Method and electronic device for supporting corporate decision-making based on heterogeneous types of public data

Abstract

The present invention relates to a method for supporting corporate decision-making based on different types of public data. The method comprises the steps of: acquiring industrial data from one or more public data sources; normalizing the industrial data according to a predefined data structure based on an industrial domain; generating a feature vector to represent at least one state among the industry, technology, or market to which the company belongs during the corporate decision-making process based on the normalized industrial data; calculating a key performance indicator of the company through a predefined calculation rule based on the feature vector; generating a judgment result regarding the state by comparing the key performance indicator with specific conditions; and outputting explanatory information regarding the judgment basis and judgment path corresponding to the judgment result to a user terminal based on the judgment result.

Description

Technology Field

[0001] The present disclosure relates to a method and an electronic device for supporting corporate decision-making based on different types of public data. Background Technology

[0002] Technologies that utilize various types of public data to support corporate decision-making have recently been receiving significant attention. These technologies help companies more accurately grasp the state of their industries, technologies, and markets, and assist in making strategic decisions based on this understanding. Previously, many decision-making support systems relied primarily on a single type of data or limited data sources. However, these methods had limitations as they failed to adequately reflect the diversity and complexity of the data.

[0003] In particular, integrating and analyzing different types of data—such as policy data, industrial statistics, technology data, patent data, academic papers, and news data—has been challenging due to structural differences. Since these data possess distinct formats and structures, normalization and structuring are essential for integrated analysis. Furthermore, to enhance the reliability of data analysis results, it is crucial to clearly present the basis for judgment and the analysis pathways.

[0004] Therefore, systems that effectively structure heterogeneous public data and support corporate decision-making based on it focus on ensuring data reproducibility and traceability. Such systems support companies in making decisions based on more reliable data and enhance the transparency of decision-making by clearly explaining the results of data analysis. A prior art related to this is Korean Patent Publication No. 10-2020-0111475 (published September 29, 2020, 'Method and system for collecting business data and generating data for achieving key performance indicators'). The problem to be solved

[0005] The objective of one embodiment is to provide a system that structures different types of public data based on industry domains and supports corporate decision-making based thereon. This system aims to determine the state of the industry, technology, or market to which the company belongs through normalized data, and to ensure the reproducibility and traceability of the decision results by clearly presenting the basis for the decision and the decision path corresponding to the decision result. means of solving the problem

[0006] A method for supporting decision-making in an enterprise according to one embodiment may acquire industry data from one or more public data sources, normalize the industry data according to a predefined data structure based on an industry domain, and generate a feature vector to represent at least one state among the industry, technology, or market to which the enterprise belongs during the decision-making process of the enterprise based on the normalized industry data. Additionally, the method may calculate the enterprise's key performance indicators through predefined calculation rules based on the feature vectors, generate a judgment result regarding the state by comparing the key performance indicators with specific conditions, and output explanatory information regarding the judgment basis and judgment path corresponding to the judgment result to a user terminal based on the judgment result.

[0007] In a method according to one embodiment, the public data source may include at least one of policy data, industry statistical data, technology data, patent data, paper data, or news data. Additionally, the previously defined data structure may be defined based on at least one of an entity classification structure, data field configuration, or data schema pre-set to correspond to the industry domain.

[0008] In a method according to one embodiment, the step of generating the feature vector may involve aligning the normalized industry data according to a time standard and combining or summarizing the industry data aligned according to the time standard to generate a feature vector for determining the state of the industry, technology, or market to which the company belongs during the decision-making process of the company.

[0009] In a method according to one embodiment, the step of calculating the key performance indicator may calculate a key performance indicator in the form of an index representing the state of the industry, technology, or market to which the company belongs by combining or weighting a plurality of elements constituting the feature vector according to the predefined calculation rule.

[0010] In a method according to one embodiment, the step of generating the judgment result may generate a judgment result for the state by comparing the key performance indicator with a preset reference value, threshold value, or comparison target to determine whether the state corresponds to any one of a plurality of discrete states.

[0011] In a method according to one embodiment, the step of outputting to the user terminal may generate descriptive information including, together with the judgment result, information configured to identify at least one of a data source, a feature vector, a key performance indicator, and a comparison condition applied until the judgment result is generated.

[0012] An electronic device according to one embodiment may include at least one processor and a memory for storing instructions. When the instructions are executed individually or collectively by the at least one processor, the electronic device may be configured to acquire industrial data from one or more public data sources, normalize the industrial data according to a predefined data structure based on an industrial domain, generate a feature vector for representing at least one state among the industry, technology, or market to which the company belongs in a decision-making process based on the normalized industrial data, calculate a key performance indicator of the company through a predefined calculation rule based on the feature vector, generate a judgment result for the state by comparing the key performance indicator with specific conditions, and output explanatory information regarding the judgment basis and judgment path corresponding to the judgment result to a user terminal based on the judgment result. Effects of the invention

[0013] The data processing system and method according to the present invention can ensure the reproducibility and traceability of judgment results by structuring different types of public data based on industry domains, determining the state of the industry, technology, or market to which a company belongs based thereon, and providing judgment grounds and judgment paths corresponding to the judgment results. Furthermore, the present invention can reliably support corporate decision-making by structurally separating the data analysis step and the explanatory information generation step. These effects can be realized by various embodiments of the present invention, and the effects of the present invention are not limited to such literal descriptions but include everything that a person skilled in the art can infer from the present invention. Brief explanation of the drawing

[0014] FIG. 1 shows an example of a decision support system according to one embodiment. FIG. 2 is a diagram illustrating the operation of each layer according to one embodiment. FIG. 3 is an example of a flowchart of a decision support method according to one embodiment. FIG. 4 shows an example of an electronic device according to one embodiment. Specific details for implementing the invention

[0015] Specific structural or functional descriptions of the embodiments are disclosed for illustrative purposes only and may be modified and implemented in various forms. Accordingly, actual implementations are not limited to the specific embodiments disclosed, and the scope of this specification includes modifications, equivalents, or substitutions included in the technical concept described by the embodiments.

[0016] Terms such as "first" or "second" may be used to describe various components, but these terms should be interpreted solely for the purpose of distinguishing one component from another. For example, the first component may be named the second component, and similarly, the second component may be named the first component.

[0017] When it is stated that a component is "connected" to another component, it should be understood that it may be directly connected to or coupled with that other component, or that there may be other components in between.

[0018] Singular expressions include plural expressions unless the context clearly indicates otherwise. In this document, phrases such as “A or B,” “at least one of A and B,” “at least one of A or B,” “A, B or C,” “at least one of A, B and C,” and “at least one of A, B, or C” may each include any one of the items listed together with the corresponding phrase, or all possible combinations thereof. In this specification, terms such as “comprising” or “having” are intended to designate the existence of the described feature, number, step, action, component, part, or combination thereof, and should be understood as not precluding the existence or addition of one or more other features, numbers, steps, actions, components, parts, or combinations thereof.

[0019] Unless otherwise defined, all terms used herein, including technical or scientific terms, have the same meaning as generally understood by those skilled in the art. Terms such as those defined in commonly used dictionaries should be interpreted as having a meaning consistent with their meaning in the context of the relevant technology, and should not be interpreted in an ideal or overly formal sense unless explicitly defined in this specification.

[0020] Hereinafter, embodiments will be described in detail with reference to the attached drawings. In the description with reference to the attached drawings, identical components are given the same reference numeral regardless of the drawing number, and redundant descriptions thereof will be omitted.

[0022] FIG. 1 shows an example of a decision support system according to one embodiment.

[0023] In one embodiment, the present invention may be applied to a technology for supporting corporate decision-making based on different types of public data. The present invention may include public data having different structures and formats, such as policy data, industrial statistical data, technical data, patent data, research paper data, or news data. The present invention may provide information necessary for corporate decision-making by integrating different data sources by industry domain. The present invention may include data normalization and structuring processes to resolve data heterogeneity. By ensuring data reliability and reproducibility, the present invention may support companies in making decisions based on objective evidence. The present invention may realize a transparent decision-making environment by making the source of data, processing procedures, and basis for judgment traceable.

[0024] In at least one embodiment of the present disclosure, the method may include the step of acquiring industrial data from one or more public data sources. An electronic device may establish access rights to a plurality of public data sources. The electronic device may access the public data sources in various ways, such as APIs, crawling, or database integration. The electronic device may identify the format and structure of the data provided by each public data source. The electronic device may store metadata, such as the time of data acquisition, the reliability of the data source, and the timeliness of the data. The electronic device may store the acquired industrial data in a temporary storage or a permanent storage.

[0025] In at least one embodiment of the present disclosure, the method may include at least one of policy data, industrial statistical data, technical data, patent data, paper data, or news data as a public data source. An electronic device may obtain government policy, statute, and regulatory information from a policy data source. An electronic device may obtain statistical information such as production volume, sales, and employment by industry from an industrial statistical data source. An electronic device may obtain information on technology trends, technology classification, and technology level from a technical data source. An electronic device may obtain information on patent applications, registrations, citations, and classifications from a patent data source. An electronic device may obtain information on academic papers, research results, and citations from a paper data source. An electronic device may obtain information on industry-related news, market trends, and issues from a news data source. An electronic device may construct industrial data by integrating data obtained from a plurality of public data sources.

[0026] In one embodiment, the method may include a step of normalizing industrial data according to a predefined data structure based on an industrial domain. An electronic device may classify acquired industrial data by industrial domain. The electronic device may predefine an entity classification structure, data field configuration, and data schema corresponding to the industrial domain. The electronic device may map the fields of each industrial data to the predefined data structure. The electronic device may process the data by standardizing its format, unit, and name to enable comparison and combination between data. The electronic device may detect and correct erroneous data, missing data, and inconsistent data during the normalization process. The electronic device may transmit the normalized data to a subsequent analysis step.

[0027] In one embodiment, the method may be defined based on at least one of a pre-configured entity classification structure, data field configuration, or data schema so that the pre-defined data structure corresponds to an industrial domain. The electronic device may set an entity classification structure for each industrial domain. The electronic device may define a list of data fields corresponding to each entity. The electronic device may clearly define the structure, relationships, and constraints of the data through the data schema. The electronic device may apply data normalization rules by referring to at least one of the entity classification structure, data field configuration, and data schema. The electronic device may ensure data consistency and scalability through the definition of the data structure.

[0028] In at least one embodiment of the present disclosure, the method may include the step of generating a feature vector to represent the state of at least one of the industry, technology, or market to which the company belongs during the decision-making process of the company, based on normalized industry data. An electronic device may collect normalized industry data as input values. The electronic device may preprocess the data in various ways, such as time-based alignment, multi-point-time synchronization, and discontinuous data processing. The electronic device may generate a multidimensional feature vector by combining or summarizing multiple industry data items. The electronic device may assign values ​​reflecting the state of the industry, technology, or market to each element of the feature vector. The electronic device may numerically represent the state of the company through the feature vector. The electronic device may input the generated feature vector into the KPI calculation step.

[0029] In at least one embodiment of the present disclosure, the method may include the step of calculating a key performance indicator of an enterprise based on a feature vector through a predefined calculation rule. An electronic device may combine each element of the feature vector according to a predefined calculation rule. The electronic device may apply various calculation methods, such as weight matrices, linear or non-linear combination formulas, and parameter settings. The electronic device may calculate the KPI by assigning weights to each element of the feature vector. The electronic device may express the KPI calculation result in various forms, such as an index, a score, or a grade. The electronic device may transmit the KPI calculation result to a state judgment step.

[0030] In one embodiment, the method may include a step of generating a judgment result regarding a state by comparing a key performance indicator with a specific condition. An electronic device may compare the KPI calculation result with a preset reference value, threshold value, or external comparison target. The electronic device may determine whether the KPI value is above, below, or falls within a specific range of the reference value. Based on the KPI value, the electronic device may determine one of a plurality of discrete states, such as growth, stability, risk, or slowdown. The electronic device may express the judgment result as a code, message, grade, etc. The electronic device may transmit the judgment result to a descriptive information generation step.

[0031] In at least one embodiment of the present disclosure, the method may include the step of outputting explanatory information regarding the judgment basis and judgment path corresponding to the judgment result to a user terminal based on the judgment result. The electronic device may generate explanatory information including relevant information such as the judgment basis, judgment path, data source, feature vector, KPI, and comparison condition along with the judgment result. The electronic device may explicitly record the input values, processing steps, and output results for each step in the explanatory information. The electronic device may transmit the explanatory information to the user terminal via a network. The electronic device may output the explanatory information to the user terminal in various ways, such as visualization, reports, and dashboards. The electronic device may ensure the reproducibility and traceability of the judgment result through the provision of explanatory information.

[0032] In one embodiment, the method may include a step of generating explanatory information configured to identify at least one of the data source, feature vector, key performance indicator, and comparison condition applied until the judgment result is generated, along with the judgment result at the step of outputting to a user terminal. The electronic device may include the identifier of the data source used to generate the judgment result, the collection time, and data version information in the explanatory information. The electronic device may record each element of the feature vector and the basis for calculating the corresponding value in the explanatory information. The electronic device may specify the calculation rules, weights, and parameter values ​​applied to the calculation of the KPI in the explanatory information. The electronic device may include condition information, such as the reference value, threshold value, and comparison target used to derive the judgment result, in the explanatory information. The electronic device may add processing logs, error details, and exception handling results for each step to the explanatory information. The electronic device may standardize the structure of the explanatory information to utilize it for reproducibility verification and external auditing. The electronic device can systematically guarantee the reliability, transparency, and traceability of the judgment result through the provision of explanatory information.

[0033] In one embodiment, the database (110) may be configured to store different types of public data, such as policy data, industrial statistics data, technical data, patent data, paper data, or news data, or to store data collected from external public data sources temporarily or permanently. The database (110) may store policy data, industrial statistics data, technical data, patent data, paper data, and news data in separate tables or collections. The database (110) may first store data collected from external public data sources in a temporary storage and then transfer it to a permanent storage after verifying data consistency and quality. The database (110) may manage information regarding the source, time of acquisition, and reliability of the data by storing metadata together with the data by data type. Depending on the data storage method, the database (110) may apply various storage structures, such as relational databases, NoSQL databases, and file systems. The database (110) may ensure the integrity and confidentiality of the data by setting access rights and security policies for the stored data.

[0034] In one embodiment, the database (110) may store multiple data types corresponding to various industrial domains. The database (110) may manage data sets classified by industrial domain in separate partitions or namespaces. The database (110) may maintain structural consistency of the data by applying a data schema specialized for each industrial domain. The database (110) may configure a mapping table between domains to support data integration and comparative analysis across multiple industrial domains. The database (110) may store log and historical information for verifying the timeliness and reliability of data by industrial domain.

[0035] In at least one embodiment of the present disclosure, the database (110) may store data by directly collecting data from an external public data source or by receiving data collected by an external system. The database (110) may support various collection methods, such as API integration with an external public data source, data crawling, and file uploading. The database (110) may receive and store data collected by an external system in a standardized data format. The database (110) may record metadata such as the source of the data, the collection method, and the collection cycle at the time of data collection. The database (110) may monitor the connection status with an external public data source and perform notification and retry procedures in the event of data collection failure or delay.

[0036] In at least one embodiment of the present disclosure, the database (110) can ensure data diversity and reliability through linkage with external public data sources. The database (110) can expand data coverage by collecting data from multiple external public data sources in parallel. The database (110) can compare data from multiple sources for the same data item and select or merge values ​​with high reliability. The database (110) can establish reliability evaluation criteria for external public data sources and store data with low reliability after undergoing a separate verification procedure. The database (110) can continuously monitor data quality by managing reliability indicators and collection history for each data source.

[0037] In one embodiment, the database (110) may store acquired data in a temporary storage or a permanent storage. The database (110) may store data in a temporary storage immediately after data collection to perform data consistency verification and preprocessing. After verification and preprocessing are completed, the database (110) may move the data to a permanent storage to be used for long-term storage and subsequent analysis. The database (110) may track the data processing process by recording the history of data movement between the temporary storage and the permanent storage. The database (110) may set a retention period for the data stored in the temporary storage to perform automatic deletion or transfer to the permanent storage after a certain period has elapsed.

[0038] In at least one embodiment of the present disclosure, the database (110) may operate by distinguishing between temporary storage and permanent storage according to the data processing stage and purpose of use. The database (110) may allocate storage specialized for each stage, such as temporary storage, preprocessing, normalization, and permanent storage, according to the data processing stage. The database (110) may operate various storage types, such as for analysis, backup, and service provision, according to the purpose of use. The database (110) may automate data movement and permission changes between storage when the purpose of data use changes. The database (110) may optimize system efficiency by applying differential capacity, access speed, and security level to the temporary storage and permanent storage.

[0039] In at least one embodiment of the present disclosure, the database (110) can store and manage structured data and unstructured data separately. The database (110) can apply a table-based relational database structure to the structured data. The database (110) can apply a non-relational storage structure, such as a file system, object storage, or NoSQL database, to the unstructured data. The database (110) can manage storage paths, indexing methods, and search methods by differentiating them according to data type. The database (110) can support integrated querying by linking association information between structured data and unstructured data as metadata.

[0040] In at least one embodiment of the present disclosure, the database (110) may apply a suitable storage method according to the structural characteristics of the data. The database (110) may automatically select an optimal storage type by analyzing the field structure, data size, access frequency, etc. of the data. For large volumes of unstructured data, the database (110) may apply a distributed file system to ensure storage efficiency and scalability. For structured data requiring real-time analysis, the database (110) may apply an in-memory database to improve processing speed. The database (110) may automate schema migration and data conversion procedures when the data structure is changed.

[0041] In one embodiment, the database (110) can structure data so that comparison and combination between data are possible in subsequent steps by converting industrial data obtained from different public data sources to match the same entity classification structure, data field configuration, or data schema. The database (110) can pre-define conversion rules for mapping data fields and entities for each public data source. The database (110) can standardize field names, data types, units, and code systems during the data normalization process. The database (110) can support data comparison, combination, and statistical analysis between industrial domains based on the structured data. The database (110) can ensure consistency and traceability of the data structure by managing the history of data schema changes.

[0042] In at least one embodiment of the present disclosure, the database (110) can ensure data consistency and compatibility by applying a predefined data schema for each industrial domain. The database (110) can predefine a data schema for each industrial domain that includes an entity classification structure, data field configuration, data type, constraints, etc. The database (110) can perform data validity verification according to the domain-specific schema when new data is introduced. The database (110) can support interoperability and integrated analysis between data based on the domain-specific schema. The database (110) can minimize compatibility issues when changing data structures through schema version management.

[0043] In at least one embodiment of the present disclosure, the database (110) may store metadata, such as the time of data acquisition, the reliability of the data source, and the data's freshness, to ensure the data's freshness. The database (110) may automatically generate and store metadata, such as timestamps, source IDs, and reliability scores, when collecting data. The database (110) may set standards for the data's freshness and perform update or disposal procedures for data that has elapsed for a certain period. The database (110) may monitor the data's freshness status in real time based on the metadata. The database (110) may maintain data quality by managing the freshness history and update cycle for each data source.

[0044] In one embodiment, the database (110) can maintain the latest state of data through periodic data update and verification procedures. The database (110) can periodically recollect or update data by setting update cycles for each data source. When updating data, the database (110) can record a change history by comparing existing data with new data. The database (110) can identify and automatically correct erroneous, missing, or inconsistent data through data verification procedures. The database (110) can log the results of data updates and verifications to be used for data quality management and auditing.

[0045] In one embodiment, the electronic device (130) may be configured to perform data processing, judgment, and the generation of explanatory information, which are core elements of the present invention. The electronic device (130) may receive industrial data input from a database (110). The electronic device (130) may process the received industrial data step by step in the order of normalization, analysis, judgment, and generation of explanatory information. The electronic device (130) may perform data processing and judgment functions through modularized software or hardware components at each step. The electronic device (130) may generate explanatory information regarding the judgment basis and judgment path along with the judgment result. The electronic device (130) may transmit the generated judgment result and explanatory information to a user terminal (120) via a network. By integrating data processing and judgment functions, the electronic device (130) can ensure the reproducibility and traceability of the judgment result.

[0046] In at least one embodiment of the present disclosure, an electronic device (130) may acquire industrial data from a database (110). The electronic device (130) may request industrial data periodically or on an event basis through a network interface with the database (110). The electronic device (130) may store metadata (source, time of acquisition, etc.) of the industrial data received from the database (110) together to ensure the reliability and traceability of the data. The electronic device (130) may integrate and process industrial data acquired from multiple data sources.

