect specification information. This enables automatic acquisition of API schema information and learning of constraints through test collaboration. These functions of the information management unit 32 realize reliable collaboration with various external systems 4 and improve the accuracy and efficiency of conversion and verification processes.

[0059] The information management unit 32 stores the content of updates detected by the update detection unit 33 and the update date and time. The information management unit 32 chronologically manages changes to the product information stored in the product DB 51. The product information update history includes detailed information such as the update date and time, the person who updated it (e.g., user ID or system identifier), the update type (e.g., new creation, update, deletion, etc.), the update target (e.g., product ID or field name), the value before the update, the value after the update, and the reason for the update (e.g., comments or update classification). This information is managed by the information management unit 32 and is used by the version management unit 40, the external synchronization unit 35, etc.

[0060] The information management unit 32 has an advanced history search function that allows it to efficiently search for past changes based on conditions such as time range, target product, updated items, and updater. The information management unit 32 also has a function for analyzing the scope of impact of changes, making it possible to track how changes to a product have affected other related products and processing flows. Furthermore, the information management unit 32 has a function for compressing and archiving history data. This automatically applies compression processing to old history data, optimizing storage usage.

[0061] The information management unit 32 functions as an audit trail, which preserves evidence of important change history from the perspective of compliance requirements and internal control. The information management unit 32 also functions as a data recovery platform, which allows data to be restored to a previous state in the event of an operational error or failure. These functions of the information management unit 32 ensure data transparency and reliability, and enable efficient differential synchronization and state management.

[0062] The information management unit 32 stores and manages the synchronization status of the external systems 4 and the synchronization processing results in the product DB 51. The synchronization processing results include processing results in which synchronization was successful and processing results in which synchronization was unsuccessful. The information management unit 32 also stores and manages information regarding the priority order when synchronizing multiple external systems 4 (hereinafter referred to as "priority order information") in the priority order DB 52. The information management unit 32 also stores and manages information regarding the specifications of the external systems 4 (hereinafter referred to as "specification information") in the specification DB 53. The information management unit 32 also stores and manages the update history of the product information in the product DB 51.

[0063] The information management unit 32 also stores and manages the content of updates and the date and time of updates detected by the update detection unit 33, which will be described later. The information management unit 32 systematically manages the change history of the product information, enabling tracking and auditing of changes. The stored information includes, for example, the date and time of the update, the person who updated (e.g., user ID or system identifier), the update target (e.g., product ID or field name), the value before the update, the value after the update, and the reason for the update (e.g., comments or update classification). This information is stored in chronological order in the product DB 51, so that the evolution of the product information can be kept track of.

[0064] For example, both differential and state recording methods are supported for recording update history. With the differential recording method, only changed fields are recorded, optimizing storage efficiency. With the state recording method, the complete state of product information at the time of each update is saved as a snapshot, making it easy to recreate the state at any point in time.

[0065] The information management unit 32 has a function for compressing historical data and an archiving function. This automatically migrates old historical data to compressed storage after a certain period of time has elapsed, enabling long-term storage while minimizing the impact on system performance. The information management unit 32 also has a function for searching update history and a filtering function. This allows for efficient extraction of history that matches specific conditions (e.g., time period, target product, updated items, etc.). Furthermore, the information management unit 32 has a function for visualizing update content, allowing intuitive understanding of changes in product information using time-series graphs and comparative displays. These functions of the information management unit 32 ensure thorough change management of product information, enabling efficient identification of causes when problems occur, audit responses, and compliance requirements. Furthermore, the information management unit 32, in cooperation with the version management unit 40, realizes advanced status management functions, such as rollback to previous versions.

[0066] Furthermore, the information management unit 32 detects the processing results (success, failure, etc.) of synchronization with each external system 4, and stores and manages them in the product DB 51. The information management unit 32 monitors the processing results of synchronization by the external synchronization unit 35 (described later) and systematically manages the status. In addition to the processing results, the information stored in the product DB 51 may include detailed information such as the synchronization execution date and time, the target external system, the data to be synchronized (e.g., product ID range, etc.), error code, error message, processing time, and the number of processed records. This information is maintained in the product DB 51 in a state where the history and current status can be referenced.

[0067] The processing result includes not only simple success or failure, but also detailed information such as partial success (for example, success for only some records) or success with a warning (for example, processing completed but with caution). The processing result may also include the analysis results of the response code and message from the external system 4. Such analysis results include, for example, the identification of the cause of failure (for example, authentication error, rate limit exceeded, data format mismatch, etc.).

[0068] The information management unit 32 has a function to detect abnormalities in synchronization processing and a function to notify the user of the abnormality. For example, if the information management unit 32 detects an abnormal state such as a synchronization failure or a long processing delay, it notifies the user of the abnormality (for example, by email, SMS, chat tool, etc.). The information management unit 32 also has a function to visualize the synchronization status through a management console. This makes it possible to get an overview of the synchronization status of all external systems 4 in dashboard format and to investigate the details of a specific synchronization process.

[0069] Furthermore, the information management unit 32 has a function for reporting the synchronization status. Specifically, the information management unit 32 periodically reports KPIs (Key Performance Indicators), such as the synchronization success rate and average processing time, enabling a qualitative evaluation of system operation. These functions of the information management unit 32 enable centralized management of the synchronization status with multiple external systems 4, enabling early detection and rapid response when a problem occurs. Furthermore, by coordinating with the external synchronization unit 35, which will be described later, automatic recovery of failed synchronization processing can be efficiently performed.

[0070] The information management unit 32 updates the standard-format product information based on the feedback acquired by the information acquisition unit 31. The information management unit 32 has a function of analyzing the information from the external system 4 acquired by the information acquisition unit 31 and reflecting the information in the standard-format product information stored in the product DB 51. This integrates changes and additional information in the external system 4, thereby realizing centralized management of information.

[0071] The information management unit 32 first compares and analyzes the acquired feedback with existing product information to identify the items to be updated and their contents. Updates can be broadly categorized into adding new information (e.g., customer reviews collected by the external system 4), updating existing information (e.g., changing prices, adjusting inventory levels), and invalidating or deleting information (e.g., setting sales to end). In determining whether to update, an evaluation of the reliability of the information plays an important role, and a comprehensive evaluation is made of factors such as the reliability of the external system 4 that serves as the information source, the appropriateness of the update frequency, and the consistency of the data.

[0072] The information management unit 32 has a function for customizing update policies. For example, detailed rules can be set for each item, such as "always overwrite," "reflect only newer information," or "only allow updates from a specific system." The information management unit 32 also has a function for recording changes in product information before and after updates. This allows a detailed change history to be accumulated. Furthermore, the information management unit 32 has a function for analyzing the scope of impact of updates. For example, the information management unit 32 identifies the impact that an update to a certain item will have on other related items (e.g., recalculating profit margins due to a price change) and performs linked updates as necessary.

[0073] The information management unit 32 has a function to optimize batch processing to efficiently execute large-scale update processing. This makes it possible to process large amounts of feedback at high speed. These functions of the information management unit 32 allow the various information collected from the external system 4 to be properly integrated and managed, and the latest and accurate product information to be maintained at all times.

[0074] The information management unit 32 also makes it possible to manage core data other than product information. For example, the information management unit 32 manages store information, customer information, supplier information, and the like in a standardized format. Store information includes store code, name, location, business hours, facility information, and contact information, and serves as the basis for managing sales channels such as physical stores and online stores. Customer information includes customer ID, basic attributes (name, contact information, etc.), purchase history information, membership rank, preference information, and the like, and serves as the basis for customer management and personalized services. Supplier information includes supplier code, company information, transaction terms, contact information, product categories handled, and the like, and is used for procurement management and product information source management.

[0075] This information is structured in a standardized format, just like the product information, and data conversion is applied as necessary when linking with each external system 4. The information management unit 32 has the function of referencing and managing master data, and manages the reference relationships between product information and related master data (for example, the association between products and stores that carry them). This maintains data consistency. The information management unit 32 also has the function of managing change histories, for example, managing the change history of each piece of master data. This makes it possible to track past status and ensure audit trails.

[0076] Furthermore, the information management unit 32 has a function for managing synchronization with the external system 4, and enables updates of each piece of master information to be linked to the related external system 4. These functions of the information management unit 32 enable systematic management of the relationships between entities involved in the distribution and sales activities of merchandise.

[0077] The information management unit 32 also stores various pieces of information in a data lake, which can store the acquired information in its original data format. A data lake is a large-scale storage platform for storing large amounts of diverse data, whether structured, semi-structured, or unstructured, in a format close to its original form. The data stored in the data lake is primarily raw data that has not undergone standardization or cleansing, and retains the format of the source system. This ensures that all data points are stored without any information loss and maintains a state that can flexibly respond to future analysis needs. The data stored in the data lake includes a wide range of data, such as detailed sales transactions from the POS system of the external system 4, customer behavior logs from the e-commerce site (e.g., browsing history, cart operations, etc.), inventory fluctuation history from the inventory management server 2, purchasing information from the purchasing management server 3, external market information (e.g., exchange rates, raw material prices, etc.), customer review data, and social media mentions.