[0047] In one embodiment, the electronic device (130) can process acquired industrial data to generate judgment results and explanatory information. The electronic device (130) can convert the acquired industrial data into a predefined data structure through a normalization module. The electronic device (130) can generate feature vectors by sorting the normalized data by time, or by combining or summarizing data from multiple time points. The electronic device (130) can calculate key performance indicators through a KPI calculation engine using the feature vectors as input. The electronic device (130) can generate a state judgment result by comparing the calculated key performance indicators with a reference value, a threshold value, or a comparison target. The electronic device (130) can generate judgment grounds and judgment paths corresponding to the judgment result through an explanatory information generation module.

[0048] In at least one embodiment of the present disclosure, the electronic device (130) can transmit the generated judgment result and explanatory information to the user terminal (120) via a network. The electronic device (130) can convert the judgment result and explanatory information into a standardized message format through an output module. The electronic device (130) can transmit data to the user terminal (120) in real-time or batch mode via a network interface. The electronic device (130) can check whether the judgment result and explanatory information transmitted to the user terminal (120) have been received, and can perform a retransmission procedure in case of transmission failure. The electronic device (130) can provide customized data formats or visualization information for each user terminal (120).

[0049] In at least one embodiment of the present disclosure, the electronic device (130) may include an external data layer (210), a data collection and normalization layer (230), an analysis layer (250), and an output layer (270). The electronic device (130) may configure independent hardware or software modules for each layer. The electronic device (130) may acquire data from a global public verification data source (211) in the external data layer (210). The electronic device (130) may perform data collection and normalization through a source-specific collection module (231) and a normalization module (233) in the data collection and normalization layer (230). The electronic device (130) may perform data analysis and state determination through a feature vector generation module (251), a KPI calculation engine (253), and a rule-based judgment engine (255) in the analysis layer (250). The electronic device (130) can generate and output judgment results and explanation information through the explanation information generation module (271) and the output module (273) in the output layer (270).

[0050] In one embodiment, the electronic device (130) can perform data acquisition, normalization, analysis, descriptive information generation, and output functions separately for each layer. The electronic device (130) can standardize the input and output interfaces of each layer by clearly defining the data flow for each layer. The electronic device (130) can increase the maintainability and scalability of the system by minimizing dependencies between modules for each layer. The electronic device (130) can limit the scope of impact in the event of a failure by separating responsibilities by function for each layer. The electronic device (130) can be designed to enable independent expansion or upgrade for each layer.

[0051] In one embodiment, the electronic device (130) can ensure data quality management, analysis reliability, and the reproducibility and traceability of explanatory information through a layered design. The electronic device (130) can apply data integrity verification and quality management policies at the data collection and normalization layer. The electronic device (130) can perform reliability verification of data processing logic and result verification procedures at the analysis layer. The electronic device (130) can ensure reproducibility and traceability by including metadata regarding judgment results and explanatory information at the output layer. The electronic device (130) can track the entire data processing path by recording the processing history of each step through the layered design.

[0052] In at least one embodiment of the present disclosure, the electronic device (130) may receive industrial data from the database (110). The electronic device (130) may use a standardized network protocol or API for communication with the database (110). The electronic device (130) may record the time of receipt, data type, and source information of the data received from the database (110). The electronic device (130) may monitor the connection status with the database (110) and perform a retry procedure in the event of a failure to receive data.

[0053] In one embodiment, the electronic device (130) can process input data in the order of normalization, analysis, judgment, and generation of explanatory information. The electronic device (130) can convert input data into a predefined data structure through a normalization module. The electronic device (130) can generate a feature vector by sorting the normalized data by time, or by combining or summarizing data from multiple time points. The electronic device (130) can calculate a key performance indicator through a KPI calculation engine using the feature vector as input. The electronic device (130) can generate a state judgment result by comparing the calculated key performance indicator with a reference value, a threshold value, or a comparison target. The electronic device (130) can generate a judgment basis and a judgment path corresponding to the judgment result through an explanatory information generation module.

[0054] In at least one embodiment of the present disclosure, the electronic device (130) can output the processed judgment result and explanatory information to the user terminal (120) through a network interface. The electronic device (130) can convert the judgment result and explanatory information into a standardized message format through an output module. The electronic device (130) can transmit data to the user terminal (120) in real-time or batch mode through a network interface. The electronic device (130) can provide customized data formats or visualization information for each user terminal (120). The electronic device (130) can check whether the judgment result and explanatory information transmitted to the user terminal (120) have been received, and can perform a retransmission procedure if transmission fails.

[0055] In at least one embodiment of the present disclosure, the electronic device (130) may be a server, a cloud computing device, or a distributed processing system in terms of hardware. The electronic device (130) may operate in a single server or a cluster environment composed of multiple servers. The electronic device (130) may perform data processing and storage by utilizing virtualized resources in a cloud computing environment. The electronic device (130) may support parallel processing and scalability of large-scale data through a distributed processing system. The electronic device (130) may dynamically allocate computing resources, storage resources, and network resources according to the hardware configuration.

[0056] In one embodiment, the electronic device (130) may be composed of a plurality of servers or distributed nodes to support large-scale data processing and parallel computation. The electronic device (130) may improve processing speed through data partitioning and parallel processing between distributed nodes. The electronic device (130) may apply data consistency and synchronization mechanisms in a distributed environment. The electronic device (130) may provide data replication and automatic failure recovery functions in the event of failure between nodes. The electronic device (130) may flexibly respond to an increase in data throughput depending on the scalability of the distributed processing system.

[0057] In one embodiment, the electronic device (130) may include redundancy and a failure recovery mechanism to ensure that service can continue even in the event of a failure. The electronic device (130) may apply a redundancy (Active-Active or Active-Standby) configuration to key hardware and network components. The electronic device (130) may perform a failure recovery procedure that automatically switches traffic to an alternate node or device in the event of a failure. The electronic device (130) may log the history of failures and the recovery process to be used for analyzing the cause of failures and for prevention. The electronic device (130) may verify the validity of the high availability policy through regular failure recovery scenario tests.

[0058] In at least one embodiment of the present disclosure, the electronic device (130) may apply horizontal scaling or clustering configurations to respond to an increase in data throughput. The electronic device (130) may ensure system scalability by dynamically adding or removing the number of servers or nodes. The electronic device (130) may distribute data processing requests through a load balancer in a clustering environment. The electronic device (130) may ensure data integrity by applying data consistency and synchronization policies during scaling. The electronic device (130) may utilize container-based deployment or orchestration tools to ensure scalability.

[0059] In one embodiment, the electronic device (130) may include a network interface for communication with the database (110) and the user terminal (120). The electronic device (130) may support various network interfaces such as Ethernet, wireless LAN, and optical communication. The electronic device (130) may use standard protocols such as TCP / IP, HTTP, and HTTPS for data transmission and reception with the database (110). The electronic device (130) may support various communication methods such as RESTful API, WebSocket, and gRPC for communication with the user terminal (120). The electronic device (130) may monitor the status of the network interface to enable automatic recovery or use an alternative path in the event of a failure.

[0060] In one embodiment, the electronic device (130) may use a standardized API or message queue for linking with an external data source and for transmitting and receiving data. The electronic device (130) may apply authentication and access control policies, such as OAuth or API Key, for linking with an external data source. The electronic device (130) may support asynchronous data transmission and reception and large-scale data processing by utilizing a message queue. The electronic device (130) may ensure the reliability of data collection by periodically checking the linkage status with the external data source. The electronic device (130) may ensure the traceability of the data flow by recording the linkage history with the external data source.

[0061] In one embodiment, the electronic device (130) may include a source-specific collection module (231) and a normalization module (233) in the data collection and normalization layer (230). The electronic device (130) may acquire industrial data from a global public verified data source (211) of the external data layer (210) through the source-specific collection module (231). The electronic device (130) may convert the acquired industrial data into a predefined data structure based on industrial domain standards through the normalization module (233). The electronic device (130) may structure the normalized data according to an entity classification structure, data field configuration, and data schema so that it can be compared and combined in a subsequent analysis step. The electronic device (130) may perform a procedure to verify the integrity, consistency, and reliability of the data during the normalization process.

[0062] In one embodiment, the electronic device (130) may include a step of sorting normalized industrial data according to a time reference. The electronic device (130) may perform sorting in ascending or descending order based on the timestamp or collection time of the normalized industrial data. The electronic device (130) may prepare the data so that time series analysis and change tracking of the data are possible through time reference sorting. The electronic device (130) may ensure the continuity of the time series data by interpolating or correcting data at discontinuous or missing time points.

[0063] In at least one embodiment of the present disclosure, the electronic device (130) may perform the operation of aligning normalized industrial data according to a time standard in the step of generating a feature vector. The electronic device (130) may group the data aligned according to a time standard according to multiple time points and assign them to each element of the feature vector. The electronic device (130) may extract statistical characteristics such as the amount of change, mean, and variance of the time series data and reflect them in the feature vector. The electronic device (130) may quantitatively express changes in the state of the enterprise through time standard alignment and statistical summarization.

[0064] In at least one embodiment of the present disclosure, the electronic device (130) may include a step of combining or summarizing industry data sorted by time in the step of generating a feature vector to generate a feature vector for determining the state of the industry, technology, or market to which the company belongs in the decision-making process of the company. The electronic device (130) may generate a multidimensional feature vector by sorting data collected from a plurality of industry data sources by time and then combining data from the same point in time or adjacent points in time. The electronic device (130) may summarize key indicators of the combined data (e.g., growth rate, market share, investment scale, etc.) and assign them to each dimension of the feature vector. The electronic device (130) may adjust the resolution and analysis efficiency of the feature vector by applying dimensionality reduction or expansion techniques. The electronic device (130) may utilize the generated feature vector as input data for an analysis layer.

[0065] In at least one embodiment of the present disclosure, the electronic device (130) can convert industrial data obtained from different public data sources through a normalization module (233) to conform to the same entity classification structure, data field configuration, or data schema. The electronic device (130) can convert field names, data types, units, etc., which differ by data source, into a standardized schema. The electronic device (130) can classify data by domain, such as industry, technology, or market, according to the entity classification structure. The electronic device (130) can structure the data to ensure consistency in the data field configuration, thereby facilitating subsequent analysis and comparison. The electronic device (130) can detect missing or erroneous data during the data schema conversion process and perform correction or exception handling.

[0066] In one embodiment, the electronic device (130) may include a feature vector generation module (251), a KPI calculation engine (253), and a rule-based judgment engine (255) in the analysis layer (250). The electronic device (130) can generate a feature vector representing the state of an enterprise from normalized industrial data through the feature vector generation module (251). The electronic device (130) can calculate a key performance indicator by receiving the feature vector as input through the KPI calculation engine (253). The electronic device (130) can generate a state judgment result by comparing the calculated key performance indicator with a reference value, a threshold value, or a comparison target through the rule-based judgment engine (255). The electronic device (130) can connect the data flow between each module of the analysis layer through a standardized interface.

[0067] In one embodiment, the electronic device (130) may include a step of combining or summarizing industry data sorted by time in the step of generating a feature vector to generate a feature vector for determining the state of the industry, technology, or market to which the company belongs during the decision-making process of the company. The electronic device (130) may group the data sorted by time according to multiple time points and assign them to each element of the feature vector. The electronic device (130) may generate a multidimensional feature vector by combining key indicators of multiple industry data. The electronic device (130) may assign domain-specific state information, such as industry, technology, or market, to each dimension of the feature vector. The electronic device (130) may utilize the generated feature vector as input data for the KPI calculation step.

[0068] In one embodiment, the electronic device (130) may include a step of calculating a key performance indicator in the form of an index representing the state of the industry, technology, or market to which the company belongs by combining or weighting multiple elements constituting a feature vector according to predefined calculation rules in the KPI calculation step. The electronic device (130) may perform weighted summation or combination operations by applying a weight matrix to each element of the feature vector in the KPI calculation step. The electronic device (130) may calculate the KPI according to predefined linear or non-linear calculation rules. The electronic device (130) may convert the KPI calculation result into a key performance indicator in the form of an index by normalizing or scaling it. The electronic device (130) may reflect parameter settings and learning results in the KPI calculation process.

[0069] In at least one embodiment of the present disclosure, the electronic device (130) may include a step of generating a judgment result for a state by comparing a key performance indicator with a preset reference value, threshold value, or comparison target in the step of generating a judgment result, and determining whether the state corresponds to one of a plurality of discrete states. The electronic device (130) may perform a comparison operation that compares the KPI calculation result with a reference value, threshold value, or external comparison target. The electronic device (130) may select one of a plurality of discrete states, such as growth, stability, risk, or slowdown, based on the comparison result. The electronic device (130) may generate a state code or message for the judgment result and transmit it to a subsequent explanation information generation step. The electronic device (130) may set a hierarchy and priority of the judgment logic to ensure the reliability and consistency of the judgment result.

[0070] In one embodiment, the electronic device (130) may include an explanation information generation module (271) and an output module (273) in the output layer (270). The electronic device (130) may generate explanation information regarding the judgment basis and judgment path corresponding to the judgment result through the explanation information generation module (271). The electronic device (130) may include metadata in the explanation information, such as the data source used in the judgment, feature vector, key performance indicator, and applied conditions. The electronic device (130) may ensure reproducibility and traceability by standardizing the structure of the explanation information. The electronic device (130) may ensure transparency of the judgment path by recording the history of explanation information generation.

[0071] In one embodiment, the electronic device (130) may generate explanatory information regarding the judgment basis and judgment path corresponding to the judgment result. The electronic device (130) may include detailed information in the explanatory information, such as the identifier of the data source used for judgment, the time of collection, each element of the feature vector, the KPI calculation formula, and the comparison criteria, along with the judgment result. The electronic device (130) may generate the judgment path in the form of visualized data or step-by-step logs. The electronic device (130) may enhance traceability by assigning a unique identifier or token to each item of the explanatory information. The electronic device (130) may record the time of creation and creator information to verify the reliability of the explanatory information.

[0072] In at least one embodiment of the present disclosure, the electronic device (130) can transmit and output the generated descriptive information to the user terminal (120) via a network. The electronic device (130) can convert the descriptive information into a standardized message format through an output module (273). The electronic device (130) can transmit the descriptive information to the user terminal (120) in real-time or in batch mode via a network interface. The electronic device (130) can provide customized descriptive information formats or visualization data for each user terminal (120). The electronic device (130) can ensure reproducibility and traceability by recording the transmission history and reception status of the descriptive information.

[0073] In at least one embodiment of the present disclosure, the user terminal (120) is a device that receives judgment results and explanatory information generated by an electronic device (130) and outputs them to the user in a visual or other manner, and may be, for example, a personal computer, tablet, smartphone, or web-based dashboard environment. The user terminal (120) may receive judgment results and explanatory information from the electronic device (130) via a network. The user terminal (120) may provide the received judgment results and explanatory information to the user in various output methods, such as a display, notification, or voice guidance. The user terminal (120) may include various output devices in hardware, such as an LCD, OLED, touchscreen, or speaker. The user terminal (120) may operate in various execution environments in software, such as a web browser, mobile application, or desktop application. The user terminal (120) may output judgment results and explanatory information in an optimized form by adapting to various system environments, such as an operating system, network environment, and display resolution. Depending on the output method of the judgment results and explanatory information, the user terminal (120) may support various visualization formats, such as tables, graphs, messages, and dashboards. When outputting the judgment result and explanation information, the user terminal (120) may display metadata (e.g., time of reception, data source, judgment path, etc.) to ensure the data's freshness, reliability, and traceability.

[0074] In at least one embodiment of the present disclosure, the user terminal (120) may operate in various hardware and software environments and may receive data from the electronic device (130) via a network. The user terminal (120) may communicate with the electronic device (130) via a wired or wireless network (e.g., Wi-Fi, LTE, 5G, Ethernet, etc.). The user terminal (120) may ensure the stability of data reception by monitoring the network connection status in real time. The user terminal (120) may apply security protocols such as encryption, authentication, and integrity verification when receiving data. The user terminal (120) may provide a retry or error notification function in the event of a network failure or failure to receive data. The user terminal (120) may run on various operating systems (e.g., Windows, macOS, Android, iOS, etc.). In a web-based environment, the user terminal (120) may display judgment results and explanatory information in the form of a dashboard using standard technologies such as HTML5, JavaScript, and CSS. The user terminal (120) can provide judgment results and explanation information in the form of a native app or a hybrid app in a mobile environment.

[0075] In one embodiment, the user terminal (120) receives judgment results and explanatory information generated by the electronic device (130) and can output them to the user in a visual or other manner. The user terminal (120) can display the judgment results and explanatory information on a display screen using visual elements such as text, icons, colors, and graphs. The user terminal (120) can display different colors or icons depending on the state of the judgment result (e.g., growth, stability, danger, slowdown, etc.) so that the user can intuitively recognize the state. The user terminal (120) can display key items of the explanatory information (e.g., judgment basis, judgment path, data source, etc.) by separating them into a separate area or pop-up window. In addition to visual output, the user terminal (120) can support various output methods such as voice guidance, vibration, and notification messages. When visualizing the judgment results and explanatory information, the user terminal (120) can increase the transparency of the information by displaying the reliability, timeliness, and source of the data together.

[0076] In at least one embodiment of the present disclosure, the user terminal (120) may display judgment results and explanatory information in various visualization formats, such as tables, graphs, messages, and dashboards. The user terminal (120) may arrange each item of the judgment results and explanatory information (e.g., status, KPI value, data source, etc.) in a table format by separating them into rows and columns. The user terminal (120) may visually represent changes in KPI values, status distribution, data trends, etc., using graphs (e.g., bar graphs, line graphs, pie charts, etc.). The user terminal (120) may integrate and display multiple judgment results, explanatory information, and metadata on a single screen in a dashboard format. The user terminal (120) may notify the user in real time of summary information of judgment results or warning messages in a message format. The user terminal (120) may provide a visualization format selection function to allow the user to check the judgment results and explanatory information in a desired manner.

[0077] In one embodiment, the user terminal (120) can visually represent at least one of the judgment basis, judgment path, data source, feature vector, key performance indicator, and applied condition included in the explanatory information so that it can be identified. The user terminal (120) can display the judgment basis as a text description, highlight, tooltip, etc., so that the user can easily understand the reason for the judgment. The user terminal (120) can clearly show the flow of the judgment process by expressing the judgment path as a flowchart, step-by-step visualization, timeline, etc. The user terminal (120) can ensure the reliability and traceability of the data by displaying data source information as an icon, link, source indication, etc. The user terminal (120) can visualize the feature vector and key performance indicator as numbers, graphs, charts, etc., so that the user can intuitively check the basis data of the status. The user terminal (120) can clearly convey the judgment criteria by displaying the applied condition (e.g., reference value, threshold value, etc.) as a separate information window or tooltip. The user terminal (120) can enhance the user's understanding by providing explanation or help functions for each visualization element.

[0078] In at least one embodiment of the present disclosure, a user terminal (120) may receive judgment results and explanatory information from an electronic device (130) via a network. Upon receiving judgment results and explanatory information, the user terminal (120) may provide a real-time notification or periodic update function. The user terminal (120) may monitor the time of data reception, the timeliness of the data, network status, etc., and notify the user. In the event of data reception failure or an error, the user terminal (120) may provide a retry or error message display function. The user terminal (120) may ensure data integrity and security by applying security functions, such as encryption and authentication, during the data reception process.

[0079] In one embodiment, the user terminal (120) may support the user in performing interactions such as viewing, searching, checking details, comparing, and downloading judgment results and explanation information. The user terminal (120) may allow the user to view a list of judgment results and select a specific result to check detailed information. The user terminal (120) may provide sorting, filtering, and search functions for each item of judgment results and explanation information. The user terminal (120) may provide a comparison view or chart function that allows multiple judgment results or explanation information to be compared. The user terminal (120) may allow the judgment results and explanation information to be downloaded in various file formats such as PDF, Excel, and images. The user terminal (120) may record the user's interaction history and provide personalization functions such as recent viewing history, favorites, and personalized recommendations.

[0080] In at least one embodiment of the present disclosure, the user terminal (120) may provide various user interaction scenarios, such as updating, filtering, and changing conditions of judgment results and explanatory information, in response to a user's request. The user terminal (120) may allow the user to check the latest judgment results and explanatory information in real time by clicking a data update button or setting an automatic update cycle. The user terminal (120) may allow the user to filter judgment results and explanatory information by selecting desired conditions (e.g., period, industry domain, KPI item, etc.). The user terminal (120) may allow the user to request new judgment results and explanatory information by changing judgment conditions, such as threshold values ​​or reference values. When conditions are changed, the user terminal (120) may update and display judgment results and explanatory information in real time according to the changed conditions. The user terminal (120) may record the history of data changes based on user interaction to ensure future reproducibility and traceability.

[0081] In one embodiment, an electronic device (130) can acquire industrial data from a database (110). The electronic device (130) can be connected to the database (110) via a network. The electronic device (130) can selectively request at least one of policy data, industrial statistics data, technical data, patent data, paper data, or news data stored in the database (110). The electronic device (130) can store the data received from the database (110) in internal memory. The electronic device (130) can verify the integrity and timeliness of the data when acquiring the data. The electronic device (130) can detect exceptional situations, such as network errors, missing data, or format inconsistencies, during the data acquisition process.

[0082] In one embodiment, the electronic device (130) can process acquired industrial data to generate judgment results and explanatory information. The electronic device (130) can convert the acquired industrial data into a predefined data structure in the data collection and normalization layer. The electronic device (130) can transmit the normalized data to the analysis layer. In the analysis layer, the electronic device (130) can generate a vector for state representation through a feature vector generation module. The electronic device (130) can calculate key performance indicators based on the feature vector through a KPI calculation engine. The electronic device (130) can generate judgment results by comparing the KPI with a reference value, threshold value, or comparison target through a rule-based judgment engine. The electronic device (130) can generate judgment grounds and judgment paths corresponding to the judgment results in the explanatory information generation module. The electronic device (130) can include identification information such as data sources, feature vectors, KPIs, and application conditions in the explanatory information. The electronic device (130) can log the process of generating judgment results and explanatory information step by step.

[0083] In at least one embodiment of the present disclosure, the electronic device (130) can transmit the generated judgment result and explanation information to the user terminal (120) via a network. The electronic device (130) can convert the judgment result and explanation information into a network packet form through an output module of the output layer. The electronic device (130) can transmit the judgment result and explanation information to the user terminal (120) using standard protocols such as TCP / IP, HTTP, and WebSocket. The electronic device (130) can apply security procedures such as encryption, authentication, and integrity verification during data transmission. The electronic device (130) can detect packet loss, transmission delay, network failure, etc., during the data transmission process and perform retransmission or error notification. The electronic device (130) can confirm from the user terminal (120) whether the transmitted judgment result and explanation information have been received. The electronic device (130) can record the transmission history and results in a system log.

[0084] In one embodiment, the electronic device (130) may sequentially perform a series of steps including data acquisition, processing, judgment, generation of explanatory information, and output. In the data acquisition step, the electronic device (130) may request and receive data from the database (110). The electronic device (130) may sequentially transmit the received data to the normalization and analysis steps. The electronic device (130) may record the status of the data, processing results, and exception situations at each step. After generating judgment results and explanatory information, the electronic device (130) may transmit the information to the user terminal (120) through the output layer. The electronic device (130) may perform reprocessing or exception handling in the event of an error at each step of the entire sequence. The electronic device (130) may generate metadata for each step to ensure the reproducibility and traceability of the data flow.

[0085] In at least one embodiment of the present disclosure, an electronic device (130) can obtain industrial data by communicating with a database (110). The electronic device (130) can communicate with the database (110) through a standardized interface such as a REST API, SQL query, or database connector. When requesting data from the database (110), the electronic device (130) can perform procedures such as authentication, authorization verification, and data format conversion. The electronic device (130) can monitor the communication status with the database (110) in real time.

[0086] In one embodiment, the electronic device (130) can internally transmit data between multiple modules, such as a data collection and normalization layer, an analysis layer, and an output layer. When transmitting data between each layer, the electronic device (130) can apply a standardized data schema and message format. The electronic device (130) can transmit normalized data from the data collection and normalization layer to the analysis layer. The electronic device (130) can transmit judgment results and explanatory information generated in the analysis layer to the output layer. When transmitting data between each module, the electronic device (130) can verify data integrity and consistency. The electronic device (130) can perform exception handling when an interface error occurs between modules.

[0087] In one embodiment, the electronic device (130) can transmit the generated judgment result and explanation information to the user terminal (120) through the output layer. The electronic device (130) can convert the judgment result and explanation information into network packets at the output layer. The electronic device (130) can manage the data transmission status, whether transmission was successful, and whether to retransmit in case of an error at the output layer. The electronic device (130) can receive a response from the user terminal (120) at the output layer to check whether data has been received.

[0088] In one embodiment, the interface between each module may be defined according to the input, processing, and output flow of data. The electronic device (130) may receive data from the database (110) at the input interface. The electronic device (130) may transmit data at each stage, such as normalization, analysis, judgment, and generation of explanatory information, at the processing interface. The electronic device (130) may transmit judgment results and explanatory information to the user terminal (120) at the output interface. The electronic device (130) may define a data format, transmission method, and error handling policy for each interface.

[0089] In at least one embodiment of the present disclosure, the electronic device (130) may perform real-time data processing and batch data processing in parallel. The electronic device (130) may operate by distinguishing between a real-time data processing mode and a batch data processing mode. The electronic device (130) may utilize event-based triggers during real-time data processing. During batch data processing, the electronic device (130) may process large volumes of data in batches through a periodic scheduler or scheduled tasks. The electronic device (130) may maintain a consistent data flow by integrating the results of real-time and batch processing.

[0090] In one embodiment, the electronic device (130) can immediately generate judgment results and explanatory information upon real-time data collection and transmit them to the user terminal (120). Upon receiving data upon real-time data collection, the electronic device (130) can sequentially perform normalization, analysis, judgment, and explanatory information generation steps. The electronic device (130) can immediately transmit real-time processing results to the user terminal (120) via a network. The electronic device (130) can detect exceptional situations, such as data delays or network failures, during the real-time processing and respond immediately.

[0091] In at least one embodiment of the present disclosure, the electronic device (130) can collect, process, and output data in a batch processing manner in the case of large volume data or periodic data. The electronic device (130) can request a large volume of data in batches from the database (110) in batch processing mode. The electronic device (130) can perform data normalization, analysis, judgment, and descriptive information generation steps in batches during batch processing. The electronic device (130) can transmit the batch processing results to the user terminal (120) at regular intervals. The electronic device (130) can detect exceptional situations, such as data omission or processing failure, during the batch processing process and perform reprocessing or error notification.

[0092] In one embodiment, the electronic device (130) can integrate real-time and batch processing results to provide a consistent information flow. The electronic device (130) can manage the real-time processing results and batch processing results in an integrated manner. The electronic device (130) can provide the integrated results to the user terminal (120) in the same format. The electronic device (130) can record metadata of the real-time and batch processing results (e.g., processing time, data source, processing method, etc.) together. The electronic device (130) can ensure consistency, reliability, and traceability of the real-time and batch processing results.

[0093] In one embodiment, an electronic device (130) can acquire industrial data from a database (110). The electronic device (130) can be connected to the database (110) via a network. The electronic device (130) can selectively request multiple types of public data stored in the database (110). The electronic device (130) can temporarily or permanently store the data received from the database (110) in an internal storage. When acquiring data, the electronic device (130) can verify the integrity, timeliness, and format consistency of the data. The electronic device (130) can detect and log exceptional situations, such as network failures, missing data, and format inconsistencies, during the data acquisition process.

[0094] In one embodiment, the database (110) may be configured to store different types of public data, such as policy data, industrial statistics data, technical data, patent data, paper data, or news data, or to store data collected from external public data sources temporarily or permanently. The database (110) may store different types of data separately through multiple data tables or schemas. The database (110) may periodically update data through integration with external public data sources. The database (110) may store metadata such as the source of the data, the time of collection, and the data type. The database (110) may include a procedure for verifying data integrity, duplication, and omission. The database (110) may manage the data retention period by distinguishing between temporary storage and permanent storage.

[0095] In one embodiment, the electronic device (130) can perform data transmission and reception with the database (110) via a network. The electronic device (130) can perform data transmission and reception with the database (110) via various network protocols such as TCP / IP, HTTP, WebSocket, and gRPC. The electronic device (130) can apply security protocols such as encryption, authentication, and integrity verification when transmitting and receiving data. The electronic device (130) can detect packet loss, transmission delay, network failure, etc., during the data transmission and reception process and perform retransmission or error notification. The electronic device (130) can record the data transmission and reception history and results in a system log.

[0096] In at least one embodiment of the present disclosure, the electronic device (130) can process and transmit the generated judgment result and explanation information to the user terminal (120) via a network. The electronic device (130) can convert the judgment result and explanation information into network packets in an output module of the output layer. The electronic device (130) can transmit the judgment result and explanation information to the user terminal (120) using standard protocols such as TCP / IP, HTTP, and WebSocket. The electronic device (130) can apply security procedures such as encryption, authentication, and integrity verification during data transmission. The electronic device (130) can detect packet loss, transmission delay, network failure, etc., during the data transmission process and perform retransmission or error notification. The electronic device (130) can confirm from the user terminal (120) whether the transmitted judgment result and explanation information have been received. The electronic device (130) can record the transmission history and results in a system log.

[0097] In at least one embodiment of the present disclosure, the output module (273) can output the explanation information generated by the explanation information generation module (271) by transmitting it to the user terminal (120) via a network. The output module (273) can receive the explanation information from the explanation information generation module (271). The output module (273) can convert the explanation information into a network packet or an API response. The output module (273) can transmit the converted explanation information to the user terminal (120) using standard protocols such as TCP / IP, HTTP, or WebSocket. The output module (273) can apply security procedures such as encryption, authentication, and integrity verification during data transmission. The output module (273) can perform retransmission or error notification if an error occurs during the data transmission process. The output module (273) can record the transmission history and results in a system log.

[0098] In one embodiment, the user terminal (120) is a device that receives judgment results and explanatory information generated by an electronic device (130) and outputs them to the user in a visual or other manner, and may be, for example, a personal computer, tablet, smartphone, or web-based dashboard environment. The user terminal (120) may receive judgment results and explanatory information from the electronic device (130) via a network. The user terminal (120) may provide the received judgment results and explanatory information to the user in various output methods, such as a display, notification, or voice guidance. The user terminal (120) may include various output devices in terms of hardware, such as an LCD, OLED, touchscreen, or speaker. The user terminal (120) may operate in various execution environments in terms of software, such as a web browser, mobile application, or desktop application. The user terminal (120) may adapt to various system environments, such as an operating system, network environment, and display resolution, to output judgment results and explanatory information in an optimized form. Depending on the output method of the judgment results and explanatory information, the user terminal (120) may support various visualization formats, such as tables, graphs, messages, and dashboards. When outputting the judgment result and explanation information, the user terminal (120) may display metadata (e.g., time of reception, data source, judgment path, etc.) to ensure the data's freshness, reliability, and traceability.

[0099] In one embodiment, the electronic device (130) can manage data integrity, security, transmission delay, etc., when transmitting judgment results and explanatory information. The electronic device (130) can apply security protocols such as encryption, authentication, and integrity verification when transmitting data. The electronic device (130) can detect packet loss, transmission delay, network failure, etc., during the data transmission process and perform retransmission or error notification. The electronic device (130) can record data transmission history and results in a system log. The electronic device (130) can ensure the reliability, consistency, and traceability of data transmission.

[0100] In at least one embodiment of the present disclosure, the electronic device (130) can detect and log errors that may occur during the process of data acquisition, processing, judgment, generation of descriptive information, and output. The electronic device (130) can detect the validity of input data, consistency of processing results, and whether exceptions occur in real time at each step. When an error occurs, the electronic device (130) can record an error code, time of occurrence, related data, and cause analysis information in a system log. The electronic device (130) can statistically analyze the frequency, type, and scope of impact of errors.

[0101] In one embodiment, the electronic device (130) can monitor the operating status, data flow, and exception situations of each module in real time. The electronic device (130) can periodically collect status information of each module and display it on a dashboard or monitoring system. The electronic device (130) can detect abnormal signs in real time, such as bottlenecks, delays, or omissions in the data flow. The electronic device (130) can send a notification to an administrator or operator when an exception occurs. Based on the monitoring results, the electronic device (130) can perform system performance improvement or failure prevention measures.

[0102] In at least one embodiment of the present disclosure, the electronic device (130) can perform appropriate response procedures, such as reprocessing, notification, and failure reporting, when an error occurs. The electronic device (130) can execute automatic reprocessing logic when an error occurs. The electronic device (130) can notify an administrator or operator of the fact that an error has occurred in real time. The electronic device (130) can automatically generate and record a failure report when a failure occurs. The electronic device (130) can apply various response measures, such as reprocessing, selection of an alternative path, and data recovery, depending on the type of error.

[0103] In at least one embodiment of the present disclosure, the electronic device (130) may record all major events to ensure the reliability and traceability of data and information flow within the system. The electronic device (130) may record major events at each stage, such as data acquisition, processing, judgment, generation of explanatory information, and output, in a system log. When recording events, the electronic device (130) may store metadata such as event type, time of occurrence, related data, and processing result together. The electronic device (130) may verify the reproducibility and traceability of data flow based on the recorded event log. The electronic device (130) may ensure the reliability, consistency, and transparency of system operation by analyzing the event log.

[0104] FIG. 2 is a diagram illustrating the operation of each layer according to one embodiment.

[0105] Referring to FIG. 2, in one embodiment, an electronic device (130) can obtain reliable data through a global public verified data source (211) in an external data layer (210). The electronic device (130) can collect various types of data, such as policies, industry statistics, technology, patents, papers, and news, from the global public verified data source (211). When obtaining data, the electronic device (130) can verify the reliability, timeliness, and integrity of the data source. During the data acquisition process, the electronic device (130) can store the source, collection time, and version information of the data together as metadata. Obtaining reliable data can serve as a basis for increasing the accuracy of subsequent data processing and judgment.

[0106] In at least one embodiment of the present disclosure, the electronic device (130) can normalize different types of industrial data in the data collection and normalization layer (230) according to a predefined data structure based on an industrial domain. The electronic device (130) can collect data of different structures through a source-specific collection module (231). The electronic device (130) can convert the collected data into the same entity classification structure, data field configuration, and data schema in the normalization module (233). The electronic device (130) can automatically convert and correct discrepancies in data format, unit, and name during the normalization process. The electronic device (130) can standardize data quality so that comparison, combination, and statistical analysis between data are possible through the normalized data. Data normalization enables integrated analysis of heterogeneous data and reliable judgment.

[0107] In one embodiment, the electronic device (130) can ensure data quality and judgment reliability by stepwise generating feature vectors, KPIs, and judgment results based on normalized data in the analysis layer (250). The electronic device (130) can generate a multidimensional feature vector by sorting the normalized data by time in the feature vector generation module (251) or by combining data from multiple time points. The electronic device (130) can calculate key performance indicators by combining or weighting elements of the feature vector in the KPI calculation engine (253). The electronic device (130) can generate discrete state judgment results by comparing the KPI with reference values, threshold values, and comparison targets in the rule-based judgment engine (255). In the stepwise calculation process, the input, processing, and output of each step are clearly distinguished, thereby minimizing the possibility of data quality degradation and errors. The stepwise processing results can be logged and verified, thereby increasing judgment reliability.

[0108] In at least one embodiment of the present disclosure, the electronic device (130) structurally separates the data analysis step and the explanation information generation step, thereby eliminating the non-determinacy of free-generation-based explanations and reliably supporting corporate decision-making. After producing a judgment result in the analysis layer (250), the electronic device (130) may separately generate explanation information regarding the judgment basis and judgment path in the explanation information generation module (271) of the output layer (270). When generating explanation information, the electronic device (130) may include identification information such as the data source used for the judgment, feature vector, KPI, and application conditions. The electronic device (130) can implement the explanation information generation logic based on fixed rules or templates to always generate the same explanation result for the same input. Structural separation of explanation information can ensure the reproducibility and traceability of the judgment result. Consistency of explanation information can increase the reliability of corporate decision-making and provide transparent grounds to the user.

[0109] In at least one embodiment of the present disclosure, the electronic device (130) can perform inter-module interaction through a standardized interface that exchanges data and control signals between each layer. The electronic device (130) can pre-define the format, protocol, and message structure of input and output data for each layer. The electronic device (130) can use standardized communication methods, such as APIs, message queues, and file transfers, when transmitting data between layers. The electronic device (130) can maintain compatibility during module upgrades by versioning the interface specifications. The standardized interface can reduce the possibility of data loss, tampering, or errors occurring between modules. Interface standardization can ensure a consistent data flow even when interacting with and expanding external systems.

[0110] In one embodiment, the electronic device (130) can transmit data acquired from the external data layer (210) to the data collection and normalization layer (230). The electronic device (130) can transmit raw data collected from the external data layer (210) to a source-specific collection module (231). When transmitting data, the electronic device (130) may include metadata such as the source of the data, the time of collection, and the data type. The data transmission process can be implemented in real-time or batch processing mode.

[0111] In at least one embodiment of the present disclosure, the electronic device (130) may transmit normalized data to the analysis layer (250). The electronic device (130) may input data converted in the normalization module (233) to the feature vector generation module (251). When transmitting data, the electronic device (130) may include the normalization status of the data, conversion history, error information, etc. The normalized data may be used directly for analysis in the analysis layer without further processing.

[0112] In at least one embodiment of the present disclosure, the electronic device (130) can transmit a judgment result and related information calculated in the analysis layer (250) to the output layer (270). The electronic device (130) can transmit outputs of the analysis layer, such as KPIs, judgment results, and analysis logs, to the explanation information generation module (271). The electronic device (130) may include detailed information, such as the basis for judgment, application rules, and comparison conditions, along with the judgment result. Data transmission between the analysis layer and the output layer can be implemented in various ways, such as an internal bus, API calls, and message queues.

[0113] In at least one embodiment of the present disclosure, the electronic device (130) can transmit descriptive information and judgment results generated at the output layer (270) to the user terminal (120) via a network. The electronic device (130) can convert the descriptive information and judgment results at the output module (273) into various formats such as network packets, API responses, and files. The electronic device (130) can transmit data to the user terminal (120) using standard protocols such as TCP / IP, HTTP, and WebSocket. The electronic device (130) can apply security procedures such as encryption, authentication, and integrity verification during data transmission. The electronic device (130) can detect packet loss, transmission delay, network failure, etc., during the data transmission process and perform retransmission or error notification. The electronic device (130) can record transmission history and results in a system log to ensure traceability of the data flow.

[0114] In at least one embodiment of the present disclosure, the external data layer (210) may provide a source of industrial data processed in the present invention. The external data layer (210) may operate as an initial entry point for acquiring industrial data. The external data layer (210) may be responsible for connecting with a global public verified data source (211) as a source of industrial data. The external data layer (210) may collect industrial data through linkage with various public data sources such as policies, industrial statistics, technology, patents, papers, and news. The external data layer (210) may support different access methods (API, file download, web crawling, etc.) depending on the data source. The external data layer (210) may transmit the acquired industrial data to a subsequent data collection and normalization layer (230). The external data layer (210) may manage metadata such as the time of data acquisition, source identifier, and data type. By clearly distinguishing the source of industrial data, the external data layer (210) may ensure the traceability and reproducibility of the data.

[0115] In at least one embodiment of the present disclosure, the external data layer (210) may include a global public verification data source (211). The external data layer (210) may provide a direct connection interface with the global public verification data source (211). The external data layer (210) may collect data from the global public verification data source (211) on a periodic or event-based basis. The external data layer (210) may verify the reliability and timeliness of the data provided by the global public verification data source (211). The external data layer (210) may transmit the data obtained from the global public verification data source (211) to the data collection and normalization layer (230). The external data layer (210) may record the history of interaction with the global public verification data source (211) to ensure the transparency of the data acquisition path.