[0078] Data lakes employ distributed storage technology for efficient management of large volumes of data and are designed for excellent horizontal scalability. Data lakes also have a data catalog function that manages metadata, such as the content, quality, and provenance of stored datasets, streamlining data exploration and analysis preparation. Data access control functions are also implemented, managing access privileges according to the data's sensitivity level, ensuring security and privacy. Data lakes serve as the foundation for advanced data analysis, abstracting and structuring raw data to extract business insights and build predictive models. Data lakes function as a "single source of truth" that aggregates all data assets within an organization and are an important information infrastructure supporting data-driven decision-making and strategy planning.

[0079] The update detection unit 33 detects that product information has been updated. Specifically, the update detection unit 33 detects that product information in the product DB 51 has been updated. The update detection unit 33 monitors changes to the product information stored in the product DB 51, and when an operation such as new addition, update, or deletion is performed, it identifies the change and triggers subsequent processing. Various techniques can be applied as a method for detecting updates, such as a method that utilizes a database trigger or log-based change data capture (CDC).

[0080] Possible methods for updating product information include manual updates via a user interface, automatic updates using data feeds from an external system 4, and bulk updates through periodic batch processing. The update detection unit 33 identifies the type of update (e.g., new addition, attribute change, deletion, etc.) and the target of the update (e.g., product name, price, inventory information, etc.), notifies the information management unit 32 of this information, and delegates processing to related functions such as the conversion unit 39 and external synchronization unit 35. Furthermore, the update detection unit 33 makes it possible to dynamically adjust the processing priority depending on the frequency and amount of updates, thereby optimizing the load balance of the entire system.

[0081] The external connection unit 34 connects the information processing system S to multiple external systems. Specifically, the external connection unit 34 establishes a communication channel between the main management server 1 and the external systems 4 (e.g., an e-commerce mall, a POS system, an inventory management system, etc.) to enable data integration. A variety of connection methods are available depending on the characteristics and interfaces of the external systems 4, such as direct integration via APIs, indirect integration via file transfers, and information acquisition via web scraping. API integration supports various protocols, such as REST, SOAP, and GraphQL, and enables authentication processing (e.g., OAuth, API keys, etc.), rate limiting, and error handling. Furthermore, file transfer integration supports protocols such as FTP, SFTP, and SCP, and automatically performs file format conversion and encryption / decryption.

[0082] The external connection unit 34 has a function of monitoring the connection status and automatically performs periodic communication checks, timeout detection, and reconnection processing, thereby maintaining a stable connection state. Furthermore, the external connection unit 34 has a function of detecting abnormal situations, such as changes in the API specifications of the external system 4 or the occurrence of a failure, and notifying the user of such situations. The external connection unit 34 individually manages connection settings for each external system 4 and enables connection processing based on connection parameters (e.g., endpoint URL, authentication information, timeout settings, etc.) stored in the specification DB 53. This allows for flexible response to the addition of a new external system 4 or changes to the settings of an existing system. The external connection unit 34 not only establishes a connection but also enables data transmission and reception processing, and realizes a series of data integration flows, such as sending product information and receiving feedback, in cooperation with the external synchronization unit 35 (described later).

[0083] The external synchronization unit 35 synchronizes the product information converted into the second format by providing it to the external system. The external synchronization unit 35 identifies the external systems 4 that require synchronization based on the changes to the product information. The external synchronization unit 35 selects and identifies the external systems 4 that require synchronization based on the changes to the product information. Rather than synchronizing all changes to all external systems 4, the external synchronization unit 35 analyzes the relevance of the changes to the external systems 4 and selects and synchronizes only the relevant systems. This reduces unnecessary synchronization processing and improves the efficiency of the entire system.

[0084] The external synchronization unit 35 first analyzes the attributes of the changed product information (e.g., category, price, inventory quantity, etc.) and evaluates whether the change is relevant to each external system 4. For example, the external synchronization unit 35 determines that a change in a product description for a specific e-commerce mall only needs to be synchronized with that e-commerce mall and is not relevant to other systems. The external synchronization unit 35 also determines that a change in price or inventory quantity is relevant to many external systems 4, but a change in the product's internal category classification is only relevant to the inventory management system.

[0085] The external synchronization unit 35 refers to the data requirements of each external system (e.g., which attributes are used) stored in the specification DB 53, and checks whether the changed attributes are used in that system. Furthermore, the external synchronization unit 35 considers the importance and urgency of the change and determines whether to synchronize important changes (e.g., price revisions, changes in sales status, etc.) with less relevant external systems 4.

[0086] The external synchronization unit 35 has the ability to customize synchronization decision rules, making it possible to adjust the rules according to business requirements. The external synchronization unit 35 also stores the results and reasons for synchronization decisions, making it possible to verify and audit them later. These functions of the external synchronization unit 35 enable efficient synchronization processing that is optimized for changes. As a result, unnecessary system load and network traffic are reduced, and a system can be established that reliably maintains the consistency of important information.

[0087] Furthermore, the external synchronization unit 35 sequentially executes synchronization based on priority when synchronizing multiple external systems 4. That is, the external synchronization unit 35 has a function of efficiently utilizing limited system resources and prioritizing synchronization processing of high importance. In determining the priority, priority information for each external system 4 stored in the priority DB 52 is referenced. The priority information includes information such as the importance of the external system 4 (e.g., main e-commerce site, auxiliary sales channel, etc.), real-time requirements (e.g., whether immediate reflection is required or regular updates are sufficient), and business impact (e.g., sales contribution, number of customers, etc.).

[0088] The external synchronization unit 35 makes a comprehensive decision based on the priority information, taking into account dynamic factors (e.g., the current processing queue status, changes in importance depending on the time period, etc.) in addition to the static information described above. For example, the external synchronization unit 35 can control the priority by time period, such as prioritizing synchronization with the POS system during business hours and synchronizing large amounts of data for batch processing outside of business hours. Priority control can be implemented using, for example, multilevel queuing. In this case, a processing queue is prepared for each priority level, and low-priority processing waits until the high-priority queue is empty. However, to prevent complete starvation of low-priority processing, an aging mechanism may be employed, in which the priority of processing that has been waiting for a certain period of time is gradually increased.

[0089] The external synchronization unit 35 adjusts resource allocation for synchronization processing based on priority information. This allows more CPU and memory resources to be allocated to processing of the external system 4 with a higher priority. The external synchronization unit 35 also has a function for dynamic throttling according to the load status of the entire system. This makes it possible to control the system so that when the entire system is under high load, only the most important synchronization processing is executed, and other processing is postponed until the load decreases. These functions of the external synchronization unit 35 enable efficient synchronization processing according to business importance even with limited system resources, ensuring the freshness and consistency of information on important channels.

[0090] Furthermore, the external synchronization unit 35 synchronizes by providing only differential updates to the external system based on the product information update history. That is, the external synchronization unit 35 has the function of improving the efficiency of the synchronization process by identifying and transmitting only data that has changed since the previous synchronization, rather than resynchronizing all data. This reduces communication volume, processing time, and the load on the external system.

[0091] When detecting differences, the external synchronization unit 35 works in conjunction with the information management unit 32 and refers to the change history since the previous synchronization. The external synchronization unit 35 applies the appropriate process for each type of change (e.g., new addition, update, deletion). For example, a CREATE operation is associated with a new addition, an UPDATE operation is associated with an update, and a DELETE operation is associated with a deletion. In the case of an update, a partial update is also performed that targets only the changed fields, minimizing unnecessary data transfer.

[0092] The external synchronization unit 35 has a function of updating product information in a chain reaction, taking into account the relevance of the product information. For example, when a product category is changed, it is possible to detect related information for all products belonging to that category as targets for update. The external synchronization unit 35 also has an adaption function according to the difference detection capability of each external system 4. This allows, for example, differential updates at the API level to be applied to external systems 4 that natively support differential updates, and differential calculation and transmission at the application level to be applied to external systems 4 that do not.

[0093] The external synchronization unit 35 also has a function to optimize batch processing of differential data. This enables efficient batch processing by queuing small differences until a certain amount is collected. The external synchronization unit 35 stores statistical information on the results of difference detection and the amount of data sent. The stored information can be used, for example, to evaluate and optimize synchronization efficiency. These functions of the external synchronization unit 35 enable efficient synchronization processing even in an environment that handles large amounts of master data, making it possible to achieve both real-time performance and processing efficiency.