[0116] In one embodiment, the global public verification data source (211) may include multilayered public data having different types and structures. The global public verification data source (211) may include multiple data types, such as policy data, industry statistics data, technology data, patent data, paper data, and news data. The global public verification data source (211) may have different data structures (e.g., tables, JSON, XML, text, etc.) for each data type. The global public verification data source (211) may support a hierarchical structure of data (e.g., by country, by industry, by year, etc.). The global public verification data source (211) may include multilayered data fields such as data metadata, body data, and reference data. Due to the heterogeneity of data types and structures, the global public verification data source (211) may require data mapping and transformation in a subsequent normalization step. The global public verification data source (211) may support comprehensive analysis of industry data through the integration of multilayered public data.

[0117] In one embodiment, the external data layer (210) may serve as a linkage point with an upper system and provide an interface for acquiring and transmitting industrial data. The external data layer (210) may provide a standardized API or message queue interface for transmitting and receiving data with an upper system (e.g., electronic device (130)). The external data layer (210) may transmit data in real-time or batch mode according to a data request from the upper system. When linking with the upper system, the external data layer (210) may pre-define data formats, transmission protocols, authentication methods, etc. The external data layer (210) may log the transmission history and status of data transmitted to the upper system to ensure traceability of the data flow. The external data layer (210) may support various linkage scenarios, such as data acquisition, transmission, and error handling, through the linkage point with the upper system.

[0118] In one embodiment, the global public verified data source (211) may refer to a publicly accessible and verified data source, such as policies, industry statistics, technical information, patents, papers, news, etc. The global public verified data source (211) may include data sources provided by credible institutions such as governments, international organizations, industry associations, patent offices, academic institutions, and media outlets. The global public verified data source (211) may support various access methods for each data, such as public APIs, data portals, open databases, and websites. The global public verified data source (211) may comply with policy requirements, such as authentication, authorization, and conditions for data usage, when accessing data. The global public verified data source (211) may provide data that has undergone official distribution channels and verification procedures to ensure data reliability. The global public verified data source (211) may provide data that meets quality standards, such as data timeliness, completeness, and consistency.

[0119] In one embodiment, the global public verification data source (211) may include multilayered public data having different types and structures. The global public verification data source (211) may include multiple data types, such as policy data, industry statistics data, technology data, patent data, paper data, and news data. The global public verification data source (211) may have different data structures (e.g., tables, JSON, XML, text, etc.) for each data type. The global public verification data source (211) may support a hierarchical structure of data (e.g., by country, by industry, by year, etc.). The global public verification data source (211) may include multilayered data fields such as data metadata, body data, and reference data. Due to the heterogeneity of data types and structures, the global public verification data source (211) may require data mapping and transformation in a subsequent normalization step. The global public verification data source (211) may support comprehensive analysis of industry data through the integration of multilayered public data.

[0120] In at least one embodiment of the present disclosure, a global public verified data source (211) may provide the source of industrial data processed in the present invention. As the initial provider of industrial data, the global public verified data source (211) may guarantee the reliability and source of the data. The global public verified data source (211) may also provide metadata such as the time of acquisition of industrial data, data type, and source identifier. By clearly distinguishing the source of industrial data, the global public verified data source (211) may ensure the traceability and reproducibility of the data. The global public verified data source (211) may support the use of the source information of industrial data as a basis for judgment in subsequent data processing, analysis, and explanatory information generation stages. The global public verified data source (211) may contribute to ensuring the transparency and reliability of the data flow within the system through the source information of industrial data.

[0121] In at least one embodiment of the present disclosure, the data collection and normalization layer (230) may include a source-specific collection module (231) and a normalization module (233). The data collection and normalization layer (230) may receive industrial data transmitted from an external data layer (210). The data collection and normalization layer (230) may acquire data from a plurality of global public verification data sources (211) through the source-specific collection module (231). The data collection and normalization layer (230) may transmit the acquired data to the normalization module (233). The data collection and normalization layer (230) may perform data structuring and standardization processing in the normalization module (233). The data collection and normalization layer (230) may transmit the structured data to the analysis layer (250). The data collection and normalization layer (230) may record the processing status and logs of the data at each stage within the pipeline. The data collection and normalization layer (230) may include automation and error recovery functions of the pipeline.

[0122] In one embodiment, the source-specific collection module (231) can acquire different types of industrial data from global public verified data sources (211) of the external data layer (210). The source-specific collection module (231) can collect various types of industrial data, such as policy data, industrial statistics data, technology data, patent data, paper data, and news data. The source-specific collection module (231) can recognize different data structures and formats for each data source. The source-specific collection module (231) can manage authentication, access rights, and data request methods (API, file download, web crawling, etc.) for each data source. The source-specific collection module (231) can record metadata such as the time of data acquisition, source identifier, and data type. The source-specific collection module (231) can transmit the acquired data to the normalization module (233).

[0123] In one embodiment, the normalization module (233) can normalize industrial data obtained by the source-specific collection module (231) according to a predefined data structure based on the industrial domain. The normalization module (233) can refer to an entity classification structure, data field configuration, and data schema predefined for each industrial domain. The normalization module (233) can perform normalization procedures such as field mapping of input source data, data type conversion, missing value processing, and unit conversion. The normalization module (233) can structure data to enable subsequent analysis and comparison by converting heterogeneous data structures into a unified format. The normalization module (233) can record logs and conversion history regarding the results of the normalization process. The normalization module (233) can transmit the normalized data to the analysis layer (250).

[0124] In one embodiment, the normalization module (233) can structure the data so that comparison and combination between data are possible in subsequent steps by converting industrial data obtained from different public data sources to match the same entity classification structure, data field configuration, or data schema. The normalization module (233) can map entity names, field names, and data types that differ by data source to a standardized entity classification structure and data fields. The normalization module (233) can verify data quality issues such as missing data fields, outliers, and duplicates, and can correct or filter them according to predefined quality standards. When converting the data schema, the normalization module (233) can verify the integrity of the data (e.g., existence of required fields, referential integrity, etc.). The normalization module (233) can record the results of the data quality verification as a quality log and separate data that does not meet quality standards into a separate storage. By ensuring the quality and integrity of the structured data, the normalization module (233) can guarantee reliable data utilization in the analysis layer (250).

[0125] In one embodiment, the data collection and normalization layer (230) may perform a structuring and standardization process of the acquired data to ensure the integrity and quality of the data. The data collection and normalization layer (230) may perform data quality verification and integrity checks in parallel from the data acquisition stage. The data collection and normalization layer (230) may increase the accuracy of comparison and combination between data by applying a standardized data structure. The data collection and normalization layer (230) may trigger an automatic notification or manual review process when data errors, omissions, or outliers occur according to a quality management policy. The data collection and normalization layer (230) may ensure transparency and reproducibility of data processing by leaving the history and results of the quality management policy as a system log.

[0126] In one embodiment, the data collection and normalization layer (230) may be designed to enable large-scale data processing and simultaneous collection and normalization of various data sources by extending multiple source-specific collection modules (231) and normalization modules (233) in parallel. The data collection and normalization layer (230) may adopt a module-specific independent execution structure so that each source-specific collection module (231) and normalization module (233) can operate in parallel. The data collection and normalization layer (230) may apply a container-based or microservices architecture to dynamically expand or contract module instances as needed. The data collection and normalization layer (230) may support real-time or batch processing of large-scale data by utilizing distributed processing technologies such as distributed queues and message brokers. Due to the scalability, the data collection and normalization layer (230) can flexibly respond to the addition of data sources, an increase in data volume, and changes in processing load. The data collection and normalization layer (230) may improve the processing performance and availability of the system through a scale-out design.

[0127] In at least one embodiment of the present disclosure, the source-specific collection module (231) can acquire different types of industrial data from a global public verification data source (211) of an external data layer (210). The source-specific collection module (231) can manage API endpoints, authentication tokens, data request parameters, etc., for connection with the global public verification data source (211). The source-specific collection module (231) can automatically recognize and parse different data formats (JSON, XML, CSV, HTML, etc.) for each data source. The source-specific collection module (231) can record metadata such as the time of data acquisition, the response code for each source, and the data size. The source-specific collection module (231) can transmit the acquired data to the normalization module (233).

[0128] In one embodiment, the source-specific collection module (231) can collect industrial data in various ways, such as APIs, web crawling, and data pipelines. The source-specific collection module (231) can process real-time data requests and responses through RESTful APIs. The source-specific collection module (231) can automatically extract unstructured data (e.g., web pages, news articles, etc.) through a web crawler. The source-specific collection module (231) can collect large files or streaming data in batches through a data pipeline. The source-specific collection module (231) may include error handling, retry, and data duplication prevention logic for each collection method. Due to the diversification of collection methods, the source-specific collection module (231) can apply an optimal collection strategy tailored to the characteristics of each data source.

[0129] In at least one embodiment of the present disclosure, the source-specific collection module (231) may include a verification procedure to verify the integrity and reliability of the acquired data. The source-specific collection module (231) may generate and compare integrity verification information, such as hash values, checksums, and data length, when collecting data. The source-specific collection module (231) may evaluate the collected data by applying reliability evaluation criteria for each data source (e.g., data recency, source reliability, error rate, etc.). If the integrity or reliability verification fails, the source-specific collection module (231) may separate the data into a separate storage or trigger re-collection. The source-specific collection module (231) may record the verification results and history as a system log to ensure transparency in data processing.

[0130] In at least one embodiment of the present disclosure, the normalization module (233) can normalize industrial data obtained by the source-specific collection module (231) according to a predefined data structure based on an industrial domain. The normalization module (233) can ensure data quality by accepting only data that has passed integrity verification as input. The normalization module (233) can perform additional quality verification (e.g., existence of required fields, data type matching, etc.) during the data normalization process. The normalization module (233) can branch the processing path of the data according to the verification results.

[0131] In at least one embodiment of the present disclosure, the source-specific collection module (231) manages the collection cycle and schedule for each data source to acquire data in real-time or batch mode. The source-specific collection module (231) can set a collection schedule for each data source on a periodic basis (e.g., daily, weekly, monthly) or an event basis (e.g., when data changes). The source-specific collection module (231) can automatically manage the collection timing and frequency of each data source through a scheduler. The source-specific collection module (231) may include a data change detection and immediate collection trigger function during real-time collection. The source-specific collection module (231) may support batch processing of large volumes of data and error recovery functions during batch collection. The source-specific collection module (231) can record the collection cycle and schedule history to ensure traceability of data acquisition.

[0132] In at least one embodiment of the present disclosure, the normalization module (233) can normalize industrial data obtained by the source-specific collection module (231) according to a predefined data structure based on an industrial domain. The normalization module (233) can map entities of input data to standard entities by referring to an entity classification structure predefined for each industrial domain. The normalization module (233) can select a suitable entity classification structure by analyzing metadata such as domain identifiers, industrial codes, and categories of the input data. The normalization module (233) can record the entity mapping results in a data field to enable consistent entity referencing in subsequent analysis steps.

[0133] In at least one embodiment of the present disclosure, the normalization module (233) can perform consistent mapping of data by applying an entity classification structure by industry domain. The normalization module (233) can standardize different entity names by data source based on the entity classification structure by industry domain. The normalization module (233) can detect cases of mapping failure or inconsistency during the entity mapping process and trigger a separate review or correction process. The normalization module (233) can record the entity mapping history and results as a system log.

[0134] In one embodiment, the normalization module (233) can structure the data so that comparison and combination between data are possible in subsequent steps by converting industrial data obtained from different public data sources to match the same entity classification structure, data field configuration, or data schema. The normalization module (233) can convert the field names, data types, units, formats, etc., of the input data to match the standard data field configuration. The normalization module (233) can utilize a data field mapping table to match different field names from different sources to standard field names. The normalization module (233) can perform data type conversion (e.g., string → number, date format unification, etc.) and unit conversion (e.g., USD → KRW, kg → ton, etc.). The normalization module (233) can detect errors such as missing essential fields, data type mismatches, and outliers during the schema conversion process and correct or filter them according to a predefined processing policy. The normalization module (233) can record the schema version of the converted data to ensure data history management and reproducibility.

[0135] In at least one embodiment of the present disclosure, the normalization module (233) can unify the structure of heterogeneous data through data field configuration and schema conversion procedures. The normalization module (233) can achieve consistent structural unification for a large amount of data by automating the data field configuration and schema conversion procedures. The normalization module (233) can verify the schema conversion results and trigger a notification or manual review in the event of conversion errors or missing data. The normalization module (233) can record the schema conversion history and results as a system log.

[0136] In at least one embodiment of the present disclosure, the normalization module (233) can structure the data so that comparison and combination between data are possible in subsequent steps by converting industrial data obtained from different public data sources to match the same entity classification structure, data field configuration, or data schema. The normalization module (233) can align the data to enable direct comparison and combination between multiple data sources by applying a standardized entity classification structure and data field configuration. When combining data, the normalization module (233) can perform data merging or joining based on a common key (e.g., industry code, corporate identifier, etc.). The normalization module (233) can verify quality issues such as data consistency, redundancy, and missing values ​​during the data comparison and combination process. By ensuring the possibility of comparison and combination, the normalization module (233) can support integrated analysis of multiple data sources at the analysis layer (250).

[0137] In at least one embodiment of the present disclosure, the normalization module (233) can ensure the possibility of comparison and combination between various data sources through a standardization processing process. The normalization module (233) can record the results of the standardization processing as metadata to track the source of the data, transformation history, combination criteria, etc. The normalization module (233) can ensure the reproducibility and transparency of data processing through the automation and history management of the standardization processing.

[0138] In one embodiment, the analysis layer (250) may include a feature vector generation module (251), a KPI calculation engine (253), and a rule-based judgment engine (255). The analysis layer (250) may receive normalized industrial data from the feature vector generation module (251). The analysis layer (250) may process the input data step by step, separating the functions of feature vector generation, KPI calculation, and status judgment into modules. The analysis layer (250) may provide an internal interface for data transfer and linkage processing between each module. The analysis layer (250) may record the processing results of each module and use them as input for subsequent steps. The analysis layer (250) may store processing logs and result data for each module to ensure the reproducibility and traceability of the judgment results.

[0139] In one embodiment, the analysis layer (250) receives normalized industry data and can perform a series of analysis processes to determine the state of the industry, technology, or market in the corporate decision-making process. The analysis layer (250) can receive normalized industry data from the data collection and normalization layer (230). The analysis layer (250) can transmit the input data to a feature vector generation module (251) to generate a multidimensional vector for state representation. The analysis layer (250) can transmit the generated feature vector to a KPI calculation engine (253) to calculate key performance indicators. The analysis layer (250) can transmit the indicators calculated by the KPI calculation engine (253) to a rule-based judgment engine (255) to generate state judgment results. The analysis layer (250) can store the processing results for each step in an internal data structure.

[0140] In one embodiment, the analysis layer (250) can derive a final judgment result from input data by linking each step of feature vector generation, KPI calculation, and state judgment. The analysis layer (250) can automatically transmit the feature vector generated by the feature vector generation module (251) to the KPI calculation engine (253). The analysis layer (250) can link the KPI calculated by the KPI calculation engine (253) to the Rule-based judgment engine (255) to use as an input value for state judgment. The analysis layer (250) can record the processing results of each step along with metadata to enable tracking of the judgment path. The analysis layer (250) can transmit the final judgment result to the output layer (270).

[0141] In at least one embodiment of the present disclosure, the analysis layer (250) may be designed with consideration for the application and scalability of various analysis algorithms. The analysis layer (250) may apply a modular structure to allow analysis algorithms to be added or replaced in the form of plugins for each module. The analysis layer (250) may support the application of new analysis logic through version management of analysis algorithms and interface standardization. The analysis layer (250) may selectively apply different analysis algorithms depending on the domain and purpose.

[0142] In at least one embodiment of the present disclosure, the analysis layer (250) may apply strategies for improving processing performance and parallelization when processing large volumes of data. When large volumes of data are input, the analysis layer (250) may improve processing speed through data partitioning and parallel computation. The analysis layer (250) may optimize data processing performance by integrating with a distributed processing framework (e.g., Spark, Hadoop, etc.). The analysis layer (250) may apply performance optimization techniques such as caching, memory management, and asynchronous processing.

[0143] In one embodiment, the analysis layer (250) can transmit the feature vector generated by the feature vector generation module (251) to the KPI calculation engine (253). The analysis layer (250) can convert the multidimensional vector data generated by the feature vector generation module (251) to fit the input format of the KPI calculation engine (253). The analysis layer (250) can identify and transmit the fields required for KPI calculation for each element of the feature vector.

[0144] In at least one embodiment of the present disclosure, the KPI calculation engine (253) can calculate key performance indicators based on input feature vectors. The KPI calculation engine (253) can calculate KPIs by combining or weighting values ​​for each dimension of the feature vector according to predefined calculation rules. The KPI calculation engine (253) can store the KPI calculation results in an internal data structure.

[0145] In at least one embodiment of the present disclosure, the rule-based judgment engine (255) may receive key performance indicators calculated by the KPI calculation engine (253) and generate a state judgment result. The rule-based judgment engine (255) may perform a state judgment by comparing the KPI calculation result with a reference value, a threshold value, or an external comparison target. The rule-based judgment engine (255) may generate the judgment result as a data structure composed of a code, a state message, a basis for judgment, etc.

[0146] In at least one embodiment of the present disclosure, the analysis layer (250) may support modularization and interface standardization to enable the application of various analysis algorithms. The analysis layer (250) may define API or interface specifications for each analysis module to facilitate the plugging in and replacement of new algorithms. The analysis layer (250) may standardize the input / output data formats of analysis algorithms to enhance inter-module connectivity. The analysis layer (250) may provide version management and a test environment for each algorithm.

[0147] In one embodiment, the analysis layer (250) may be designed to be expandable without changing the existing structure when new analysis logic or indicators are added. When new analysis logic is added, the analysis layer (250) may insert new functions in the form of plugins while maintaining the interface of the existing module. The analysis layer (250) may apply a structure that minimizes dependencies between modules to ensure scalability. When new indicators are added, the analysis layer (250) may maintain the existing data flow and result structure.

[0148] In one embodiment, the analysis layer (250) can apply performance optimization strategies such as parallel computation, distributed processing, and caching when processing large volumes of data. The analysis layer (250) can perform parallel computation by dividing input data into multiple partitions. The analysis layer (250) can distribute the computational load by allocating data blocks to each node in a distributed processing environment. The analysis layer (250) can improve processing speed during repetitive computation by caching intermediate results. The analysis layer (250) can identify bottlenecks and increase overall processing efficiency by applying asynchronous processing or a pipeline structure.

[0149] In at least one embodiment of the present disclosure, the analysis layer (250) may configure a processing structure to operate efficiently in both real-time analysis and batch analysis environments. When performing real-time analysis, the analysis layer (250) may utilize event-based triggers and stream processing engines. When performing batch analysis, the analysis layer (250) may apply schedulers and large-scale data batch processing functions. The analysis layer (250) may provide a data structure capable of integrated management of real-time and batch processing results.

[0150] In at least one embodiment of the present disclosure, the feature vector generation module (251) can generate a feature vector to represent the state of at least one of the industry, technology, or market to which the company belongs during the decision-making process of the company, based on normalized industry data. The feature vector generation module (251) can analyze the input normalized industry data based on time, domain, or statistics. The feature vector generation module (251) can produce a multidimensional vector that numerically represents the state of the subject of analysis. The feature vector generation module (251) can transmit the generated feature vector to the KPI calculation engine (253). The feature vector generation module (251) can record the data source, time point, and domain information used in the vector generation process together as metadata.

[0151] In one embodiment, the feature vector generation module (251) can generate a multidimensional vector that numerically represents the state of a judgment target from input normalized industrial data. The feature vector generation module (251) can map key indicators of each industrial data to individual dimensions of the vector. The feature vector generation module (251) may include time series data, statistical summary values, combined indicators, etc., in each dimension of the vector. The feature vector generation module (251) can record the structure of the generated vector and the meaning of each element as metadata.

[0152] In one embodiment, the feature vector generation module (251) can transmit the generated feature vector to the KPI calculation engine (253). The feature vector generation module (251) can convert the vector data to fit the input format of the KPI calculation engine (253). When transmitting the vector, the feature vector generation module (251) can verify data integrity and consistency.

[0153] In at least one embodiment of the present disclosure, the feature vector generation module (251) may generate data in the form of a multidimensional vector representing the state of a judgment target by aligning normalized industrial data based on time, reflecting changes in data according to multiple time points, or combining or summarizing multiple industrial data. The feature vector generation module (251) may perform time-series alignment based on the timestamps of the input data. The feature vector generation module (251) may include a rate of change or trend indicator in the vector by calculating the difference in data from multiple time points. The feature vector generation module (251) may combine multiple industrial domain data to generate a vector reflecting the cross-domain state. The feature vector generation module (251) may calculate statistical summary values ​​(e.g., mean, standard deviation, maximum / minimum value, etc.) as vector elements. The feature vector generation module (251) may verify and correct data omissions, outliers, and duplicates during the combining and summarizing process.

[0154] In at least one embodiment of the present disclosure, the feature vector generation module (251) can form a multidimensional vector that includes multiple domain-specific indicators, such as industry, technology, and market, as each dimension. The feature vector generation module (251) can assign key indicators for each industry domain (e.g., production volume, sales, number of patent applications, etc.) to individual dimensions of the vector. The feature vector generation module (251) can include indicators for each technology domain (e.g., technology adoption rate, R&D investment, etc.) as separate dimensions. The feature vector generation module (251) can include indicators for each market domain (e.g., market share, growth rate, etc.) as additional dimensions.

[0155] In at least one embodiment of the present disclosure, the feature vector generation module (251) may include time series data, statistical summary values, combined indicators, etc., for each element of the vector. The feature vector generation module (251) may calculate various statistical indicators, such as time series change amounts, moving averages, and cumulative sums, for each vector element. The feature vector generation module (251) may calculate combined values ​​of multiple domain indicators (e.g., industry x technology growth rates, etc.) and include them in the vector. The feature vector generation module (251) may record units, ranges, and interpretation criteria as metadata for each element of the vector.

[0156] In one embodiment, the KPI calculation engine (253) receives a feature vector generated by the feature vector generation module (251) as input and can calculate the company's key performance indicators according to predefined calculation rules. The KPI calculation engine (253) can perform operations necessary for calculating KPIs for each element of the input feature vector. The KPI calculation engine (253) can calculate KPIs by applying predefined formulas, combination formulas, and algorithms. The KPI calculation engine (253) can transmit the KPI calculation results to the Rule-based judgment engine (255). The KPI calculation engine (253) can record the calculation rules, weights, and parameter information used in the KPI calculation process as metadata.

[0157] In one embodiment, the KPI calculation engine (253) can calculate an indicator in the form of an index to determine the state of an industry, technology, market, or company in the decision-making process of the company. The KPI calculation engine (253) can calculate various KPI items, such as an industry-specific growth index, a technology innovation index, and a market stability index. The KPI calculation engine (253) can define a calculation formula, units, and interpretation criteria for each KPI item. The KPI calculation engine (253) can express the KPI calculation results in various forms, such as codes, numerical values, and grades.

[0158] In at least one embodiment of the present disclosure, the KPI calculation engine (253) may calculate KPIs by applying a predefined formula, combination formula, or algorithm to an input feature vector. The KPI calculation engine (253) may apply various calculation rules, such as linear combinations, weighted averages, and non-linear functions. The KPI calculation engine (253) may manage the calculation rules applied to each KPI item as metadata.

[0159] In one embodiment, the KPI calculation engine (253) can transmit the KPI calculation result to the Rule-based judgment engine (255). The KPI calculation engine (253) can transmit the result to the Rule-based judgment engine (255) after verifying the integrity and consistency of the KPI calculation result. The KPI calculation engine (253) can transmit the result along with the calculation basis information.

[0160] In one embodiment, the KPI calculation engine (253) can calculate key performance indicators by combining or weighting multiple feature vector elements. The KPI calculation engine (253) can assign weights to each feature vector element and reflect them in the KPI calculation according to importance. The KPI calculation engine (253) can set combination parameters for multiple elements and apply them to the KPI calculation formula. The KPI calculation engine (253) can record the weight and parameter setting values ​​as metadata.

[0161] In at least one embodiment of the present disclosure, the KPI calculation engine (253) may set weights, combination parameters, normalization coefficients, etc. for each KPI item. The KPI calculation engine (253) may standardize the unit and range of the calculated value by applying a normalization coefficient for each KPI item. The KPI calculation engine (253) may manage the change history of the weights, parameters, and normalization coefficients.

[0162] In one embodiment, the rule-based judgment engine (255) can generate a judgment result regarding the state of a judgment target by comparing the key performance indicators calculated by the KPI calculation engine (253) with specific conditions, e.g., a reference value, a threshold value, or a comparison target. The rule-based judgment engine (255) can receive the KPI calculation result and perform a comparison operation with a predefined reference value or threshold value. The rule-based judgment engine (255) can assist in the state judgment by comparing the KPI with an external comparison target (e.g., industry average, competitor indicator, etc.). The rule-based judgment engine (255) can generate the judgment result as a data structure consisting of a code, a status message, and the basis for judgment. The rule-based judgment engine (255) can transmit the judgment result to an output layer (270).

[0163] In at least one embodiment of the present disclosure, the rule-based judgment engine (255) may perform a state judgment on the KPI calculation result by referring to a reference value, a threshold value, or an external comparison target. The rule-based judgment engine (255) may pre-define a reference value and a threshold value for each KPI item and apply them to the judgment logic. The rule-based judgment engine (255) may periodically update data of an external comparison target (e.g., industry average, market benchmark, etc.) and use it as a judgment standard. The rule-based judgment engine (255) may record the change history of the reference value, threshold value, and comparison target as metadata.

[0164] In one embodiment, the judgment result may be expressed as one of a plurality of discrete states, such as a stable state, a dangerous state, a growth state, or a slowdown state. The rule-based judgment engine (255) can determine whether the KPI calculation result falls within each threshold range and classify it into a state corresponding to that range (e.g., growth, stable, dangerous, slowdown, etc.). The rule-based judgment engine (255) can manage multiple state classification criteria in a priority or hierarchical structure. The rule-based judgment engine (255) may include exception handling logic for edge cases (e.g., threshold boundary values, missing data, etc.).

[0165] In one embodiment, the rule-based judgment engine (255) may generate a judgment result as a data structure composed of a code, a status message, and a basis for judgment. The rule-based judgment engine (255) may include a judgment result code (e.g., 0=stable, 1=growth, 2=risk, etc.) in the data structure. The rule-based judgment engine (255) may include a status message (e.g., "growth status," "risk status," etc.) in the judgment result. The rule-based judgment engine (255) may include a basis for judgment (e.g., used KPI, reference value, comparison target, etc.) in the data structure.

[0166] In at least one embodiment of the present disclosure, the rule-based judgment engine (255) may transmit a judgment result to an output layer (270). The rule-based judgment engine (255) may transmit a judgment path, basis, and metadata to the output layer (270) along with the judgment result data. The rule-based judgment engine (255) may transmit a log of the judgment process along with the judgment result to ensure reproducibility and traceability of the judgment result.

[0167] In one embodiment, the output layer (270) may include an explanation information generation module (271) and an output module (273). The output layer (270) may be configured with the explanation information generation module (271) and the output module (273) physically separated. The output layer (270) may transmit explanation information generated by the explanation information generation module (271) to the output module (273). The output layer (270) may provide an internal interface for data transmission between each module. The output layer (270) may record processing results for each module and use them as input for a subsequent step.

[0168] The output layer (270) can perform the role of transmitting the judgment result and related explanatory information generated by the analysis layer (250) to the user terminal (120). The output layer (270) can receive the judgment result and explanatory information from the analysis layer (250). The output layer (270) can store the received judgment result and explanatory information in an internal data structure. The output layer (270) can transmit the stored judgment result and explanatory information to the user terminal (120). The output layer (270) can apply information structuring to ensure transparency and reproducibility of the judgment result and customized display strategies for each user terminal. The output layer (270) can convert the judgment result and explanatory information into a structured format. The output layer (270) can generate an information structure including the judgment basis, judgment path, data source, feature vector, key performance indicator, and application conditions. The output layer (270) can selectively apply a display method and data format depending on the type of user terminal (120). The output layer (270) can ensure transparency and reproducibility of the judgment results through information structuring and customized display strategies.

[0169] In at least one embodiment of the present disclosure, the output layer (270) may be intended to transmit a judgment result and explanatory information generated by an electronic device (130) to a user terminal (120). The output layer (270) may receive the judgment result and explanatory information from the electronic device (130). The output layer (270) may store the received judgment result and explanatory information in an internal data structure. The output layer (270) may transmit the stored judgment result and explanatory information to the user terminal (120) via a network. The output layer (270) may ensure the reproducibility and traceability of the judgment result by providing explanatory information regarding the judgment basis and judgment path together with the judgment result. The output layer (270) may generate the judgment basis and judgment path information corresponding to the judgment result in an explanatory information generation module (271). The output layer (270) may transmit the generated explanatory information together with the judgment result to the user terminal (120). The output layer (270) can ensure the reproducibility and traceability of the judgment result by including judgment basis and judgment path information. The output layer (270) can adaptively display the judgment result and explanation information visually or in other ways depending on the type of user terminal (120).

[0170] The output layer (270) can identify the device type of the user terminal (120) (e.g., PC, tablet, smartphone, web dashboard, etc.). The output layer (270) can determine the display format of the judgment result and explanatory information based on the identified terminal type. The output layer (270) can output information in a visual display (e.g., graph, chart, text) or in other ways such as voice, notifications, etc. The output layer (270) can enhance the user experience by applying a terminal-specific adaptive display strategy.

[0171] In at least one embodiment of the present disclosure, the output layer (270) can transmit a judgment result and explanation information to a user terminal (120) through an explanation information generation module (271) and an output module (273). The output layer (270) can transmit explanation information generated by the explanation information generation module (271) to the output module (273). The output layer (270) can transmit explanation information and judgment result to a network through the output module (273). The output layer (270) can support the user terminal (120) receiving the judgment result and explanation information in real time.

[0172] In at least one embodiment of the present disclosure, the output layer (270) can ensure the reproducibility and traceability of the judgment result by providing not only the judgment result but also the judgment basis and the judgment path. The output layer (270) can generate judgment basis and judgment path information corresponding to the judgment result in an explanation information generation module (271). The output layer (270) can include the judgment basis and judgment path information as structured data together with the judgment result. The output layer (270) can include metadata such as data sources used in the judgment, feature vectors, key performance indicators, and application conditions in the explanation information. By including the judgment basis and path information, the output layer (270) can ensure the reproducibility and transparency of the judgment result.

[0173] In one embodiment, the output layer (270) may adaptively display judgment results and explanatory information visually or in other ways depending on the type of user terminal (120). The output layer (270) may identify the device type of the user terminal (120). The output layer (270) may select a display format (e.g., graph, table, text, notification, etc.) depending on the terminal type. The output layer (270) may increase the efficiency of information delivery by applying a rendering method optimized for each terminal.

[0174] In one embodiment, the explanation information generation module (271) can generate explanation information regarding the judgment basis and judgment path corresponding to the judgment result based on the judgment result generated by the rule-based judgment engine (255). The explanation information generation module (271) can receive the judgment result from the rule-based judgment engine (255). The explanation information generation module (271) can extract the judgment basis corresponding to the input judgment result from data sources, feature vectors, key performance indicators, application conditions, etc. The explanation information generation module (271) can record the judgment path step by step and include it in the explanation information. The explanation information generation module (271) can generate explanation information by combining the judgment basis and judgment path information.

[0175] The explanation information generation module (271) can configure explanation information so that at least one of the data source used in the judgment, the feature vector, the key performance indicator, and the applied condition can be identified along with the judgment result. The explanation information generation module (271) can include the identifier (ID) and timestamp of the data source used in the judgment in the explanation information. The explanation information generation module (271) can record the value and structure of the feature vector in the explanation information. The explanation information generation module (271) can include the key performance indicator (KPI) and the applied condition (e.g., reference value, threshold value, etc.) in the explanation information. The explanation information generation module (271) can increase the traceability of the explanation information by tokenizing each element.

[0176] The explanation information generation module (271) can ensure the reproducibility and transparency of the judgment result by including the judgment basis and tracking information of the judgment path. The explanation information generation module (271) can record each step of the judgment basis and judgment path as metadata. The explanation information generation module (271) can include the flow of the judgment path (e.g., data acquisition, normalization, feature vector generation, KPI calculation, status determination, etc.) in the explanation information step by step. By including the judgment basis and tracking information of the path, the explanation information generation module (271) can ensure the reproducibility and transparency of the judgment result.

[0177] In one embodiment, the explanation information generation module (271) can generate explanation information regarding the judgment basis and judgment path corresponding to the judgment result based on the judgment result generated by the rule-based judgment engine (255). The explanation information generation module (271) can extract the judgment basis corresponding to the judgment result from data sources, feature vectors, KPIs, application conditions, etc. The explanation information generation module (271) can record each step of the judgment path and include it in the explanation information. The explanation information generation module (271) can generate explanation information by combining the judgment basis and judgment path information.

[0178] In at least one embodiment of the present disclosure, the description information generation module (271) may be configured to identify at least one of the data source used in the judgment, the feature vector, the key performance indicator, and the applied condition, along with the judgment result. The description information generation module (271) may include the identifier and timestamp of the data source in the description information. The description information generation module (271) may record the value and structure of the feature vector in the description information. The description information generation module (271) may include the key performance indicator (KPI) and the applied condition (e.g., reference value, threshold value, etc.) in the description information. The description information generation module (271) may tokenize each element to increase the traceability of the description information.

[0179] In one embodiment, the explanation information generation module (271) can ensure the reproducibility and transparency of the judgment result by including the judgment basis and tracking information of the judgment path. The explanation information generation module (271) can record each step of the judgment basis and judgment path as metadata. The explanation information generation module (271) can include the flow of the judgment path (e.g., data acquisition, normalization, feature vector generation, KPI calculation, state judgment, etc.) step by step in the explanation information. By including the judgment basis and tracking information of the path, the explanation information generation module (271) can ensure the reproducibility and transparency of the judgment result.

[0180] In one embodiment, the output module (273) can transmit the explanation information generated by the explanation information generation module (271) to the user terminal (120) via a network and output it. The output module (273) can receive the explanation information from the explanation information generation module (271). The output module (273) can transmit the received explanation information to the user terminal (120) via a network protocol (e.g., HTTP, WebSocket, etc.). The output module (273) can convert the data format so that the transmitted explanation information can be output at the user terminal (120).

[0181] The output module (273) can output judgment results and explanatory information in various rendering formats depending on the type of user terminal (120). The output module (273) can identify the device type of the user terminal (120). Depending on the terminal type, the output module (273) can convert judgment results and explanatory information into various rendering formats such as graphs, tables, text, and notifications. The output module (273) can increase the efficiency of information delivery by applying a rendering method optimized for each terminal. The output module (273) can support real-time information output based on the user terminal (120), such as real-time notifications or dashboard updates. The output module (273) can immediately display judgment results and explanatory information on the user terminal (120) through a real-time notification function. The output module (273) can periodically update judgment results and explanatory information through a dashboard update function. The output module (273) can provide the latest status information to the user through a real-time information output function.

[0182] In one embodiment, the output module (273) can transmit the description information generated by the description information generation module (271) to the user terminal (120) via a network and output it. The output module (273) can transmit the description information to the user terminal (120) using a network protocol (e.g., HTTP, WebSocket, etc.). The output module (273) can convert the description information into a standardized data format (e.g., JSON, XML, etc.) by applying an API response structure. The output module (273) can convert the data format so that the transmitted description information can be output at the user terminal (120).

[0183] In at least one embodiment of the present disclosure, the output module (273) can output judgment results and explanatory information in various rendering formats depending on the type of user terminal (120). The output module (273) can identify the device type of the user terminal (120). Depending on the terminal type, the output module (273) can convert judgment results and explanatory information into various rendering formats such as graphs, tables, text, and notifications. The output module (273) can increase the efficiency of information transmission by applying a rendering method optimized for each terminal.

[0184] In at least one embodiment of the present disclosure, the output module (273) may support real-time information output based on the user terminal (120), such as real-time notifications or dashboard updates. The output module (273) may immediately display judgment results and explanatory information on the user terminal (120) through a real-time notification function. The output module (273) may periodically update judgment results and explanatory information through a dashboard update function. The output module (273) may provide the latest status information to the user through a real-time information output function.

[0185] In at least one embodiment of the present disclosure, in one embodiment, the electronic device (130) may configure the inter-layer data flow so that data can flow sequentially between the external data layer (210), the data collection and normalization layer (230), the analysis layer (250), and the output layer (270). The electronic device (130) may pre-define the input and output data structures of each layer. The electronic device (130) may transmit data collected from the external data layer (210) to the data collection and normalization layer (230). The electronic device (130) may transmit normalized data from the data collection and normalization layer (230) to the analysis layer (250). The electronic device (130) may transmit judgment results and related data generated in the analysis layer (250) to the output layer (270). The electronic device (130) may verify the integrity and consistency of the data when transmitting data between each layer. The electronic device (130) can ensure the traceability of the steps of data processing through the inter-layer data flow. By structuring the inter-layer data flow, the electronic device (130) can ensure the reproducibility of the data processing.

[0186] In at least one embodiment of the present disclosure, the electronic device (130) can trigger the operation of a subsequent layer when the operation of each layer is completed or when a specific condition occurs. The electronic device (130) can detect a signal indicating the completion of data processing of each layer. When the electronic device (130) detects the signal indicating the completion of data processing, it can automatically activate the data processing module of the subsequent layer. The electronic device (130) can trigger the operation of the subsequent layer when a specific condition (e.g., data update, reaching a threshold, occurrence of an external event, etc.) is satisfied. When the trigger occurs, the electronic device (130) can transmit a data transfer and processing command to the subsequent layer. The electronic device (130) can improve the automation and efficiency of inter-layer data processing through trigger-based operation.

[0187] In one embodiment, the electronic device (130) can apply inter-layer data transfer and operation trigger policies by combining real-time event-based processing and periodic batch processing. When a real-time event occurs, the electronic device (130) can immediately perform data processing of the layer corresponding to the event. The electronic device (130) can set a batch schedule periodically to perform inter-layer data transfer and processing in batches at set intervals. By combining real-time events and batch processing, the electronic device (130) can simultaneously ensure speed and stability of data processing. The electronic device (130) can flexibly respond to fluctuations in data throughput through a mixed processing policy.

[0188] In one embodiment, the electronic device (130) can process immediately when an event occurs at each layer, such as data collection, normalization, analysis, and output, or process in batches at regular intervals. The electronic device (130) can transmit data to the normalization layer in real time whenever new data is received from the data collection layer. When data accumulates in the normalization layer, the electronic device (130) can transmit it to the analysis layer in batches at regular intervals. The electronic device (130) can transmit a judgment result to the output layer immediately whenever a result is generated in the analysis layer. The electronic device (130) can selectively apply event-based processing and batch processing to each layer. The electronic device (130) can increase the processing efficiency and scalability of the system through the flexible selection of processing methods for each layer.

[0189] In at least one embodiment of the present disclosure, the electronic device (130) may define an internal interface specification for inter-layer data transfer. The electronic device (130) may clearly define a data format, protocol, and message structure for data exchange between each layer. The electronic device (130) may standardize data formats, field names, data types, required / optional items, etc., according to the interface specification. The electronic device (130) may document the interface specification so that it can be referenced during system maintenance and expansion. The electronic device (130) may ensure inter-layer data compatibility and consistency through compliance with the interface specification.

[0190] In one embodiment, the electronic device (130) may provide an interface specification including input and output data structures, calling methods, error handling methods, etc. for each layer. The electronic device (130) may define the input data structure and output data structure in detail for each layer. When transferring data between layers, the electronic device (130) may selectively apply a synchronous or asynchronous calling method. When an error occurs during data transfer or processing, the electronic device (130) may include an error code and an exception handling method in the interface specification. Based on the interface specification, the electronic device (130) may increase the reliability and predictability of communication between layers.

[0191] In at least one embodiment of the present disclosure, the electronic device (130) can trigger the operation of a subsequent layer when an event occurs in which new data is collected or the data state changes in the external data layer (210) or the data collection and normalization layer (230). The electronic device (130) can detect a data collection event in the external data layer (210). The electronic device (130) can detect a data state change event (e.g., data update, error occurrence, etc.) in the data collection and normalization layer (230). When an event is detected, the electronic device (130) can automatically transmit data transfer and processing commands to a subsequent layer (e.g., normalization, analysis, output layer). The electronic device (130) can ensure real-time data processing through event-based triggering. The electronic device (130) can minimize the delay between the time of event occurrence and the operation of the subsequent layer.