[0094] The external synchronization unit 35 performs synchronization processing as follows. That is, the external synchronization unit 35 synchronizes by providing each external system 4 with product information that has been converted from a standard format to an external format by a conversion unit 39 (described later). The external synchronization unit 35 has a function of delivering product information that has been optimized for the external system 4 by the conversion unit 39 (described later) to each external system 4 via the external connection unit 34.

[0095] The synchronization process by the external synchronization unit 35 supports both immediate synchronization triggered by an update of product information and batch synchronization based on a regular schedule. In immediate synchronization, related external systems 4 are identified based on changes detected by the update detection unit 33, and synchronization is selectively performed only for the identified external systems 4. On the other hand, in batch synchronization, batch synchronization of all target product information is performed based on a specified schedule (e.g., late night every day, every Monday, etc.). For efficiency reasons, a differential synchronization mechanism is adopted in the synchronization process, and only product information that has changed since the previous synchronization is synchronized. This difference detection utilizes the change history recorded in the information management unit 32.

[0096] The external synchronization unit 35 also enables priority control according to the load status and importance of each external system 4, and adjusts the synchronization order and the number of simultaneous executions based on the priority information stored in the priority DB 52. The external synchronization unit 35 has a function to monitor the results of synchronization processing, tracks the synchronization results (success, failure, etc.), and records the processing results.

[0097] If synchronization fails, the external synchronization unit 35 applies retry logic according to the error content and automatically recovers from failures such as temporary communication failures. If synchronization fails for a long period of time, the external synchronization unit 35 also enables flexible countermeasures, such as attempting synchronization via an alternative route (for example, switching from API to file transfer). These processes by the external synchronization unit 35 maintain the consistency of product information between multiple external systems 4.

[0098] The external synchronization unit 35 also attempts to resynchronize with the external system 4 with which synchronization failed. That is, the external synchronization unit 35 has the function of automatically executing recovery processing if the synchronization process fails. The external synchronization unit 35 references information about the synchronization failure stored in the product DB 51 and identifies the synchronization process that should be retried. When retrying, appropriate countermeasures are taken depending on the cause of the failure. For example, if the failure is due to a temporary network failure, a simple retransmission is performed, but if the failure is due to a data format mismatch, a retransmission is performed after adjusting the conversion rules.

[0099] The timing and frequency of synchronization retry attempts are dynamically adjusted depending on the type and severity of the failure. Minor transient errors are retried multiple times at short intervals, while serious errors are retried only after user confirmation. The external synchronization unit 35 can implement an exponential backoff algorithm, which gradually lengthens the retry interval in the event of successive failures. This prevents the external system 4 from becoming overloaded.

[0100] The external synchronization unit 35 also has a function for automatically repairing specific error patterns. This enables, for example, automatic token updating for failures due to authentication token expiration, and automatic correction processing for minor data format errors. Furthermore, the external synchronization unit 35 has a function for attempting an alternative synchronization route (e.g., switching from API to file transfer) if synchronization is unsuccessful after a certain number of retries. The external synchronization unit 35 stores detailed records of retry attempts and their results. The stored information can be used to evaluate the effectiveness of retry patterns and analyze the root cause of problems. These functions of the external synchronization unit 35 automatically recover from synchronization failures due to temporary failures or minor errors, thereby maintaining a high synchronization success rate while minimizing manual user intervention.

[0101] The display control unit 36 ​​controls the display of a setting screen on the user terminal 5 for setting rules for converting a standard format into an external format (hereinafter referred to as "conversion rules"). Specifically, the display control unit 36 ​​provides a user interface for the user to define the details of the conversion process. The setting screen is implemented as an intuitive GUI (Graphical User Interface) that runs on a web browser and is designed to be easily operable even without specialized programming knowledge.

[0102] The settings screen can display various setting items, such as conversion rule definitions for each external system, item mapping settings, data format conversion rules, validation rules, and generation AI utilization parameters. Of these, item mapping settings allow users to visually define which fields in the standard format correspond to which fields in the external system, and allows users to set value conversion and merge / split rules as needed. Data format conversion rules also allow users to define standardization rules for data representation, such as date format, numeric format, and unit conversion. Generation AI utilization parameters also allow users to set adjustment items for AI processing, such as the generation and tagging of appeal statements (e.g., writing style, features to emphasize, prohibited expressions, etc.). Specific examples of the settings screen will be described later with reference to Figure 5.

[0103] The display control unit 36 ​​has a setting content preview function and enables the display of a screen that simulates the conversion results using actual product information. This allows the validity of the conversion rules to be confirmed before applying them to the production environment. The display control unit 36 ​​also enables the display of a screen that manages the version of the setting content. This makes it possible to restore past setting states and track the history of setting changes. Furthermore, it enables the display of a screen that replicates settings between multiple environments (e.g., development, verification, production, etc.). This makes it possible to apply verified settings to the production environment all at once. These processes by the display control unit 36 ​​allow the user to visually and systematically manage the complex conversion requirements for each external system 4.

[0104] The prompt generation unit 37 generates a prompt to be given to the generation AI based on the set conversion rules. The prompt generation unit 37 automatically generates specific and clear prompts to effectively utilize the generation AI. When generating a prompt, the prompt generation unit 37 refers to the conversion rules and processing requirements defined via the display control unit 36 ​​and performs processing to translate them into a format that is easy for the generation AI to understand.

[0105] The generated prompts comprehensively include information necessary for task execution, such as the type of task (e.g., generating a promotional statement, correcting spelling variations, etc.), a description of the input data, the expected output format, constraints (e.g., character limit, prohibited expressions, etc.), processing priority, etc. Specifically, for example, a prompt for generating a promotional statement will specify basic information about the product, as well as the characteristics of the target customer base, points of differentiation from competing products, and the company's branding policy.

[0106] The prompt generation unit 37 has a function to optimize prompts taking into account the characteristics of the generation AI, and adopts structures and expressions that the generation AI can process efficiently. For example, when complex processing is required, the prompt generation unit 37 employs techniques such as breaking down instructions into stages, emphasizing important points, and presenting specific examples. The prompt generation unit 37 also has a mechanism for improving prompts by reflecting feedback from past processing results, enabling the generation AI to continuously improve its response quality for similar tasks.

[0107] The prompt generation unit 37 has a so-called multi-model support function that can support multiple generation AI models, making it possible to optimize the format and expression of the prompt according to the characteristics of the AI ​​model being used. Furthermore, the prompt generation unit 37 has a function for managing prompt versions, and by saving or reusing effective prompt patterns, it makes it possible to systematically accumulate know-how in prompt engineering.

[0108] The abstraction unit 38 reduces the amount of data related to the product information by abstracting the product information. For example, the abstraction unit 38 abstracts the product information by matching organization names (e.g., company names) included in the product information. The abstraction unit 38 has a function of summarizing and aggregating detailed information in order to efficiently process large amounts of product information. When abstracting the product information, the abstraction unit 38 uses an approach that omits redundant information and detailed variations while maintaining the essential characteristics and trends of the data.

[0109] The abstraction unit 38 abstracts product information using various dimension reduction techniques, such as range classification of numerical data (e.g., price range categorization), summarization of text data, topic extraction, aggregation of time-series data (e.g., daily to monthly), and categorization of similar products. The abstraction unit 38 utilizes generative AI to semantically abstract product descriptions, extracting only essential functions and features from detailed descriptions and converting them into concise expressions. Furthermore, in the process of merging organizational names, the abstraction unit 38 standardizes names due to spelling variations and organizational changes, enabling, for example, standardization so that the same company can be consistently identified. This name merging process applies a highly accurate matching algorithm that utilizes not only simple string comparison but also auxiliary information such as addresses and telephone numbers.

[0110] The abstraction unit 38 has the ability to adjust the level of abstraction, allowing for flexible changes in the information compression rate depending on the purpose of use and the granularity of analysis. For example, it allows for the application of high-level abstraction for grasping overall trends and low-level abstraction for detailed analysis. The abstraction unit 38 also has the ability to evaluate the importance of information lost through abstraction, and performs the abstraction process while ensuring that important features and differentiating elements are maintained. These processes by the abstraction unit 38 enable efficient processing of large amounts of data that exceed the input limits of the generation AI and facilitate the understanding of trends in vast amounts of product information, thereby realizing strategic product management and analysis.

[0111] When an update to the product information stored in the standard format is detected, the conversion unit 39 converts the product information from the standard format to an external format. The conversion unit 39 also converts product search and analysis instructions written in natural language into command statements compatible with the information processing system S using a generation AI. Specifically, the conversion unit 39 performs a conversion process to adapt the product information managed in the standard format to a format required by each external system 4 (e.g., an e-commerce mall, a POS system, etc.). The elements to be converted are diverse, including not only data structure (e.g., item names, item order, etc.), but also text format (e.g., character limit, prohibited characters, etc.) and product image format (e.g., size, aspect ratio, resolution, etc.).