[0192] In at least one embodiment of the present disclosure, the electronic device (130) may automatically start processing of the next layer when a specific condition is met in the analysis layer (250) or the output layer (270). When a judgment result is generated in the analysis layer (250), the electronic device (130) may immediately trigger data transmission and output processing in the output layer (270). The electronic device (130) may automatically start subsequent operations (e.g., user notification, logging, etc.) when a specific condition is met in the output layer (270), such as completion of data transmission or occurrence of an error. The electronic device (130) may monitor in real time whether conditions for each layer are met. The electronic device (130) may increase the automation and reliability of data processing through condition-based triggers.

[0193] In at least one embodiment of the present disclosure, the electronic device (130) can set a batch processing schedule for each layer, such as data collection, normalization, analysis, and output. The electronic device (130) can pre-set a batch processing cycle (e.g., one hour, one day, etc.) for each layer. The electronic device (130) can perform data collection, normalization, analysis, and output operations in batches at set intervals according to the batch schedule. The electronic device (130) can efficiently process a large amount of data during batch processing. The electronic device (130) can respond to changes in data throughput by dynamically adjusting the batch processing schedule.

[0194] In one embodiment, the electronic device (130) can perform inter-layer data transfer and processing according to a set time or period according to a batch schedule. The electronic device (130) can control the start time of data processing for each layer through a batch scheduler. After the batch processing is completed, the electronic device (130) can transmit data transfer and trigger signals to subsequent layers. The electronic device (130) can record the batch processing results to ensure data processing history and traceability. The electronic device (130) can optimize system resource utilization when applying the batch processing method.

[0195] In one embodiment, the electronic device (130) can detect and log errors when an error occurs during inter-layer data transfer or operation execution. The electronic device (130) can detect errors occurring during data transfer or processing in real time. The electronic device (130) can record detailed information about the detected error in a log file or database. The electronic device (130) can also store specific information such as the time of error occurrence, the type of error, and the scope of impact. The electronic device (130) can improve the reliability and maintainability of the system through error logging.

[0196] In at least one embodiment of the present disclosure, the electronic device (130) may record the error information when an error occurs and, if necessary, perform follow-up actions such as retrying or notifying an administrator. When an error occurs, the electronic device (130) may automatically execute retry logic to repeat data transmission or processing. If an error occurs repeatedly, the electronic device (130) may send a notification message or warning to an administrator. The electronic device (130) may systematically manage the history of error occurrences and follow-up actions. The electronic device (130) may increase the availability and stability of the system through error handling and logging strategies.

[0197] In one embodiment, the electronic device (130) can manage versions of data schemas when exchanging data between layers or modules. The electronic device (130) can include version information of the data schema in each data packet or message. When changing the data schema version, the electronic device (130) can verify compatibility with the previous version. The electronic device (130) can manage data structures and field definitions for each data schema version. The electronic device (130) can ensure data compatibility during system upgrades and expansions through data schema version management.

[0198] In at least one embodiment of the present disclosure, the electronic device (130) may include data schema version information to maintain backward compatibility and version history when changing the data structure. The electronic device (130) may record the history of data structure changes in a separate version management system. To maintain backward compatibility, the electronic device (130) may support both an old version data structure and a new version data structure in parallel. The electronic device (130) may enable each module to process data in a suitable data structure by referring to the data schema version information. The electronic device (130) may support the stable operation of the system by managing version history and ensuring backward compatibility.

[0199] In one embodiment, the electronic device (130) may design an API (Application Programming Interface) or message queue-based interface for data transfer between layers or modules. The electronic device (130) may selectively apply various interface methods, such as RESTful API, gRPC, and message queues, for data exchange between each layer or module. The electronic device (130) can ensure asynchronous and scalable data transfer through the API or message queue. When designing the interface, the electronic device (130) can clearly define data formats, authentication methods, and error handling methods. The electronic device (130) can reduce coupling between system modules and improve maintainability through API and message queue-based interfaces.

[0200] In one embodiment, the electronic device (130) may apply various protocols and message queue systems such as RESTful API, gRPC, AMQP, and Kafka. The electronic device (130) may support HTTP-based data exchange through RESTful API. The electronic device (130) may support high-performance binary protocol-based data exchange through gRPC. The electronic device (130) may support asynchronous processing and distributed processing of large volumes of data through message queue systems such as AMQP and Kafka. The electronic device (130) may increase the scalability and flexibility of the system by applying various protocols and message queue systems.

[0201] In at least one embodiment of the present disclosure, the electronic device (130) may apply security and access control policies, such as authentication, authorization, and data encryption, when exchanging data between layers or modules. The electronic device (130) may perform an authentication procedure to verify the identity of the user or system when exchanging data. The electronic device (130) may restrict data access by granting access rights for each layer or module. The electronic device (130) may ensure the confidentiality of data by applying an encryption algorithm when transmitting data. The electronic device (130) may prevent data leakage and unauthorized access through security and access control policies.

[0202] In one embodiment, the electronic device (130) can prevent data leakage and unauthorized access by separating access rights by layer or module. The electronic device (130) can set granular permission policies by layer or module. The electronic device (130) can block attempts to access data from layers or modules that do not have access rights. The electronic device (130) can send a warning or notification to an administrator in the event of a violation of the access control policy. The electronic device (130) can enhance the security level of the system through the separation and monitoring of access rights.

[0203] FIG. 3 is an example of a flowchart of a decision support method according to one embodiment.

[0204] In one embodiment, at step 310, the electronic device (110) may acquire industrial data from one or more public data sources. The electronic device (110) may access multiple public data sources that provide different types of public data, such as policy data, industrial statistics data, technical data, patent data, paper data, or news data. The electronic device (110) may collect industrial data from a global public verified data source (211) of an external data layer (210) via API calls, data crawling, or file downloads. The electronic device (110) may store the collected industrial data in a temporary storage or database (110). The electronic device (110) may verify the integrity and reliability of the data when collecting data. The electronic device (110) may acquire data in real-time or batch mode depending on the collection cycle or the occurrence of an event. The electronic device (110) may record metadata such as the source of the acquired industrial data, the time of collection, and the data type.

[0205] In at least one embodiment of the present disclosure, at step 320, the electronic device (110) can normalize acquired industrial data according to a predefined data structure based on an industrial domain. The electronic device (110) can transmit industrial data acquired by a source-specific collection module (231) to a normalization module (233). The normalization module (233) may refer to an entity classification structure, data field configuration, or data schema predefined by industrial domain. The normalization module (233) can convert data acquired from different public data sources into the same data structure. The normalization module (233) can perform procedures such as field mapping, data type conversion, missing value processing, and unit unification in stages. The normalization module (233) can structure the normalized data so that it can be compared and combined in a subsequent analysis stage. The normalized data can be transmitted to a database (110) or an analysis layer (250).

[0206] In one embodiment, at step 330, the electronic device (110) can generate a feature vector to represent the state of at least one of the industry, technology, or market to which the company belongs in the decision-making process, based on normalized industry data. The feature vector generation module (251) can sort the normalized industry data by time. The feature vector generation module (251) can generate a time series vector by reflecting data changes at multiple points in time. The feature vector generation module (251) can generate a multidimensional vector by combining multiple industry data items or calculating statistical summary indicators. The feature vector generation module (251) can assign elements that distinguish domain-specific states, such as industry, technology, and market, to each dimension of the vector. The feature vector can quantify and represent various state information, such as the company's current state, growth trends, and risk signals. The generated feature vector can be transmitted to the KPI calculation engine (253).

[0207] In one embodiment, at step 340, the electronic device (110) can calculate a key performance indicator of an enterprise through predefined calculation rules based on a generated feature vector. The KPI calculation engine (253) can receive a feature vector input from the feature vector generation module (251). The KPI calculation engine (253) can apply predefined linear or non-linear calculation rules. The KPI calculation engine (253) can calculate a key performance indicator by multiplying or combining a weight matrix with multiple elements of the feature vector. The KPI calculation engine (253) can calculate an exponential KPI, such as a growth index, a risk index, a stability index, etc. The KPI calculation engine (253) can record the unit, range, and interpretation criteria of the calculated KPI. The calculated key performance indicator can be transmitted to a rule-based judgment engine (255).

[0208] In one embodiment, at step 350, the electronic device (110) can generate a judgment result for the state by comparing the calculated key performance indicator with a specific condition. The rule-based judgment engine (255) can receive the key performance indicator input from the KPI calculation engine (253). The rule-based judgment engine (255) can pre-set specific conditions, such as a reference value, a threshold value, or an external comparison target. The rule-based judgment engine (255) can generate one of a plurality of discrete states (e.g., growth, stability, risk, slowdown, etc.) as a judgment result by comparing the calculated key performance indicator with the specific condition. The rule-based judgment engine (255) can record the conditions used in the judgment, the comparison target, the reference value, etc., along with the judgment result. The generated judgment result can be transmitted to the explanation information generation module (271).

[0209] In one embodiment, at step 360, the electronic device (110) can output explanatory information regarding the judgment basis and judgment path corresponding to the judgment result to a user terminal based on the generated judgment result. The explanatory information generation module (271) can receive the judgment result input from the rule-based judgment engine (255). The explanatory information generation module (271) can identify the judgment basis corresponding to the judgment result, for example, the used data source, feature vector, KPI, application condition, etc. The explanatory information generation module (271) can include the judgment path, that is, the step-by-step processing flow from data acquisition to judgment, as trace information. The explanatory information generation module (271) can generate explanatory information by integrating the judgment result, the judgment basis, and the judgment path. The output module (273) can transmit the generated explanatory information to the user terminal (120) via a network. The user terminal (120) can visualize and output the received explanatory information in various ways, such as a dashboard, report, or notification. The output of explanatory information can provide the user with reproducibility and traceability of the judgment result.

[0210] In one embodiment, the electronic device (110) may acquire industrial data from one or more public data sources. The electronic device (110) may collect industrial data from a global public verified data source (211) of an external data layer (210). The electronic device (110) may transmit the collected data to a data collection and normalization layer (230).

[0211] In one embodiment, the electronic device (110) can normalize acquired industrial data according to a predefined data structure based on an industrial domain. The electronic device (110) can transmit data acquired from a source-specific collection module (231) to a normalization module (233). The normalization module (233) can generate normalized data by performing entity classification, field configuration, and schema transformation of the data. The normalized data can be transmitted to an analysis layer (250).

[0212] In one embodiment, the electronic device (110) can generate a feature vector to represent the state of at least one of the industry, technology, or market to which the company belongs in the decision-making process of the company, based on normalized industrial data. The feature vector generation module (251) can receive normalized data as input and generate a feature vector through procedures such as time-based sorting, data combination, and statistical summarization. The generated feature vector can be transmitted to the KPI calculation engine (253).

[0213] In one embodiment, the electronic device (110) can calculate a key performance indicator of an enterprise through a predefined calculation rule based on a generated feature vector. The KPI calculation engine (253) can receive the feature vector as input and calculate the key performance indicator through procedures such as applying weights, combinatorial calculation, and normalization. The calculated key performance indicator can be transmitted to a rule-based judgment engine (255).

[0214] In one embodiment, the electronic device (110) can generate a judgment result for the state by comparing the calculated key performance indicator with a specific condition. The rule-based judgment engine (255) can generate a judgment result by comparing the key performance indicator with a reference value, a threshold value, or a comparison target. The generated judgment result can be transmitted to the explanation information generation module (271).

[0215] In at least one embodiment of the present disclosure, the electronic device (110) may output explanatory information regarding the judgment basis and judgment path corresponding to the judgment result to a user terminal based on the generated judgment result. An explanatory information generation module (271) may generate explanatory information including the judgment result, the judgment basis, and the judgment path. An output module (273) may transmit the explanatory information to a user terminal (120) via a network. The user terminal (120) may visually output the received explanatory information.

[0216] In one embodiment, the electronic device (130) may include an external data layer (210), a data collection and normalization layer (230), an analysis layer (250), and an output layer (270). The electronic device (130) may be composed of independent hardware or software modules for each layer. Each layer may process input data and output data by clearly distinguishing between them. Data flow between layers may be standardized according to internal interface specifications.

[0217] In at least one embodiment of the present disclosure, the external data layer (210) may include a global public verified data source (211). The global public verified data source (211) may include reliable public data sources such as policies, industry statistics, technology, patents, papers, and news. The external data layer (210) may manage access and authentication procedures for a plurality of data sources.

[0218] In one embodiment, the source-specific collection module (231) can acquire different types of industrial data from a global public verified data source (211) of an external data layer (210). The source-specific collection module (231) can access the data source in various ways, such as API, crawling, or file download. The source-specific collection module (231) can record metadata such as the type, format, and time of collection of the data. The source-specific collection module (231) can transmit the collected data to the normalization module (233).

[0219] In at least one embodiment of the present disclosure, the normalization module (233) can normalize industrial data obtained by the source-specific collection module (231) according to a predefined data structure based on an industrial domain. The normalization module (233) can perform procedures such as entity classification, field mapping, data type conversion, missing value processing, and unit unification in stages. The normalization module (233) can transmit the normalized data to the analysis layer (250).

[0220] In at least one embodiment of the present disclosure, the feature vector generation module (251) can generate a feature vector to represent the state of at least one of the industry, technology, or market to which the company belongs during the decision-making process of the company, based on normalized industry data. The feature vector generation module (251) can generate a multidimensional feature vector through procedures such as time-based alignment, data combination, statistical summarization, dimensionality reduction, or expansion. The generated feature vector can be transmitted to the KPI calculation engine (253).

[0221] In at least one embodiment of the present disclosure, the KPI calculation engine (253) receives a feature vector generated by the feature vector generation module (251) as input and can calculate the key performance indicators of the enterprise according to a predefined calculation rule. The KPI calculation engine (253) can calculate the KPI by applying procedures such as linear or non-linear combination formulas, weight matrices, normalization, and scaling. The KPI calculation engine (253) can record the unit, range, and interpretation criteria of the calculated KPI together. The calculated KPI can be transmitted to the rule-based judgment engine (255).

[0222] In one embodiment, the rule-based judgment engine (255) can generate a judgment result regarding the state of the judgment target by comparing the key performance indicator calculated by the KPI calculation engine (253) with a specific condition, e.g., a reference value, a threshold value, or a comparison target. The rule-based judgment engine (255) can generate one of a plurality of discrete states as a judgment result by comparing the KPI with the reference value, threshold value, comparison target, etc. The judgment result can be transmitted to the explanation information generation module (271).

[0223] In at least one embodiment of the present disclosure, the explanation information generation module (271) can generate explanation information regarding the judgment basis and judgment path corresponding to the judgment result based on the judgment result generated by the rule-based judgment engine (255). The explanation information generation module (271) can generate explanation information including metadata such as the judgment result, judgment basis, judgment path, data source, feature vector, KPI, and application condition. The generated explanation information can be transmitted to the output module (273).

[0224] In one embodiment, the output module (273) can transmit the explanation information generated by the explanation information generation module (271) to the user terminal (120) via a network and output it. The output module (273) can transmit the explanation information to the user terminal (120) in various ways, such as API responses, real-time notifications, and dashboard updates. The user terminal (120) can visually output the received explanation information or save it in a report format. The output of the explanation information can provide the user with reproducibility and traceability of the judgment result.

[0225] In one embodiment, the electronic device (110) may acquire industrial data from one or more public data sources. The electronic device (110) may acquire industrial data from a global public verified data source (211) of an external data layer (210). The electronic device (110) may access multiple public data sources that provide different types of public data, such as policy data, industrial statistics data, technical data, patent data, paper data, or news data. Depending on the access method of each public data source, the electronic device (110) may apply various collection methods, such as API calls, data crawling, or file downloads. The electronic device (110) may store the collected industrial data in a temporary storage or database (110). The electronic device (110) may verify the integrity and reliability of the data when collecting data. The electronic device (110) may record metadata such as the source of the acquired industrial data, the time of collection, and the data type.

[0226] In one embodiment, the public data source may be a source that provides different types of public data, such as policy data, industrial statistics data, technology data, patent data, paper data, or news data. The public data source may provide data types that include various information such as industrial policy, government policy, regulatory information, production volume by industry, sales, employment, technology trends, patent applications, paper publications, and industrial issues. The data structure, format, update cycle, and access rights of the public data source may differ for each data type. The electronic device (110) may select a data collection and processing method according to the characteristics of each public data source.

[0227] In at least one embodiment of the present disclosure, the public data source may include multilayer public data having different types and structures. The public data source may include various data structures such as structured data, unstructured data, and semi-structured data. The public data source may provide not only single-layer data but also metadata, detailed data, summary data, etc. composed of multiple layers. The electronic device (110) may analyze the structure of each layer of the multilayer public data and perform data collection and normalization procedures step by step.

[0228] In one embodiment, the electronic device (110) may acquire industrial data by directly collecting data from the public data source or by receiving data collected by an external system. When collecting directly, the electronic device (110) may request data in real-time or periodically by accessing the API endpoint of the public data source. The electronic device (110) may receive data collected by an external system through various interfaces such as files, databases, and message queues. The electronic device (110) may selectively apply one or both of direct collection and externally linked collection.

[0229] In one embodiment, the electronic device (110) may perform direct collection and external linkage collection in parallel when collecting data. The electronic device (110) may mutually verify the reliability of the data by performing direct collection and external linkage collection in parallel for the same data item. The electronic device (110) may manage the collection cycle, data format, and error handling policy separately for each data collection path.

[0230] In at least one embodiment of the present disclosure, the electronic device (110) can verify the integrity and reliability of the data when collecting data. The electronic device (110) can perform integrity verification procedures, such as checksums, hash values, and signatures, on the collected data. The electronic device (110) can calculate a data reliability score according to reliability evaluation criteria of the data source. The electronic device (110) can record the results of the data quality verification as metadata.

[0231] In one embodiment, the electronic device (110) may record metadata such as the source of acquired industrial data, the time of collection, and the data type. The electronic device (110) may generate metadata for each data item, such as a data source identifier, a collection timestamp, a data type, a collection method, and a quality verification result. The electronic device (110) may store the metadata in a database (110) to ensure data tracking and reproducibility in subsequent steps.

[0232] In one embodiment, the public data source may provide data including industrial policies, government policies, regulatory information, etc., such as policy data. Policy data may include policy documents, statutes, regulatory guidelines, etc., officially published by government agencies, industry associations, international organizations, etc. Policy data may be provided in various formats such as text, tables, PDF, XML, etc. The electronic device (110) may also collect metadata such as the source, publication date, and scope of application of the policy data.

[0233] In one embodiment, the public data source may provide data including statistical information such as production volume, sales, and employment by industry, such as industrial statistical data. Industrial statistical data may include official statistical data provided by national statistical agencies, industry associations, international organizations, etc. Industrial statistical data may be provided in various time series units, such as by year, quarter, or month. The electronic device (110) may record metadata such as the unit, measurement method, and aggregation standard of the industrial statistical data.

[0234] In at least one embodiment of the present disclosure, the public data source may provide data containing various information such as technology trends, research results, and industry issues, such as technical data, patent data, paper data, or news data. Technical data may be extracted from technical reports, standard documents, technical white papers, etc. Patent data may include patent application, registration, and citation information provided by the Patent Office, international patent databases, etc. Paper data may include full text, abstract, and citation information provided by academic journals, academic societies, open access databases, etc. News data may include articles, press releases, and summaries of industry issues provided by major media outlets, industry journals, online portals, etc. The electronic device (110) may separately manage data structure, update cycle, and reliability evaluation criteria for each data type.

[0235] In at least one embodiment of the present disclosure, the electronic device (110) may directly collect data from the public data source. The electronic device (110) may access the API endpoint of the public data source and request data in real time or at scheduled intervals. The electronic device (110) may extract unstructured data from websites, open data portals, etc., through data crawling. The electronic device (110) may directly collect data in various formats such as CSV, Excel, PDF, etc., via file download. The electronic device (110) may store the directly collected data in a temporary storage and then perform data integrity verification and normalization procedures.

[0236] In one embodiment, the electronic device (110) may apply various direct collection methods, such as API calls, data crawling, and file downloads. The electronic device (110) may separately implement data access rights, authentication procedures, and error handling logic for each collection method. The electronic device (110) may monitor the quality of the directly collected data in real time.