[0112] In the conversion process performed by the conversion unit 39, the specification information of each external system 4 stored in the information management unit 32 is referenced, and conversion rules that conform to each constraint condition are applied. Furthermore, by utilizing generation AI, not only simple format conversion but also advanced conversion (e.g., correction of spelling variations, summarization, style adjustment, etc.) based on an interpretation of the meaning of the product information is realized. Furthermore, the conversion unit 39 has the function of converting search and analysis instructions written in natural language into formats such as structured query language (SQL) and API calls. This allows users to perform advanced searches and analyses without specialized knowledge.

[0113] The conversion unit 39 has a tag generation unit 391 as tag generation means, an appeal statement generation unit 392 as appeal statement generation means, a spelling variation correction unit 393 as spelling variation correction means, and an image conversion unit 394 as image conversion means. The conversion unit 39 also has a formatting unit 395 as formatting means, a summarization unit 396 as summarization means, a name matching unit 397 as name matching means, and a style matching unit 398 as style matching means. The conversion unit 39 also has a categorization unit 399 as categorization means, an appeal generation unit 400 as appeal generation means, and a preprocessing unit 401 as preprocessing means.

[0114] The tag generation unit 391 uses a generation AI to generate tags to be attached to commercial materials. The tag generation unit 391 provides a function to comprehensively analyze information such as the name, description, and image of the commercial material, and automatically generate tags to appropriately classify the commercial material. Tags are generated from various perspectives, such as the characteristics of the commercial material (e.g., color, size, material, etc.), category (e.g., clothing, furniture, food, etc.), use (e.g., indoor use, commercial use, etc.), and target (e.g., for men, for children, etc.).

[0115] In generating tags, the tag generation unit 391 utilizes the natural language understanding capabilities of the generation AI to extract key features from product descriptions and structure them as appropriate tags. Furthermore, the tag generation unit 391 learns existing tagging patterns for similar products to generate consistent tags. The generated tags are not only stored in the product DB 51, but are also optimized and transmitted according to the requirements of each external system 4. For example, if the external system 4 is a search engine, keyword tags that take SEO (Search Engine Optimization) into consideration are generated. Furthermore, if the external system 4 is an e-commerce mall, tags that match the product category are generated. Furthermore, the tag generation unit 391 periodically evaluates the effectiveness of tags and automatically optimizes them based on indicators such as search hit rate and related product recommendation accuracy.

[0116] The appeal statement generation unit 392 uses a generation AI to generate an appeal statement for the product. The appeal statement generation unit 392 has a function of automatically creating an explanatory statement to effectively communicate the features and advantages of the product. When generating an appeal statement, the appeal statement generation unit 392 considers not only basic information about the product (e.g., functions, performance, materials, etc.), but also the expected target demographic, usage scenarios, and points of differentiation from competing products. By utilizing the generation AI, the appeal statement generation unit 392 enables the creation of appealing sentences that appeal to the user's emotions and desires, rather than simply listing product specifications.

[0117] The appeal statement generation unit 392 also generates appeal statements optimized for each channel by adopting a tone and expression style that matches the characteristics of each sales channel and the tendencies of buyers in that channel. For example, the appeal statement generation unit 392 can adjust the appeal statement, such as adopting sophisticated expressions for a high-end e-commerce site and casual, friendly expressions for a platform aimed at younger customers. Furthermore, the appeal statement generation unit 392 can emphasize appeal points according to the product category and price range, and update the expressions to match the season and trends. In the process of generating appeal statements, the appeal statement generation unit 392 can refer to the results of analysis of past sales records and customer reviews of similar products. This allows effective appeal points to be prioritized in the product's appeal statement.

[0118] The spelling variation correction unit 393 uses the generation AI to correct spelling variations in the product information. Examples of spelling variations include "blue" and "blue" and "notebook PC" and "notebook PC." The spelling variation correction unit 393 has a function to detect spelling variations in the product information and standardize them to a standard spelling. Since spelling variations can cause a decrease in search accuracy and errors in data analysis, correcting them plays an important role in improving the quality of product information.

[0119] When correcting spelling variations, the spelling variation correction unit 393 first performs a dictionary-based primary check to detect common synonyms and spelling variations. Next, the spelling variation correction unit 393 performs advanced correction processing using context understanding by generative AI. This allows appropriate correction of spelling variations that cannot be handled by dictionaries alone, such as industry-specific terminology and product-specific expressions. In addition, the spelling variation correction unit 393 aims to unify expressions not only at the word level but also at the phrase and sentence structure level. For example, the spelling variation correction unit 393 unifies functional expressions such as "waterproof" and "waterproof compatible," and standardizes the composition patterns of product descriptions.

[0120] The spelling variation correction unit 393 has a customization function to accommodate the spelling rules of each company or industry, and makes it possible to define the preferred spelling format through a setting screen. The product information corrected by the spelling variation correction unit 393 is stored in the product DB 51 along with the correction history, making it possible to perform post-verification and auditing.

[0121] The image conversion unit 394 uses a generation AI to convert the format of the product image so that it satisfies the constraints of the external system 4. In other words, the image conversion unit 394 has the function of optimizing image data related to merchandise to meet the requirements of each external system 4 (e.g., an e-commerce mall, SNS, advertising platform, etc.). The format of the product image to be converted includes various elements such as resolution, file size, aspect ratio, color format, and file format. Unlike conventional simple image resizing, the generation AI is used to automatically recognize the main parts of the merchandise, and adjustments are made to maximize the appeal of the merchandise even when cropping or enlarging it.

[0122] For example, when resizing a product image, the image conversion unit 394 detects the product's edges to identify the main features and optimizes the cropping position so that the main features are centered. The image conversion unit 394 also improves visibility on each platform by automatically removing backgrounds, correcting color tones, and adjusting contrast. Some e-commerce platforms impose restrictions on the amount and placement of text within an image. Therefore, the image conversion unit 394 detects text elements within the image to comply with these restrictions and adjusts or deletes them as necessary. During the conversion process, the image conversion unit 394 references the image requirements of each external system 4 stored in the information management unit 32 and automatically generates multiple versions of images optimized for each system. The image conversion unit 394 stores the converted product images in the product DB 51. The stored converted product images are used when synchronizing with each external system 4.

[0123] The formatting unit 395 uses a generation AI to format descriptions of product information so that they meet the specifications of the external system 4. The formatting unit 395 uses a generation AI to format descriptions of product information so that they meet the specifications of the external system 4. The formatting unit 395 has a function to optimize text data such as product descriptions and specification information to meet the requirements of each external system 4 (e.g., character limit, prohibited expressions, required fields, etc.). When formatting descriptions of product information, the formatting unit 395 not only performs simple character reduction and word replacement, but also performs advanced editing that takes into account the meaning and importance of the text by utilizing the language understanding capabilities of the generation AI. For example, if there is a character limit, the formatting unit 395 determines what content should be prioritized based on the importance of the information, and deletes redundant expressions and rephrases them concisely.

[0124] The formatting unit 395 also detects expressions prohibited by a particular platform (e.g., excessive exaggeration, claims of medicinal efficacy, etc.) and replaces them with acceptable expressions. The formatting unit 395 also optimizes the appropriate placement of keywords and paragraph structure, taking into account the search algorithms and user experience of each platform. The formatting unit 395 has the function of comparing text before and after formatting to verify that semantic consistency is maintained, and finalizes the conversion results after confirming that no excessive information loss or semantic change has occurred. These processes by the formatting unit 395 are performed based on the requirements definitions of each external system 4 stored in the information management unit 32, and the formatted text data is stored in the product DB 51.

[0125] The summarization unit 396 uses the generation AI to summarize the product information in accordance with the data volume limit of the external system 4. The summarization unit 396 has the function of extracting important points from detailed product information and summarizing it concisely according to a specified number of characters and format. In processing the summary, the summarization unit 396 utilizes the natural language understanding capabilities of the generation AI to appropriately reduce redundant expressions and secondary information while maintaining the essential value and characteristics of the product.

[0126] The level of summarization by the summarization unit 396 is adjusted according to the limitations of each external system 4, allowing summarization of various levels of granularity, from detailed specification-level information to a short description of a few lines. For example, a short summary that succinctly conveys the main features of a product is generated for a product list page on an e-commerce mall, while a medium-level summary including more detailed information is generated for a product detail page.

[0127] Furthermore, when summarizing, the summarization unit 396 dynamically adjusts the criteria for determining importance depending on the product category and target user demographic. For example, in the case of technical products, performance and functionality are prioritized, while in the case of fashion products, design and materials are prioritized. Furthermore, to improve the readability and naturalness of the summary results, the summarization unit 396 enables rephrasing to expressions that aggregate multiple pieces of information or converting to a bulleted list format, rather than simply deleting sentences. The summarization unit 396 has a function for verifying the consistency of information before and after summarization, and finalizes the results after confirming that no important information is missing or incorrect information is mixed in.

[0128] The name matching unit 397 uses generation AI to match and align product information acquired from multiple data sources. The name matching unit 397 integrates product information collected from different systems and information sources, making it possible to eliminate duplication and inconsistencies. In the name matching process, the name matching unit 397 first performs a matching process to identify similar product records. Specifically, in addition to primary matching using identifiers such as product codes, JAN codes, and product names, the name matching unit 397 performs advanced matching based on similarity evaluation results that take into account product features and attribute information. Furthermore, by utilizing generation AI, the name matching unit 397 is able to absorb variations in spelling and differences in item configuration, making it possible to accurately match products that are deemed semantically identical.

[0129] The name matching unit 397 performs a merge process to reconcile matched records, selecting the most appropriate data based on the relationship between old and new information and reliability assessments. This process involves sophisticated processing, not simply overwriting, but selectively adopting the most reliable value for each attribute and combining and integrating information from multiple sources. The name matching unit 397 also performs business rule-based verification to verify the consistency of the reconciliation results, such as checking the relationship between price and cost and checking the consistency between specifications. The name matching unit 397 has the function of recording the history and judgment basis of the name matching and reconciliation processes, allowing for reference in the event of post-verification or manual adjustment. The results of the reconciliation process performed by the name matching unit 397 are reflected in the product DB 51 and centrally managed as standardized product information.

[0130] The style matching unit 398 uses a generation AI to generate product information that matches the expression style of the person providing the product (e.g., a company) or the industry. The style matching unit 398 has a function of adjusting text data such as product descriptions and appealing statements to a style and expression that matches the company's brand image and industry conventions. In the process of generating product information that matches the expression style of the person providing the product or the industry, the style matching unit 398 extracts characteristic expression patterns from, for example, the target company's existing marketing materials and product descriptions, and builds a text generation model that conforms to them. During this process, the style matching unit 398 utilizes the language modeling capabilities of the generation AI to analyze detailed style characteristics such as vocabulary selection, writing style, modifiers, and the use of technical terms.

[0131] The style matching unit 398 converts input product information based on the constructed model to match the style of the target company or industry. For example, a sophisticated and elegant expression is applied to luxury brands, while a casual and friendly expression is applied to brands aimed at young people. The style matching unit 398 also takes into account the expression patterns and terminology specific to specific industries (e.g., cosmetics, home appliances, food, etc.). This allows for natural expressions that conform to industry practices.

[0132] The style matching unit 398 has the function of managing style models for multiple brands and industries, making it possible to selectively apply an appropriate style to each product. Furthermore, the style matching unit 398 can set constraints to comply with a company's brand guidelines and prohibited expressions, allowing for expression adjustments while maintaining the company's brand policy. The results of processing by the style matching unit 398 are stored in the product DB 51 and are used when synchronizing with each external system 4.

[0133] The categorization unit 399 uses the generation AI to categorize merchandise. The categorization unit 399 has a function for systematically managing groups of merchandise with variations by categorizing them through association. In the categorization process, the categorization unit 399 first analyzes merchandise information to identify products in the same series and variations. In this identification, the generation AI's image recognition and natural language understanding capabilities are utilized to identify variation relationships based not only on explicit information such as product names and codes, but also on similarities in appearance and function.

[0134] The categorization unit 399 determines the type of variation, such as color, size, material, or function, for the identified group of variation products and associates appropriate tags. For example, shirts with the same design are associated with color tags such as "blue," "red," and "white" and size tags such as "S," "M," and "L." The categorization unit 399 manages information related to these tags in a hierarchical structure. This enables multi-layered classification, such as product family > model > color > size.

[0135] The categorization unit 399 displays the categorized product groups as a list of related products when displayed on an e-commerce site, and uses them to switch between variations when searching for products. This improves the user experience. In addition, the categorized information can be used for inventory management and sales analysis, enabling trend analysis at a granular level, such as "blue medium size is a best-seller." The categorization unit 399 has a function to associate new product variations with existing categorizations when they are added, and a function to periodically verify the consistency of the categorization. This allows for consistent categorization management to be maintained even when product information changes.

[0136] The appeal generation unit 400 generates text to appeal products similar to best-selling products. The appeal generation unit 400 has a function of analyzing the appeal points of proven products and generating appeal statements that can be applied to other products with similar characteristics. In the process of generating appeal statements, the appeal generation unit 400 first comprehensively analyzes the appeal statements, customer reviews, sales data, etc. of the best-selling products, and extracts the reasons why the products are popular and their effective appeal points.

[0137] The appeal generation unit 400 utilizes the natural language understanding capabilities of the generation AI to analyze the appeal points of proven products. This allows consideration of not only the frequency of words and expressions, but also the context and modifying expressions that elicit customer responses. The appeal patterns extracted by the appeal generation unit 400 are then compared with the characteristics of the new product to be targeted, and applicable elements are selected.

[0138] During this process, the appeal generation unit 400 closely examines the similarities (for example, the same category, similar functions, common target demographic, etc.) and differences between the products. This allows for appropriate customization to match the characteristics of the new product, rather than simply copying. The generated appeal text follows the appeal techniques of best-selling products, while also appropriately expressing the uniqueness and advantages of the new product.

[0139] If a product is popular because it emphasizes a particular material, the appeal generation unit 400 can generate appeal statements that highlight the characteristics of other products that use the same material. The appeal generation unit 400 has a function for evaluating the effectiveness of the generated appeal statements. This makes it possible to analyze sales performance and customer reactions after the appeal statements are generated. As a result, continuous improvement is achieved. The appeal generation unit 400 also has a function for flexibly adjusting appeal points in accordance with changes in seasons and trends. This allows the appeal generation unit 400 to always maintain the latest appeal effects.

[0140] The preprocessing unit 401 uses the generation AI to generate a program for preprocessing data to be input to the generation AI. The preprocessing unit 401 has a function of generating a program that automates data formatting and conversion in order to efficiently process large amounts of complex product-related data using the generation AI. In the process of generating a preprocessing program, the preprocessing unit 401 first analyzes the structure and characteristics of the target data and identifies issues to be processed (for example, noise removal, structure conversion, normalization, etc.).

[0141] Next, the preprocessing unit 401 uses the code generation capabilities of the generative AI to generate program code that addresses these challenges. The generated program code is an executable script that includes a series of steps from data reading, conversion processing, and output, and is provided in a format suitable for batch processing or periodic execution. The preprocessing includes various data cleansing processes, such as removing unnecessary spaces and special characters, standardizing dates and numbers, converting category codes, filling in missing values, and processing outliers.

[0142] Furthermore, the preprocessing unit 401 performs advanced preprocessing such as natural language processing, such as text segmentation, morphological analysis, entity extraction, and sentiment analysis, as well as image processing, such as image noise removal, normalization, and feature extraction. The preprocessing unit 401 has a function for verifying the operation of the generated program, evaluates the execution results using small-scale sample data, and finalizes the program after confirming that the intended processing is performed correctly.

[0143] The preprocessing unit 401 also has the ability to update programs in response to changes in data structure and processing requirements, enabling the system as a whole to flexibly respond to changes. These preprocessing programs generated by the preprocessing unit 401 are used to convert the large amounts of unstructured and semi-structured data stored in the data lake into a format that is easy to handle with generative AI and analysis tools. This significantly improves the data processing efficiency of the entire system.

[0144] The version management unit 40 manages versions of the product information. For example, the version management unit 40 rolls back versions of the product information based on the stored update history. That is, the version management unit 40 has the function of utilizing the change history stored by the information management unit 32 to return the state of the product information to any point in the past. This makes it possible, for example, to cancel an erroneous update or reproduce the state of a specific point in time. In version management by the version management unit 40, a unique version identifier (for example, a timestamp or sequence number) is assigned to each update, and transition to a specific version is achieved by specifying it.

[0145] The rollback process by the version management unit 40 not only allows for a simple rewind of the state, but also allows for the specification of the rollback scope (for example, all product information or only specific products, etc.) and target items (for example, all fields or only specific fields, etc.). In addition, the rollback operation itself is recorded in the update history as one of the updates, making it possible to cancel the rollback operation and perform audit tracking.

[0146] The version management unit 40 has a function to display the differences between versions, allowing for visual comparison of changes made between any two points in time. This provides information for making decisions about version selection and makes it easier to identify appropriate rollback points. The version management unit 40 also has version branching and merging functions, making it possible to manage multiple change paths starting from a specific version in parallel and to integrate different change systems.