[0237] In one embodiment, the electronic device (110) can acquire industrial data by receiving data collected by an external system. The electronic device (110) can receive data collected by the external system through various interfaces such as file transfer, database linkage, message queue, and API response. The electronic device (110) can analyze the format, structure, and metadata of the data received from the external system and convert it to fit its internal data structure.

[0238] In at least one embodiment of the present disclosure, the electronic device (110) can perform data format conversion, transmission protocol management, etc. when connected with an external system. The electronic device (110) can apply a mapping table, a conversion script, a data parser, etc. for data format conversion when connected with an external system. When managing transmission protocols, the electronic device (110) can support various protocols such as HTTP, FTP, SFTP, and message queue. The electronic device (110) can detect data transmission errors, delays, duplicate receptions, etc., and perform exception handling during the connection with an external system.

[0239] In at least one embodiment of the present disclosure, the electronic device (110) can verify the integrity and reliability of the data when collecting data. The electronic device (110) can perform integrity verification procedures, such as checksums, hash values, and digital signatures, on the collected data in stages. The electronic device (110) can calculate a data reliability score according to reliability evaluation criteria of the data source. The electronic device (110) can record the results of the data quality verification as metadata.

[0240] In at least one embodiment of the present disclosure, the electronic device (110) may perform quality verification procedures on collected data, such as error detection, missing value verification, and data consistency check. The electronic device (110) may check the range, format, and logical consistency of data values ​​to detect errors. The electronic device (110) may check for the omission of required fields to verify missing values. The electronic device (110) may perform value comparison between multiple sources of the same item and verification of the continuity of time-series data to check data consistency. The electronic device (110) may perform follow-up measures in stages, such as data correction, re-collection, and error log recording, depending on the results of the quality verification.

[0241] In one embodiment, the electronic device (110) can normalize acquired industrial data according to a predefined data structure based on an industrial domain. The electronic device (110) can transmit the acquired industrial data through a source-specific collection module (231) to a normalization module (233). The normalization module (233) can refer to at least one of a predefined entity classification structure, data field configuration, and data schema for each industrial domain. The normalization module (233) can map each item of the acquired industrial data to a corresponding field of a predefined data structure. The normalization module (233) can convert attributes such as data type, unit, and name according to standardization rules. The normalization module (233) can store the converted data as a normalized data set. The normalized data can be utilized in a series of processes, such as generating feature vectors, calculating KPIs, and determining status, in a subsequent analysis layer (250). The normalization process ensures structural consistency of the data, thereby providing a basis for comparison and combination between heterogeneous data.

[0242] In at least one embodiment of the present disclosure, the electronic device (110) can ensure data consistency and efficiency of subsequent processing during the normalization process of industrial data. The normalization module (233) can prevent data interpretation errors by converting data items into consistent names, types, and units. The normalized data can be structured to enable automated processing and parallel computation in the analysis layer (250). Ensuring data consistency can increase the processing efficiency of subsequent steps, such as KPI calculation, status determination, and descriptive information generation. Data reliability can be improved by performing data quality verification and error detection in parallel during the normalization process.

[0243] In at least one embodiment of the present disclosure, the electronic device (110) can structure industrial data so that comparison and combination between data are possible in subsequent steps by converting industrial data obtained from different public data sources to match the same entity classification structure, data field configuration, or data schema. The normalization module (233) can analyze the source data structure of each public data source and map it to a standard entity classification structure defined by domain. If the names, types, and units of the data fields differ, the normalization module (233) can convert and integrate them according to the standard field configuration. The normalization module (233) can analyze the hierarchical structure, required / optional fields, reference relationships, etc. of the data schema and reconstruct it to match the standard schema. The structured data can be designed to enable direct comparison, combination, and statistical analysis between data collected from multiple sources. Domain-specific schema conversion can resolve data heterogeneity and increase the accuracy of subsequent analysis.

[0244] In one embodiment, the electronic device (110) can customize the data schema according to the characteristics of each industrial domain. The normalization module (233) can predefine essential data items, data types, units, reference relationships, etc. required for each industry. The normalization module (233) can ensure semantic consistency of the data by applying schema transformation rules specialized for each industrial domain. The customized data schema can be optimized for domain-specific analysis, such as calculating industry-specific KPIs and determining status.

[0245] In one embodiment, the electronic device (110) can structure data so that comparison and combination between data can be performed in subsequent steps using normalized industrial data. The normalization module (233) can structure the data to enable direct comparison between data by applying the same entity, field, and schema criteria to each data item. The normalized data can be designed to enable various subsequent processing, such as combining data from multiple sources, time series analysis, and cross-comparison. Data structuring can provide the basis for complex operations, such as feature vector generation and KPI calculation, in the analysis layer (250).

[0246] In at least one embodiment of the present disclosure, the electronic device (110) may apply a standardized data structure to enhance the possibility of data comparison and combination. The normalization module (233) can ensure compatibility and interoperability between data by applying the standardized data structure. The standardized structure can enhance efficiency in various utilization scenarios, such as data integration, linkage with external systems, and large-scale data analysis. Ensuring the possibility of data comparison and combination can strengthen the reliability and reproducibility of corporate decision support.

[0247] In at least one embodiment of the present disclosure, the electronic device (110) can convert industrial data obtained from different public data sources to fit the same entity classification structure. The normalization module (233) can map entities (e.g., enterprise, industry, technology, market, etc.) of each data source according to a standard entity classification scheme. The entity classification structure may include a hierarchical classification scheme, codes, identifiers, etc. defined by industry domain. The conversion to fit the entity classification structure can ensure semantic consistency between data and increase the accuracy of subsequent analysis.

[0248] In at least one embodiment of the present disclosure, the electronic device (110) may predefine an entity classification structure of industrial data and use it as a standard for data normalization. The normalization module (233) may predefine an entity classification structure and verify whether each data item is suitable for the corresponding classification system. The predefined entity classification structure may be used as a reference point in various processing steps, such as data normalization, comparison, combination, and statistical analysis.

[0249] In one embodiment, the electronic device (110) can convert industrial data obtained from different public data sources to match the same data field configuration. The normalization module (233) can convert the field names, types, units, formats, etc. of each data source according to the standard field configuration. The data field configuration may include required / optional fields, data types, units, format rules, etc., required by each industrial domain. The conversion to match the same data field configuration can provide a basis for direct comparison, combination, and statistical processing between data.

[0250] In one embodiment, the electronic device (110) can ensure data compatibility by standardizing the names, types, units, etc. of data fields. The normalization module (233) can unify the names of data fields into standard terms and convert data types and units according to a consistent standard. Standardized data fields can be designed to be suitable for various applications, such as linking with external systems, large-scale data integration, and automated analysis. Ensuring data field compatibility can contribute to improving data quality and reliability.

[0251] In at least one embodiment of the present disclosure, the electronic device (110) can convert industrial data obtained from different public data sources to fit the same data schema. The normalization module (233) can analyze the schema structure (hierarchy, fields, reference relationships, etc.) of each data source and convert it to fit the standard data schema. The data schema may include a set of fields, a hierarchical structure, reference rules, etc. defined by industrial domain. Conversion to fit the same data schema can ensure structural consistency and interoperability between data.

[0252] In one embodiment, the electronic device (110) may predefine the structure of a data schema and map each data item to the corresponding schema during the normalization process. The normalization module (233) may predefine the fields, hierarchy, and reference relationships of the data schema and verify whether each data item is properly mapped to the corresponding structure. The predefined data schema may be used as a standard for subsequent processing such as data normalization, comparison, combination, and analysis.

[0253] In at least one embodiment of the present disclosure, the electronic device (110) can map each field of acquired industrial data to a field of a predefined data structure and convert the data type or unit. The normalization module (233) can map each field name of the source data to a standard field name. The normalization module (233) can convert the data type (e.g., string, integer, real number, etc.) or unit (e.g., USD, KRW, %, count, etc.) to a standard specification. The field mapping and type conversion process can ensure structural consistency of the data and accuracy of interpretation.

[0254] In one embodiment, the electronic device (110) can verify the accuracy and consistency of the data during the type conversion process of the data field. The normalization module (233) can verify whether the range, type, and unit of the data value after type conversion conform to standard specifications. Data in which errors are found during the verification process may be branched to a separate error handling procedure. Verification of the accuracy and consistency of the data can increase the reliability of subsequent analysis steps.

[0255] In one embodiment, the electronic device (110) can structure normalized industrial data to enable comparison and combination of data in a subsequent step. The normalization module (233) can generate key fields (e.g., entity ID, date, industry code, etc.) required for comparison and combination of the normalized data. The normalized data can be designed to facilitate combination and cross-analysis of data by including key fields, reference values, indexes, etc. The preparation for data comparison and combination can provide the basis for complex operations, such as feature vector generation and KPI calculation, in the analysis layer (250).

[0256] In at least one embodiment of the present disclosure, the electronic device (110) may generate key fields, reference values, indexes, etc. for data comparison and combination. The normalization module (233) may assign a unique identifier, reference key, and index value to each data item. The generated key fields and reference values ​​may be utilized in various combination operations, such as joining, merging, and cross-analysis between data. Preparing for data comparison and combination can increase the efficiency of data integration and analysis.

[0257] In at least one embodiment of the present disclosure, if an error occurs or there is missing data during the normalization process, the electronic device (110) may identify the data and apply a separate processing policy. The normalization module (233) may automatically detect errors during the data normalization process, such as missing required fields, data type mismatches, or values ​​exceeding a range. The error or missing data may be recorded in a separate error log or processed via branching according to a subsequent processing policy. Identifying error data and applying a processing policy can contribute to ensuring data quality and reliability.

[0258] In one embodiment, the electronic device (110) may perform procedures such as correction, replacement, deletion, or log recording on error or missing data. The normalization module (233) may apply predefined correction rules (e.g., mean value replacement, previous value interpolation, etc.) to the error data. Missing data may be replaced with external reference data, a prediction model, a default value, etc. Severe errors or unrecoverable data may be deleted or recorded in an error log so that they can be reviewed later. The error and missing data processing procedures can enhance the reliability and reproducibility of the entire data processing pipeline.

[0259] In one embodiment, the electronic device (110) can generate a feature vector to represent at least one state of the industry, technology, or market to which the company belongs during the decision-making process of the company, based on normalized industrial data. The electronic device (110) can receive normalized industrial data as input. The electronic device (110) can analyze each item of the input normalized industrial data to extract an indicator that reflects at least one state of the industry, technology, or market. The electronic device (110) can assign the extracted indicators to each element of a multidimensional vector. The electronic device (110) can configure each element of the vector to numerically represent a specific industry, technology, or market state. The electronic device (110) can store the generated feature vector as input data for subsequent KPI calculation and state judgment steps. The feature vector generation process can simultaneously ensure structural consistency of the data and quantitative representation of the state.

[0260] In at least one embodiment of the present disclosure, the electronic device (110) can quantitatively represent the state of the input data in the subsequent KPI calculation and state determination steps by generating a feature vector. The electronic device (110) can quantify the state by industry, technology, and market for each element of the feature vector. The electronic device (110) can input the quantified feature vector into the KPI calculation engine (253). The KPI calculation engine (253) can calculate key performance indicators based on the input feature vector. The quantitatively represented feature vector can be used as a reference value for discrete state classification in the state determination engine (255). The quantitative representation of the feature vector can increase the reliability and reproducibility of the judgment results.

[0261] In one embodiment, the electronic device (110) can generate a feature vector representing the state of a judgment target by aligning normalized industrial data based on time, reflecting data changes according to multiple points in time, or combining or summarizing multiple industrial data. The electronic device (110) can assign a timestamp to each item of the normalized industrial data. The electronic device (110) can form a time-series data structure by aligning the data according to time order. The electronic device (110) can calculate temporal characteristics such as the amount of change, the rate of increase or decrease, and trends by comparing data collected at multiple points in time. The electronic device (110) can extract and combine data at the same point in time from multiple industrial data sources. The electronic device (110) can convert the combined data into summary statistics (e.g., mean, variance, maximum / minimum value, etc.). The electronic device (110) can assign the summarized statistical indicators to each element of the feature vector. This integration strategy can express the state of the judgment target in a multidimensional way by reflecting information from multiple points in time and multiple sources.

[0262] In one embodiment, the electronic device (110) can integrate data from multiple time points to generate a vector that reflects time series changes or trends. The electronic device (110) can analyze the fluctuation patterns of the data for each time point. The electronic device (110) can calculate time series characteristic indicators such as the amount of time series change, moving average, and cumulative sum. The electronic device (110) can include the calculated time series characteristic indicators as elements of the feature vector. The feature vector reflecting time series changes can accurately represent the dynamic state of the industry, technology, and market.

[0263] In one embodiment, the electronic device (110) may apply quality control procedures, such as data normalization, missing value processing, and outlier correction, to maintain the quality of the feature vector. The electronic device (110) may replace missing values ​​in the input data with the mean, median, or recent value. The electronic device (110) may identify outliers in the vector elements by applying an outlier detection algorithm. The electronic device (110) may correct outliers or mark them with a separate flag to reflect them in subsequent processing. The electronic device (110) may adjust the distribution of the vector elements to a certain range by applying data normalization techniques (e.g., Min-Max normalization, Z-score standardization, etc.). Quality control procedures can increase the reliability of the feature vector and the accuracy of subsequent analysis.

[0264] In one embodiment, the electronic device (110) may apply scale management techniques, such as standardization, normalization, and unit conversion, to maintain a consistent scale of the feature vector. If the units differ for each vector element, the electronic device (110) may convert them to a predefined standard unit. The electronic device (110) may normalize each element of the vector to the same scale to ensure consistency in comparison between data. The electronic device (110) may input the standardized vector into the KPI calculation and status judgment stages to increase the reliability of the analysis results. Scale management techniques can ensure the accuracy of data integration between multiple data sources and time points.

[0265] In one embodiment, the electronic device (110) can sort normalized industrial data based on time. The electronic device (110) can sort the data in chronological order by assigning a timestamp to each data item. The electronic device (110) can generate a time series vector based on the sorted data. Time series sorting preserves the temporal flow of the data and can provide a basis for dynamic state analysis.

[0266] In one embodiment, the electronic device (110) can be configured to preserve the chronological order of a single time series data so that each element of the vector represents a specific point in time. The electronic device (110) can assign a data value of the corresponding point in time to each element of the vector. The electronic device (110) can identify a time series pattern by analyzing the variability of the data at each point in time. Single time series alignment can precisely track changes in state at a specific point in time.

[0267] In one embodiment, the electronic device (110) can reflect data changes according to multiple time points. The electronic device (110) can collect data from multiple time points and form a data set for each time point. The electronic device (110) can calculate temporal characteristics such as the amount of data change between time points, the rate of increase or decrease, and trends. Changes in data from multiple time points can reflect the dynamic state of the industry, technology, and market.

[0268] In one embodiment, the electronic device (110) can synchronize data from different points in time to generate a vector that enables comparison and combination on the same time axis. The electronic device (110) can perform synchronization based on the timestamps of the data for each point in time. The electronic device (110) can assign the synchronized data to the same index of the vector to enable direct comparison between points in time. Multi-point synchronization can improve the accuracy of time series analysis and trend prediction.

[0269] In one embodiment, the electronic device (110) can manage the continuity of a vector by interpolating missing intervals for discontinuously collected data or by separately marking discontinuous intervals. The electronic device (110) can apply mathematical interpolation techniques, such as linear interpolation or spline interpolation, to the missing intervals. If interpolation is not possible, the electronic device (110) can mark the interval as a missing value or assign a separate flag. Clear marking of discontinuous intervals can prevent data interpretation errors in subsequent analysis.

[0270] In one embodiment, the electronic device (110) may include information regarding missing values, outliers, or data collection intervals within a vector, reflecting the characteristics of discontinuous data. The electronic device (110) may add metadata, such as whether there are missing values, outliers, or data collection intervals, to each element of the vector. The electronic device (110) may utilize the metadata to enhance the interpretation and quality control of discontinuous data. Discontinuous data processing may reflect the realistic collection environment of industrial data, thereby increasing the reliability of the analysis results.

[0271] In at least one embodiment of the present disclosure, the electronic device (110) can generate a feature vector representing the state of a judgment target by combining or summarizing a plurality of industrial data. The electronic device (110) can extract data of the same entity or at the same point in time from a plurality of industrial data sources. The electronic device (110) can combine the extracted data to create an integrated data set. The electronic device (110) can calculate statistical summary indicators, such as mean, variance, maximum value, and minimum value, for the integrated data set. The electronic device (110) can assign the calculated statistical indicators to each element of the feature vector. The use of statistical summary indicators can aggregate information from multiple data sources to enhance the representativeness of the state representation.

[0272] In one embodiment, the electronic device (110) may calculate statistical summary indicators, such as the mean, variance, maximum / minimum value, etc. of multiple industrial data and include them as each element of a vector. The electronic device (110) may calculate statistical summary indicators for each industrial data item. The electronic device (110) may assign the calculated statistical indicators to designated indices of the vector. Vectorization of statistical summary indicators can reduce data complexity and increase the efficiency of subsequent analysis.

[0273] In one embodiment, the electronic device (110) can adjust the dimension of a vector by applying a dimension reduction technique (e.g., principal component analysis, singular value decomposition, etc.) or a dimension expansion technique (e.g., generation of derived variables, etc.) when combining multiple industrial data. The electronic device (110) can extract only the key information of the vector by applying statistical dimension reduction techniques such as principal component analysis (PCA) and singular value decomposition (SVD). The electronic device (110) can include additional information suitable for the purpose of analysis in the vector by applying dimension expansion techniques such as generation of derived variables and polynomial expansion. The electronic device (110) can reflect the results of the dimension reduction or expansion in the final structure of the vector. The dimension adjustment technique can increase processing efficiency while minimizing information loss.

[0274] In at least one embodiment of the present disclosure, the electronic device (110) can reduce or expand the dimension of a vector according to the purpose of analysis, thereby increasing processing efficiency while minimizing information loss. The electronic device (110) can determine the optimal number of dimensions by considering the purpose of analysis, data characteristics, computational resources, etc. The electronic device (110) can evaluate the information preservation rate when reducing the dimension and adjust it so that key information is maintained. When expanding the dimension, it can verify the significance of derived variables to prevent the addition of unnecessary information. Optimized dimension adjustment can simultaneously secure the interpretability of the vector and computational efficiency.

[0275] In at least one embodiment of the present disclosure, the electronic device (110) can perform normalization processing on a vector generated after combining multiple industrial data to consistently scale each element. The electronic device (110) can apply normalization techniques, such as Min-Max normalization and Z-score standardization, to each element of the vector. The electronic device (110) can input the normalized vector into a subsequent KPI calculation and status determination step. Normalization processing can increase the consistency of comparison between data and the reliability of analysis results.

[0276] In at least one embodiment of the present disclosure, the electronic device (110) can ensure the reliability of comparison between data and subsequent KPI calculation through vector normalization. The electronic device (110) can increase the accuracy of indicator calculation by inputting the normalized vector into the KPI calculation engine (253). The electronic device (110) can minimize the influence of outliers by analyzing the distribution characteristics of the data during the normalization process. Vector normalization can provide a basis for data integration between multiple data sources and time points.

[0277] In at least one embodiment of the present disclosure, the electronic device (110) can represent at least one state of an industry, technology, or market to which the company belongs through a feature vector. The electronic device (110) can map each element of the feature vector to an industry-specific, technology-specific, or market-specific state indicator. The electronic device (110) can pre-define the definition and calculation method of each state indicator. Based on the mapped state indicators, the electronic device (110) can represent the company's industry, technology, and market states in a multidimensional manner. State mapping can enable simultaneous analysis and comparison of multiple states.

[0278] In at least one embodiment of the present disclosure, the electronic device (110) can provide multidimensional state information by mapping each element of a vector to an industry-specific, technology-specific, and market-specific state indicator. The electronic device (110) can assign various state indicators, such as industry-specific growth rates, technology-specific innovation indices, and market share, as vector elements. The electronic device (110) can utilize the multidimensional state information in the KPI calculation and state judgment stages. Providing multidimensional state information can increase the precision and reliability of corporate decision-making.

[0279] In one embodiment, the electronic device (110) can design a feature space such that each element of the feature vector reflects meaningful state information. The electronic device (110) can place domain-specific state indicators, such as industry, technology, and market, in each dimension of the feature space. The electronic device (110) can clearly define the meaning, unit, and data type of each dimension. The feature space design can increase the interpretability and analysis efficiency of the vector.