[0147] The version management unit 40 also has the function of adding labels and comments to versions, making it possible to assign meaningful names to important version points (for example, the time of release, the start of a campaign, etc.) and manage them. These advanced version management functions of the version management unit 40 make it possible to systematically manage the change history of complex product information, ensuring the consistency of information and enabling quick responses when problems occur.

[0148] The query generation unit 41 causes the generation AI to generate and divide queries based on the content of a question created by a user. Specifically, the query generation unit 41 generates a query using the generation AI and divides the query according to the content of the question. The query generation unit 41 has a function of converting search and analysis requests expressed in natural language into a format that can be executed by a database or data analysis engine (e.g., SQL, NoSQL query, analysis API parameters, etc.).

[0149] In the process of generating a query, the query generation unit 41 first analyzes and structures the natural language question. This extracts elements such as intent understanding (e.g., search or analysis), target entity (e.g., what product information is being targeted), conditions (e.g., what constraints exist), and desired results (e.g., what format of output is desired). In this process, the query generation unit 41 utilizes the language understanding capabilities of the generation AI to appropriately interpret ambiguous expressions and complex conditions.

[0150] The query generation unit 41 converts the extracted elements into an appropriate query language or API format. However, there are cases where a complex question cannot be answered with a single query. In such cases, the query generation unit 41 logically divides the question and generates multiple subqueries. For example, a question such as "What are the sales and inventory status of blue T-shirts?" is divided into a sales query and an inventory query, and the results of each are integrated to form the final answer.

[0151] The query generation unit 41 has a function to verify the validity of the generated query, and confirms that there are no performance problems (e.g., overly complex joins, inefficient filters, etc.) or security problems (e.g., attempts to access outside the authorized scope, etc.). In addition, the query generation unit 41 has a function to ask confirmation questions before generating a query if the user's question intention is unclear. Furthermore, the query generation unit 41 has a function to learn frequently occurring query patterns, enabling efficient generation of queries for similar questions. With these functions, the query generation unit 41 realizes an environment in which even users with no knowledge of databases or query languages ​​can perform advanced data searches and analyses through natural dialogue.

[0152] The consistency verification unit 42 verifies whether the attributes of the product information conform to the constraints of the external systems 4 based on the specification information. The consistency verification unit 42 has a function of checking in advance whether the product information to be sent satisfies the technical or business constraints of each external system 4 before synchronizing with the external systems 4. This makes it possible to prevent synchronization errors and inappropriate data display. The constraints to be verified for consistency include both technical constraints (e.g., field length, data type, required items, etc.) and business constraints (e.g., price range, validity of inventory quantity, consistency of category system, etc.).

[0153] During the verification process, the consistency verification unit 42 references the constraint definitions for each system stored in the information management unit 32 and checks each attribute value of the data to be sent to see if it satisfies those conditions. For example, if a particular e-commerce mall requires product names to be 50 characters or less, any product name exceeding that number will be detected as a verification error. Business-related validity checks are also performed, such as checking whether inventory quantities are negative or prices are abnormally high.

[0154] The consistency verification unit 42 also has a function for classifying the severity of errors. This enables classification into levels such as "critical" (a fatal problem that makes synchronization impossible), "warning" (a problem that requires correction), and "information" (a minor problem that is provided as reference information). The consistency verification unit 42 also has a function for reporting the verification results, and provides the user with a list of detected problems and statistical information. Furthermore, by cooperating with the attribute correction unit 43 (described later), the consistency verification unit 42 can delegate problems that can be automatically corrected to correction processing.

[0155] The consistency verification unit 42 has the ability to customize verification rules, which allows the rules to be adjusted in response to changes in business requirements or changes in the specifications of the external system 4. These functions of the consistency verification unit 42 systematically detect potential problems before synchronization with the external system 4, significantly improving data quality and the reliability of the synchronization process.

[0156] The attribute correction unit 43 corrects attributes of product information that do not conform to the constraints of the external system 4. The attribute correction unit 43 has a function to process problems detected by the consistency verification unit 42 described above that can be automatically corrected based on certain rules. This allows many consistency problems to be resolved without manual user intervention, making the synchronization process more efficient and improving the success rate.

[0157] The attribute correction unit 43 corrects problems for which correction methods can be defined by clear rules, such as excessive character count, format mismatch, and out-of-range numeric values. For example, product names that exceed character limits are automatically shortened while preserving important information, and incompatible date formats are converted to standard formats. Numeric values ​​that exceed the allowable range are also automatically rounded to the maximum or minimum value.

[0158] The attribute correction unit 43 has an advanced correction function that utilizes generative AI. This enables processes such as reducing the number of characters while preserving the meaning of the text content and replacing prohibited expressions with appropriate alternatives. The level of correction processing can be adjusted by settings, allowing flexible settings according to operational requirements, such as fully automatic mode (e.g., automatically processing all correctable issues), semi-automatic mode (e.g., requiring approval for important corrections), and advice mode (e.g., displaying only correction suggestions).

[0159] The attribute correction unit 43 has a function for recording correction history. This allows for detailed storage of which attributes have been corrected and how. As a result, transparency of the correction process is ensured, and it is possible to restore the state before correction if necessary. The attribute correction unit 43 also has a function for learning correction patterns. This allows for the accumulation of frequently occurring problem patterns and effective correction methods, thereby continuously improving correction accuracy. These functions of the attribute correction unit 43 automatically resolve many consistency problems in synchronization with the external system 4, for example, reducing the burden on the user and significantly improving data quality and the reliability of the synchronization process.

[0160] The transmission control unit 44 controls the communication unit 19 (see FIG. 2) to transmit various types of information to the inventory management server 2, the purchase management server 3, and the external system 4, respectively.

[0161] The product DB 51 allows product information to be stored in a standard format. The product DB 51 functions as a central database that serves as the basis for all product information and is a core storage area for unified information management. Product information managed in a standard format includes a wide range of data, such as basic information (e.g., product ID, name, description, etc.), classification information (e.g., category, tag, etc.), specification information (e.g., size, weight, material, etc.), price information (e.g., list price, selling price, cost price, etc.), inventory information (e.g., inventory quantity, expected arrival date, etc.), related information (e.g., similar products, set products, etc.), media information (e.g., images, videos, etc.), and status information (e.g., sales status, display status, etc.). This information is structured based on a clearly defined schema, and the meaning and constraints of each attribute are managed uniformly.

[0162] The product DB 51 has advanced database functions for efficiently managing large amounts of product information, and achieves high-speed data access through index optimization, partitioning, caching mechanisms, etc. The product DB 51 also has a function for managing history, and in cooperation with the information management unit 32, it stores the change history of product information, making it possible to reproduce the state at any point in time.

[0163] Furthermore, product DB 51 has a function for managing data quality, and has mechanisms for maintaining data accuracy and consistency, such as validating input values, ensuring referential integrity, detecting duplicate data, etc. These functions of product DB 51 enable product DB 51 to function as a "single source of truth" for highly reliable product information, and enable the stable provision of standard data that will be distributed to external system 4.

[0164] The product DB 51 allows product information to be stored in an external format. The product DB 51 also allows product information in a standard format to be converted to meet the requirements of each external system and then stored. This allows data sets in a format optimized for each external system 4 (e.g., an e-commerce mall, a POS system, etc.) to be managed, improving the efficiency and reliability of synchronization processing.

[0165] Various conversion processes are applied to product information in external formats, such as converting field names and structures (for example, merging or splitting items), adjusting text formats (for example, complying with character limits, replacing prohibited characters, etc.), optimizing product image formats (for example, resizing, format conversion, etc.), and category mapping (for example, associating standard categories with external system categories). Storing this converted data in the product DB 51 in advance reduces the conversion load during synchronization processing, enabling faster delivery.

[0166] The product DB 51, together with the information management unit 32, has a function for managing conversion versions. This makes it possible to track which external format data has been regenerated when standard format data is updated. Furthermore, previous versions of external format data are retained for a certain period of time. Furthermore, the product DB 51, together with the information management unit 32, has a function for verifying the conversion results, and automatically checks whether the generated external format product information satisfies the constraints of each external system 4.

[0167] Additionally, the product DB 51, together with the information management unit 32, has a function for analyzing the conversion efficiency for each external system 4, and is used to optimize the conversion process and identify conversion patterns that frequently cause problems. With these functions of the product DB 51, the product DB 51 functions as an intermediate buffer that supports fast and reliable data delivery to each external system, and plays a role in significantly improving the stability and efficiency of the synchronization process.

[0168] The priority DB 52 stores priority information for each external system 4. The priority DB 52 is a storage area for managing reference information for determining the order of priority when executing synchronization processing with multiple external systems 4. The priority information includes a variety of evaluation elements, such as the importance rank of the external system (e.g., highest, high, medium, low, etc.), real-time requirements (e.g., immediate, near real-time, batch, etc.), time-of-day priority (e.g., high priority during business hours), and business impact indicators (e.g., sales contribution, number of customers, etc.).

[0169] The priority information stored in the priority DB 52 is referenced by the external synchronization unit 35 and is used to determine the execution order when the number of synchronization processes that can be executed simultaneously is limited. The priority DB 52 stores the history of temporary priority adjustments (for example, in response to a campaign at a specific time). The priority DB 52 also has a function to link with a management screen for setting priorities. This allows the user to adjust priorities using an intuitive interface.

[0170] Furthermore, the priority DB 52 has a function for analyzing execution statistics. This makes it possible to analyze the relationship between the actual processing order based on the set priorities and the processing results, which is useful for optimizing priority setting. The priority DB 52 has a function for managing priority versioning. This allows a change history of priority setting to be maintained, making it possible to revert to previous settings as necessary. These functions of the priority DB 52 enable the most effective synchronous processing order within limited system resources and flexible processing control according to business requirements.

[0171] The product DB 51 also stores the synchronization status with each external system 4. The product DB 51 is a storage area for managing the history and current status of synchronization processing performed between the main management server 1 and the external systems 4. The recorded information includes detailed data such as the synchronization execution date and time, target external system, synchronization direction (e.g., send / receive), target data range, processing result (e.g., success / failure / partial success), error information (e.g., error code, message, etc.), and processing statistics (e.g., number of processed records, processing time, etc.). This information is recorded and updated by the external synchronization unit 35 and is used to monitor the synchronization status and diagnose problems.

[0172] The product DB 51, together with the information management unit 32, has a function for managing synchronization history in a timeline format, making it possible to track chronological changes in the synchronization status with each external system 4. The product DB 51, together with the information management unit 32, also has a function for aggregating and analyzing the synchronization status. This makes it possible to calculate KPIs such as the synchronization success rate and average processing time for each external system 4, and evaluate the performance of each external system 4. Furthermore, the product DB 51, together with the information management unit 32, has a function for classifying error patterns, making it possible to analyze the type and frequency of errors that have occurred and to use this information to help plan measures to prevent recurrence.

[0173] The product DB 51, together with the information management unit 32, has the function of managing the history of alerts. This allows the content of notifications issued in the event of a synchronization failure and the status of their transmission to be recorded. These functions of the product DB 51 make it possible to visualize the status of complex external linkages, detect problems early, and respond quickly. Furthermore, by linking with the external synchronization unit 35, the product DB 51 enables efficient execution control of synchronization processing.

[0174] <Main Management Server 1 Processing> FIG. 4 is a flowchart showing an example of the processing flow of the main management server 1. The main management server 1 stores product information in a standard format (step S1). When the product information stored in the standard format is updated (YES in step S2), the main management server 1 detects the update (step S3) and converts the product information from the standard format to an external format (step S4). This ends the processing of the main management server 1 (END). On the other hand, if the product information is not updated (NO in step S2), the main management server 1 repeats the determination processing of step S2.

[0175] FIG. 5 is a diagram showing a specific example of a setting screen displayed on the user terminal 5 constituting the information processing system S of FIG. The setting screen shown in Fig. 5 is implemented as a GUI that runs on a web browser and can display various setting items such as conversion rule definitions for each external system 4, item mapping settings, data format conversion rules, validation rules, and generation AI utilization parameters. Of these, Fig. 5 shows an example of the GUI that enables item mapping settings.

[0176] The setting screen shown in FIG. 5 makes it possible to visually define which field names in the standard format correspond to which field names in the external system 4 indicated in the external format. Specifically, the setting screen shown in FIG. 5 indicates that the field name in the external format corresponding to the field name "Product Name" in the standard format is "Title," and that the field name in the external format corresponding to the field name "Product Description" in the standard format is "Body." It also indicates that the field name in the external format corresponding to the field name "SKU" in the standard format is "Handle," and that the field name in the external format corresponding to the field name "Publication Status" in the standard format is "Status." It also indicates that the field name in the external format corresponding to the field name "Purchase Price" in the standard format is "Cost per item," and that the field name in the external format corresponding to the field name "Tags" in the standard format is "Tags." It also indicates that the field name in the external format corresponding to the field name "Inventory Quantity" in the standard format is "Variant Inventory Qty," and that the field name in the external format corresponding to the field name "Supplier" in the standard format is "Vendor." It also indicates that the field name in the external format corresponding to the field name "Tax Rate" in the standard format is "Variant Tax."

[0177] Furthermore, the setting screen shown in Figure 5 allows users to set value conversion and join / split rules as needed. The setting screen shown in Figure 5 allows users to easily operate it even if they do not have specialized programming knowledge.

[0178] <Advantageous Effects of the Present Embodiment> According to the above-described embodiment, it is possible to realize the centralized management of product master data across multiple sales channels, thereby improving the efficiency of information updates and ensuring consistency.

[0179] <Other> Although one embodiment of the present invention has been described above, the present invention is not limited to the above-described embodiment, and modifications, improvements, etc. within the scope of achieving the object of the present invention are included in the present invention.

[0180] For example, the above-described series of processes can be executed by hardware or software. In other words, the above-described functional configuration is merely an example and is not particularly limited. In other words, it is sufficient for the information processing system to be provided with a function that can execute the above-described series of processes as a whole, and the type of functional block used to realize this function is not particularly limited to the above-described example.

[0181] Furthermore, the location of the functional blocks is not particularly limited and may be arbitrary. For example, the functional blocks of the main management server 1 may be transferred to another device, or the functional blocks of another device may be transferred to a server. Furthermore, one functional block may be configured as a single piece of hardware, a single piece of software, or a combination of both.

[0182] When a series of processes is executed by software, the programs constituting the software are installed onto a computer or the like from a network or a recording medium. The computer may be a computer incorporated into dedicated hardware. The computer may also be a computer capable of executing various functions by installing various programs, such as a server, a general-purpose smartphone, or a personal computer.

[0183] The recording medium containing such a program may be configured as a removable medium (not shown) that is distributed separately from the device main body in order to provide the program to users, etc., or may be configured as a recording medium that is pre-installed in the device main body and provided to users, etc. Since the program can be distributed via a network, the recording medium may be installed in or accessible from a computer that is connected or connectable to the network.

[0184] In this specification, the steps describing the program recorded on the recording medium include not only processes that are performed in chronological order, but also processes that are not necessarily performed in chronological order but are performed in parallel or individually. Also, in this specification, the term "system" means an overall device composed of multiple devices or multiple means, etc.

[0185] In other words, the information processing system to which the present invention is applied can take various forms having the following configurations. [1] That is, the information processing system S to which the present invention is applied is an information processing system characterized by having a first format storage means (e.g., information management unit 32 in Figure 3) that stores product information related to a product in a first format, a second format storage means (e.g., information management unit 32 in Figure 3) that stores the product information in a second format, an update detection means (e.g., update detection unit 33 in Figure 3) that detects that the product information has been updated, and a format conversion means (e.g., conversion unit 39 in Figure 3) that converts the product information from the first format to the second format when an update of the product information stored in the first format is detected. This allows for centralized management of product master data across multiple sales channels, making it possible to update information more efficiently and ensuring consistency.

[0186] [2] The second format may be a format of the product information in an external system (for example, external system 4 in FIG. 1).

[0187] [3] It may further be characterized by having an external connection means (e.g., external connection unit 34 in Figure 3) that connects to the external system, and an external synchronization means (e.g., external synchronization unit 35 in Figure 3) that synchronizes the product information converted to the second format by providing it to the external system.

[0188] [4] The system may further include a display control means (e.g., the display control unit 36 ​​in FIG. 3) for displaying a setting screen (e.g., the setting screen in FIG. 5) for setting a conversion rule for converting the first format into the second format, and a prompt generation means (e.g., the prompt generation unit 37 in FIG. 3) for generating a prompt to be given to the generation AI based on the set conversion rule.

[0189] [5] The format conversion means may further include one or more of the following means (1) to (11): (1) tag generation means for generating tags to be attached to the merchandise using a generation AI (e.g., the tag generation unit 391 in FIG. 3); (2) appeal statement generation means for generating appeal statements for the merchandise using a generation AI (e.g., the appeal statement generation unit 392 in FIG. 3); (3) spelling variation correction means for correcting spelling variations in the merchandise information using a generation AI (e.g., the spelling variation correction unit 393 in FIG. 3); (4) image conversion means for converting the format of a merchandise image using a generation AI to satisfy the constraints of an external system (e.g., the image conversion unit 394 in FIG. 3); (5) formatting means for formatting descriptions of merchandise information using a generation AI to satisfy the specifications of the external system (e.g., the formatting unit 395 in FIG. 3); (6) summarization means for summarizing the merchandise information in accordance with the data volume limit of the external system using a generation AI (e.g., the summarization unit 396 in FIG. 3). (7) A name matching unit (e.g., the name matching unit 397 in FIG. 3) that uses a generation AI to match and / or align the product information acquired from multiple data sources. (8) A style matching unit (e.g., the style matching unit 398 in FIG. 3) that uses a generation AI to generate the product information that matches the expression style of the person providing the product and / or the industry. (9) A categorization unit (e.g., the categorization unit 399 in FIG. 3) that uses a generation AI to categorize the product. (10) An appeal generation unit (e.g., the appeal generation unit 400 in FIG. 3) that uses a generation AI to generate text to appeal products similar to the best-selling product. (11) A preprocessing unit (e.g., the preprocessing unit 401 in FIG. 3) that uses a generation AI to generate a program for preprocessing data to be input to the generation AI.

[0190] [6] The system may further include an abstraction means for reducing the amount of data related to the product information by abstracting the product information, and the abstraction means (e.g., the abstraction unit 38 in Figure 3) may be characterized in that it performs name matching of organization names included in the product information.

[0191] [7] The system may further include a command sentence conversion means (e.g., the conversion unit 39 in FIG. 3) for converting search and / or analysis instructions for the commercial product in natural language into a command sentence in a predetermined format using a generation AI.

[0192] [8] The system may further include an update history storage means (e.g., information management unit 32 in FIG. 3) that stores the content of updates detected by the update detection means and the update date and time, and a version management means (e.g., version management unit 40 in FIG. 3) that rolls back the version of the product information based on the stored update history.

[0193] [9] The system may further include a query generation means (e.g., the query generation unit 41 in Figure 3) that causes the generation AI to generate and divide queries based on the content of a question created by a user.

[0194]

[10] The system may further include a synchronization result storage means (e.g., information management unit 32 in FIG. 3) for storing the synchronization status of the external system and the processing result of the synchronization, and the external synchronization means may be characterized in that it attempts to synchronize again with the external system with which synchronization has failed.

[0195]

[11] The system may further include a priority storage means (e.g., the information management unit 32 in FIG. 3) for storing priority information regarding the priority when synchronizing multiple external systems, and the external synchronization means may sequentially execute synchronization based on the priority.

[0196]

[12] The system may further include a synchronization identification means (e.g., the external synchronization unit 35 in Figure 3) that identifies the external system that requires synchronization based on the changes in the product information.

[0197]

[13] The system may further include a specification storage means (e.g., information management unit 32 in FIG. 3) for storing specification information relating to the specifications of the external system, a consistency verification means (e.g., consistency verification unit 42 in FIG. 3) for verifying whether the attributes of the product information conform to the constraints of the external system based on the specification information, and an attribute correction means (e.g., attribute correction unit 43 in FIG. 3) for correcting the attributes that do not conform to the constraints.

[0198]

[14] The system may further include an update history storage means (e.g., information management unit 32 in Figure 3) for storing the update history of the product information, and the external synchronization means may synchronize by providing only differential updates to the external system based on the update history of the product information.

[0199]

[15] The system may further include a feedback acquisition means (e.g., information acquisition unit 31 in Figure 3) for acquiring feedback on the product information from the external system, and the first format storage means may update the product information stored in the first format based on the acquired feedback.

[0200] Furthermore, the information processing method to which the present invention is applied can take various forms having the following configurations.

[16] That is, an information processing method to which the present invention is applied is an information processing method characterized by including the steps of: storing product information relating to a product in a first format; storing the product information in a second format; detecting that the product information has been updated; and, when an update of the product information stored in the first format is detected, converting the product information from the first format to the second format.

[0201] Furthermore, the program to which the present invention is applied can take various forms having the following configurations.

[17] In other words, the program to which the present invention is applied is a program for causing a computer to execute control processing including the steps of storing product information relating to a product in a first format, storing the product information in a second format, detecting that the product information has been updated, and, when an update of the product information stored in the first format is detected, converting the product information from the first format to the second format. [Explanation of symbols]

[0202] 1: Main management server, 2: Inventory management server, 3: Purchase management server, 4: External system, 5: User terminal, 11: CPU, 16: Output unit, 17: Input unit, 18: Memory unit, 19: Communication unit, 31: Information acquisition unit, 32: Information management unit, 33: Update detection unit, 34: External connection unit, 35: External synchronization unit, 36: Display control unit, 37: Prompt generation unit, 38: Abstraction unit, 39: Conversion unit, 40: Version control unit, 41: Query generation unit, 42: Consistency verification unit, 43: Attribute modification unit, 44: Transmission control unit, S: Information processing system, N: Network

Claims

1. a first format storage means for storing product information relating to the product in a first format; a second format storage means for storing the product information in a second format; update detection means for detecting that the product information has been updated; a format conversion means for converting the product information stored in the first format into the second format when an update of the product information stored in the first format is detected; An information processing system comprising:

2. The second format is a format of the product information in an external system. The information processing system according to claim 1 .

3. an external connection means for connecting to the external system; an external synchronization means for synchronizing the product information converted into the second format with the external system by providing the product information converted into the second format with the external system; 3. The information processing system according to claim 2, further comprising:

4. a display control means for displaying a setting screen for setting a conversion rule for converting the first format into the second format; a prompt generating means for generating a prompt to be given to the generation AI based on the set conversion rule; 2. The information processing system according to claim 1, further comprising:

5. The format conversion means further includes one or more of the following means (1) to (11): The information processing system according to claim 1 . (1) Tag generation means for generating tags to be attached to the merchandise using a generation AI (2) Appeal statement generation means for generating an appeal statement for the product using generation AI (3) A spelling variation correction means for correcting spelling variations in the product information using a generating AI. (4) Image conversion means for converting the format of a product image using a generation AI to satisfy the constraints of an external system. (5) A formatting means for formatting descriptions of product information using a generating AI to meet the specifications of an external system. (6) A summarization means for summarizing the product information in accordance with the data volume limit of the external system using a generation AI. (7) A name matching unit that uses the generation AI to match and / or match the product information acquired from multiple data sources. (8) Style matching means for generating the product information that matches the expression style of the person providing the product and / or the industry using a generating AI (9) Categorization means for categorizing the commercial materials using a generation AI (10) An appeal generation means for generating text to appeal a product similar to the best-selling product using a generation AI. (11) A preprocessing means for generating a program for preprocessing data to be input to the generation AI using the generation AI.

6. further comprising an abstraction means for abstracting the product information to reduce the amount of data relating to the product information; The abstraction means performs name matching of organization names included in the product information.

2. The information processing system according to claim 1, wherein:

7. The system further includes a command sentence conversion means for converting the search and / or analysis instructions for the commercial products in natural language into a command sentence in a predetermined format using a generation AI.

2. The information processing system according to claim 1, wherein:

8. an update history storage means for storing the content of updates detected by the update detection means and the update date and time; a version management means for performing a rollback of the version of the product information based on the stored update history; 2. The information processing system according to claim 1, further comprising:

9. The system further includes a query generation means for causing the generation AI to generate and divide queries based on the content of a question created by a user.

2. The information processing system according to claim 1, wherein:

10. The system further includes a synchronization result storage means for storing a synchronization state of the external system and a processing result of the synchronization, the external synchronization means attempts to synchronize again with the external system with which synchronization has failed; 4. The information processing system according to claim 3, wherein:

11. The system further includes a priority storage means for storing priority information regarding the priority order when synchronizing the plurality of external systems, the external synchronization means sequentially executes synchronization based on the priority order; 4. The information processing system according to claim 3, wherein:

12. The system further includes a synchronization specification unit that specifies the external system that requires synchronization based on the change content of the product information.

3. The information processing system according to claim 2, wherein:

13. a specification storage means for storing specification information relating to the specifications of the external system; a consistency verification means for verifying whether the attributes of the product information conform to constraints of the external system based on the specification information; attribute modification means for modifying the attributes that do not conform to the constraints; 3. The information processing system according to claim 2, further comprising:

14. The system further includes an update history storage means for storing an update history of the product information, the external synchronization means synchronizes the product information by providing only differential updates to the external system based on the product information update history; 4. The information processing system according to claim 3, wherein:

15. The system further includes a feedback acquisition means for acquiring feedback on the product information from the external system, the first format storage means updates the product information stored in the first format based on the acquired feedback.

3. The information processing system according to claim 2, wherein:

16. An information processing method executed by an information processing system, storing product information relating to the product in a first format; storing the product information in a second format; detecting that the product information has been updated; When an update of the product information stored in the first format is detected, converting the product information from the first format to the second format; An information processing method comprising:

17. On the computer, storing product information relating to the product in a first format; storing the product information in a second format; detecting that the product information has been updated; When an update of the product information stored in the first format is detected, converting the product information from the first format to the second format; A program for executing control processing including:

Images