[0280] In at least one embodiment of the present disclosure, the electronic device (110) can flexibly define the structure of the feature space according to the purpose of analysis, data characteristics, and domain requirements. The electronic device (110) can adjust the number of dimensions, the arrangement order, and the configuration of elements of the feature space according to the purpose of analysis. The electronic device (110) can design a suitable feature space structure according to data characteristics (e.g., continuous, categorical, discrete, etc.). Depending on domain requirements, specific state indicators may be emphasized or unnecessary dimensions may be removed. The flexible feature space design can respond to various analysis scenarios.

[0281] In at least one embodiment of the present disclosure, the electronic device (110) can predefine a sample format of a vector for representing a state to clearly specify the name, unit, data type, time point information, etc. of each element. The electronic device (110) may include the name, unit, data type, time point information, definition of a state indicator, etc. of each element in the sample vector format. The electronic device (110) can generate a feature vector according to the predefined format to ensure consistency in data interpretation. Clear definition of the vector format can contribute to the automation and standardization of subsequent KPI calculation and state judgment steps.

[0282] In at least one embodiment of the present disclosure, the electronic device (110) can enable consistent data processing in subsequent KPI calculation and status determination steps based on a sample vector format. The electronic device (110) can transmit data to the KPI calculation engine (253) and the status determination engine (255) according to the sample vector format. By maintaining a consistent data format, the electronic device (110) can increase the reproducibility and traceability of the analysis results. The application of the sample vector format can support standardization of data flow and quality control throughout the system.

[0283] In one embodiment, the KPI calculation engine (253) receives a feature vector generated by the feature vector generation module (251) as input and can calculate the company's key performance indicators according to predefined calculation rules. The KPI calculation engine (253) can receive a multidimensional feature vector as input from the feature vector generation module (251). The KPI calculation engine (253) can analyze each element of the input feature vector and perform calculations according to predefined calculation rules. The KPI calculation engine (253) can calculate the calculation result as a key performance indicator. The KPI calculation engine (253) can provide the calculated key performance indicator as an input value to the subsequent state determination step (350). The calculation process of the KPI calculation engine (253) can simultaneously ensure structural consistency of the input data and reproducibility of the result.

[0284] In one embodiment, the key performance indicator may be an index-type indicator used to assess the state of an industry, technology, market, or company during the company's decision-making process. The key performance indicator may be calculated by combining or weighting multiple feature vector elements. The key performance indicator may numerically represent the state of the industry, technology, market, or company. The key performance indicator may be utilized as a reference value for assessing the state during the company's decision-making process. In a subsequent step, the key performance indicator may be interpreted through comparison with a reference value, a threshold value, or a comparison target.

[0285] In at least one embodiment of the present disclosure, the KPI calculation step (340) may provide indicators capable of quantitatively evaluating the state of an industry, technology, market, or company to support corporate decision-making. The KPI calculation step (340) may quantify the state of an industry, technology, market, or company based on input feature vectors. The KPI calculation step (340) may provide the quantitatively calculated indicators to a corporate decision support system. The KPI calculation step (340) may objectively evaluate the state of a company through quantitative indicators. The KPI calculation step (340) may increase the reliability and reproducibility of the evaluation results.

[0286] In at least one embodiment of the present disclosure, the KPI calculation engine (253) can calculate key performance indicators of an enterprise according to predefined calculation rules. The KPI calculation engine (253) can perform calculations by referring to predefined formulas, algorithms, or rule sets. The KPI calculation engine (253) can apply calculation rules to each element of an input feature vector. The KPI calculation engine (253) can calculate the calculation result as a key performance indicator. The KPI calculation engine (253) can calculate various KPIs depending on changes or extensions to the calculation rules.

[0287] In one embodiment, the KPI calculation step (340) can calculate an exponential KPI by combining or weighting multiple feature vector elements. The KPI calculation step (340) can selectively combine multiple elements of the feature vector. The KPI calculation step (340) can perform weighted summation or weighted combination by assigning weights to each element. The KPI calculation step (340) can convert the weighted result into an exponential KPI. The KPI calculation step (340) can calculate various forms of KPIs depending on the combination and weighting methods.

[0288] In at least one embodiment of the present disclosure, the KPI calculation step (340) may include result interpretation guidelines so that the calculated key performance indicators can be used as criteria for judging the state of the enterprise. The KPI calculation step (340) may pre-define a method for interpreting the calculated KPI values. The KPI calculation step (340) may provide criteria for interpreting the state when the KPI value is above or below a specific threshold value. The KPI calculation step (340) may map discrete states such as growth, stability, and risk according to the range of KPI values. The KPI calculation step (340) may transmit the interpretation guidelines to the subsequent state judgment step (350).

[0289] In one embodiment, the KPI calculation result can be interpreted in a subsequent state judgment step (350) through comparison with a reference value, a threshold value, or a comparison target. The KPI calculation result can classify the state through comparison with the reference value. The KPI calculation result can calculate the difference with the threshold value or an external comparison target. The KPI calculation result can be converted into a discrete state in the state judgment engine (255). The interpretation of the KPI calculation result can be used as an objective basis in the corporate decision-making process.

[0290] In one embodiment, the KPI calculation step (340) can calculate a key performance indicator by combining multiple feature vector elements into a linear or non-linear combination formula. The KPI calculation step (340) can apply a linear combination (e.g., weighted summation) or a non-linear function (e.g., multiplication, exponential, logarithm, etc.) to the feature vector elements. The KPI calculation step (340) can calculate various KPIs depending on the form of the combination formula. The KPI calculation step (340) can adjust the sensitivity and interpretability of the state evaluation according to the selection of the combination formula.

[0291] In one embodiment, the KPI calculation engine (253) may use a linear combination or a non-linear function (e.g., multiplication, exponential, logarithm, etc.) to which weights are applied to each element of the input feature vector. The KPI calculation engine (253) may apply a linear combination formula that multiplies each feature vector element by a weight and then sums them. The KPI calculation engine (253) may calculate KPIs by applying non-linear functions such as multiplication, exponential, and logarithm. The KPI calculation engine (253) may control the characteristics of the KPI calculation results by adjusting the parameters of the combination formula. The KPI calculation engine (253) can calculate customized KPIs by industry and domain through the application of various combination formulas.

[0292] In at least one embodiment of the present disclosure, the KPI calculation step (340) may apply weights differentially according to the importance of each feature vector element. The KPI calculation step (340) may evaluate the importance for each feature vector element. The KPI calculation step (340) may assign different weights to each element according to the importance. The KPI calculation step (340) may increase the sensitivity and accuracy of the KPI calculation through the differential application of weights. The KPI calculation step (340) may define weight setting criteria by domain.

[0293] In one embodiment, the KPI calculation engine (253) can ensure flexibility in KPI calculation by applying different weight matrices by domain, industry, and time point. The KPI calculation engine (253) can select a weight matrix according to industry domain, market characteristics, and data characteristics by time point. The KPI calculation engine (253) can apply the selected weight matrix to a feature vector through multiplication operations, etc. The KPI calculation engine (253) can support various KPI calculation scenarios by changing the weight matrix. The KPI calculation engine (253) can reflect the results of applying the weight matrix in the KPI calculation results.

[0294] In at least one embodiment of the present disclosure, the KPI calculation step (340) may predefine weights and combination parameters or set them through a data-based learning method. The KPI calculation step (340) may predefine weights and combination parameters through expert knowledge or domain analysis. The KPI calculation step (340) may automatically set parameters through data-based learning methods such as historical data, statistical analysis, and machine learning. The KPI calculation step (340) may selectively apply a parameter setting method. The KPI calculation step (340) may increase the accuracy and reliability of KPI calculation through parameter optimization.

[0295] In one embodiment, the KPI calculation engine (253) can optimize parameters by utilizing past data, expert knowledge, statistical analysis results, etc. The KPI calculation engine (253) can adjust parameters by comparing past KPI calculation results with actual state data. The KPI calculation engine (253) can correct weight and combination parameters by reflecting expert feedback. The KPI calculation engine (253) can verify the significance of parameters based on statistical analysis results. The KPI calculation engine (253) can store the optimized parameters and apply them to subsequent KPI calculations.

[0296] In one embodiment, the KPI calculation step (340) may select and combine elements that are meaningful for KPI calculation among a plurality of feature vector elements. The KPI calculation step (340) may analyze each element of the input feature vector to select elements that contribute significantly to KPI calculation. The KPI calculation step (340) may increase the efficiency and accuracy of KPI calculation by combining only the selected elements. The KPI calculation step (340) may define the criteria for selecting elements by domain and by industry. The KPI calculation step (340) may transmit the result of element selection to the KPI calculation engine (253).

[0297] In at least one embodiment of the present disclosure, the KPI calculation engine (253) can analyze each element of the input feature vector and select only the elements suitable for KPI calculation. The KPI calculation engine (253) can evaluate the statistical significance, correlation, contribution, etc. of each element. The KPI calculation engine (253) can select only the elements suitable for KPI calculation based on the evaluation results. The KPI calculation engine (253) can combine the selected elements to form a KPI calculation formula. The KPI calculation engine (253) can automate the element selection process to enable consistent KPI calculation.

[0298] In at least one embodiment of the present disclosure, the KPI calculation step (340) may adjust the contribution of each element by applying a weight matrix to selected feature vector elements. The KPI calculation step (340) may apply the weight matrix to selected feature vector elements through multiplication operations, etc. The KPI calculation step (340) may numerically reflect the contribution of each element of the weight matrix. The KPI calculation step (340) may reflect the result of applying the weight matrix in the KPI calculation formula. The KPI calculation step (340) may adjust the sensitivity of the KPI calculation result by changing the weight matrix.

[0299] In at least one embodiment of the present disclosure, the KPI calculation engine (253) can calculate a final KPI value by applying a weight matrix to a multiplication operation or the like. The KPI calculation engine (253) can perform operations on selected feature vector elements and a weight matrix using a matrix multiplication method. The KPI calculation engine (253) can calculate the result of the operation as a final KPI value. The KPI calculation engine (253) can apply vectorization or parallel processing for the efficiency of matrix operations. The KPI calculation engine (253) can transmit the calculated KPI value to a subsequent state determination step (350).

[0300] In at least one embodiment of the present disclosure, the KPI calculation step (340) may perform normalization or scaling processing on the feature vector elements and the KPI calculation result. The KPI calculation step (340) may analyze the unit difference, distribution characteristics, etc. of the input data. The KPI calculation step (340) may apply standardization techniques such as min-max normalization and z-score normalization. The KPI calculation step (340) may input the normalized data into the KPI calculation formula. The KPI calculation step (340) may adjust the range of values ​​by applying scaling processing to the KPI calculation result. The KPI calculation step (340) may facilitate the interpretation and comparison of KPI values ​​through normalization and scaling.

[0301] In one embodiment, the KPI calculation engine (253) may apply a normalization technique (e.g., min-max, z-score, etc.) by considering the unit difference and distribution characteristics of the input data. The KPI calculation engine (253) may calculate the minimum value, maximum value, mean, standard deviation, etc., of each feature vector element. The KPI calculation engine (253) may normalize each element using the calculated statistical values. The KPI calculation engine (253) may derive consistent results by applying the normalized data to the KPI calculation formula. The KPI calculation engine (253) may reflect the normalization and scaling results in the KPI calculation results.

[0302] In at least one embodiment of the present disclosure, at step 340, the electronic device (110) can calculate a key performance indicator of an enterprise based on a generated feature vector through a predefined calculation rule. The electronic device (110) can receive a feature vector generated from a feature vector generation module (251) as an input value. The electronic device (110) can perform calculations by referring to a predefined calculation rule. The electronic device (110) can calculate the result of the calculation as a key performance indicator. The electronic device (110) can transmit the calculated key performance indicator to a subsequent state determination step (350).

[0303] In at least one embodiment of the present disclosure, the key performance indicator may be an index-type indicator for judging the state of an industry, technology, market, or company in the decision-making process of a company, and may be calculated by combining or weighting multiple feature vector elements. The key performance indicator may be calculated by combining multiple feature vector elements in a linear or non-linear manner. The key performance indicator may be calculated by summing the elements after multiplying each element by a weight, or by applying a non-linear function (e.g., exp, log, etc.). The key performance indicator may be expressed in various forms depending on the combination and weighting methods. The key performance indicator may be utilized as a reference value for quantitatively evaluating the state of a company.

[0304] In one embodiment, the KPI calculation step (340) may apply a calculation formula such as, for example, KPI = w1*x1 + w2*x2 + ... + wn*xn or KPI = exp(w1*x1 + w2*x2). The KPI calculation step (340) may apply a linear combination formula in which each feature vector element (x1, x2, ..., xn) is multiplied by a weight (w1, w2, ..., wn) and then summed. The KPI calculation step (340) may apply a non-linear function, such as an exponential function (exp) or a logarithmic function (log), to the summation result. The KPI calculation step (340) may adjust the parameters of the calculation formula according to the domain. The KPI calculation step (340) may store the result of applying the formula as the KPI calculation result.

[0305] In at least one embodiment of the present disclosure, the KPI calculation step (340) may provide interpretation guidelines so that the calculated KPI value can be used as a standard for quantitatively evaluating the state of the enterprise. The KPI calculation step (340) may pre-define the interpretation criteria for the KPI value. The KPI calculation step (340) may set state classification criteria according to the range of the KPI value. The KPI calculation step (340) may transmit the interpretation guidelines to the subsequent state judgment step (350). The KPI calculation step (340) may clearly explain the meaning of the KPI value through the interpretation guidelines.

[0306] In at least one embodiment of the present disclosure, the KPI calculation result may be interpreted in a subsequent state judgment step (350) through comparison with a reference value, a threshold value, or a comparison target. The KPI calculation result may be classified into discrete states such as growth, stability, or risk through comparison with a reference value. The KPI calculation result may calculate the difference from a threshold value or an external comparison target. The KPI calculation result may be used as an input value for state judgment in a state judgment engine (255). The interpretation of the KPI calculation result may be provided as an objective basis in the decision-making process of the enterprise.

[0307] In at least one embodiment of the present disclosure, the KPI calculation step (340) may predefine the unit and range of values ​​of the calculated KPI or set them by domain. The KPI calculation step (340) may predefine the unit and range of values ​​according to the KPI calculation formula. The KPI calculation step (340) may set the unit and range of KPI values ​​by domain and by industry. The KPI calculation step (340) may reflect the results of setting the unit and range in the KPI interpretation guidelines. The KPI calculation step (340) may facilitate the comparison and interpretation of KPI values ​​through consistency in units and ranges.

[0308] In at least one embodiment of the present disclosure, the KPI calculation engine (253) may apply unit consistency and range limits to facilitate the interpretation and comparison of KPI values. The KPI calculation engine (253) may perform unit conversion or scaling processing according to the KPI calculation formula. The KPI calculation engine (253) may prevent outliers or error values ​​by applying range limits to KPI values. The KPI calculation engine (253) may reflect the results of unit consistency and range limits in the KPI calculation results. The KPI calculation engine (253) may increase the reliability and interpretability of KPI values ​​through consistent unit and range settings.

[0309] In at least one embodiment of the present disclosure, the electronic device (110) can generate a judgment result for the state by comparing a calculated key performance indicator with a specific condition. The electronic device (110) can receive a key performance indicator calculated by a KPI calculation engine (253) as an input value. The electronic device (110) can compare the input key performance indicator with a specific condition, such as a reference value, a threshold value, or a comparison target. The electronic device (110) can evaluate whether the state of the judgment target corresponds to one of a plurality of predefined discrete states through a comparison operation. The electronic device (110) can generate a result code or message for each state based on the comparison result. The electronic device (110) can transmit the judgment result to a subsequent explanation information generation step. The electronic device (110) can ensure consistency and reproducibility of the state judgment through the comparison and judgment process.

[0310] In at least one embodiment of the present disclosure, the specific condition may be a reference value, a threshold value, or a comparison target, and the electronic device (110) may determine whether the state of the judgment target corresponds to one of a plurality of discrete states. The specific condition may be a reference value, a threshold value, or an external comparison target that is pre-stored in the database (110) or internal memory. The electronic device (110) may classify the state according to whether the input key performance indicator is greater than or equal to the reference value, less than the threshold value, or relative to the external comparison target. The electronic device (110) may apply different state classification criteria for each condition. The electronic device (110) may select one of the plurality of discrete states and output it as a judgment result. The electronic device (110) may record the classification result by condition as a basis for judgment.

[0311] In at least one embodiment of the present disclosure, the electronic device (110) may define a plurality of discrete states and pre-set a standard for each state. The electronic device (110) may store a list of a plurality of discrete states in a database (110) or in internal memory. The electronic device (110) may pre-set a standard value, threshold value, or comparison target applied to each state. When the standard for each state changes, the electronic device (110) may update and apply the standard value. The electronic device (110) may use the standard information for each state in conjunction with a judgment logic.

[0312] In one embodiment, the electronic device (110) may store a reference value, threshold value, or comparison target applied by state in a database (110) or internal memory. The electronic device (110) may query the reference value by state from the database (110). The electronic device (110) may refer to the threshold value or comparison target stored in internal memory. The electronic device (110) may record the change history of the reference value, threshold value, or comparison target. The electronic device (110) may perform a state determination in real-time or in batch mode based on the stored reference information.

[0313] In one embodiment, the electronic device (110) may hierarchically organize the logic for discrete state determination. The electronic device (110) may design the state determination logic by separating it into multiple layers. The electronic device (110) may evaluate the determination criteria that are applied preferentially in the upper layer. The electronic device (110) may perform additional determinations in the lower layer based on the results of the upper layer. The electronic device (110) may record the determination results for each layer step by step.

[0314] In at least one embodiment of the present disclosure, the electronic device (110) assigns priorities to a plurality of state judgment criteria and may omit the evaluation of lower criteria when a higher priority criterion is satisfied. The electronic device (110) may assign a priority value to each state judgment criterion. The electronic device (110) may stop the evaluation of lower criteria when an input key performance indicator satisfies a higher priority criterion. The electronic device (110) may store the priority evaluation results as a basis for judgment. The electronic device (110) may improve computational efficiency through priority-based judgment.

[0315] In at least one embodiment of the present disclosure, the electronic device (110) may allow only one of a plurality of states to be selected as the final judgment result through a hierarchical structure of state judgment logic. The electronic device (110) may implement branching logic to select only a single state as the final result within the hierarchical structure. The electronic device (110) may select only one state according to priority when a plurality of states are satisfied simultaneously. The electronic device (110) may transmit the final judgment result to a subsequent explanation information generation step. The electronic device (110) may ensure consistency of the judgment result through the hierarchical structure.

[0316] In one embodiment, the electronic device (110) may classify exception situations in which a reference value, threshold value, or comparison target is not all satisfied into a separate state (e.g., undefined state, pending state, etc.). The electronic device (110) may classify an input key performance indicator into an exception state if it does not satisfy all reference values, threshold values, or comparison target conditions. The electronic device (110) may designate the exception state as an undefined state, pending state, or other separate code. The electronic device (110) may transmit the judgment result to a subsequent step when an exception state occurs. The electronic device (110) may log the exception state classification history.

[0317] In at least one embodiment of the present disclosure, the electronic device (110) may, through exception handling logic, withhold a judgment result or set it to a warning state when a missing, error, or abnormal value of input data is detected. The electronic device (110) may pre-verify whether there is a missing input data. The electronic device (110) may execute exception handling logic when a data error or abnormal value is detected. The electronic device (110) may set the judgment result to a withhold state or a warning state according to the exception handling logic. The electronic device (110) may record the exception handling result along with the basis for judgment.

[0318] In at least one embodiment of the present disclosure, the electronic device (110) may record the basis for judgment and the processing path as a log when an edge case occurs. The electronic device (110) may generate the basis for judgment data when an edge case occurs. The electronic device (110) may store the processing path and the basis for judgment in a log file or a database (110). The electronic device (110) may refer to the recorded log in a subsequent step of generating explanatory information. The electronic device (110) may ensure traceability of the judgment process through log recording.

[0319] In at least one embodiment of the present disclosure, the electronic device (110) can evaluate the state of a subject to judgment by comparing a calculated key performance indicator with a reference value. The electronic device (110) can receive a key performance indicator calculated by a KPI calculation engine (253) as an input value. The electronic device (110) can query a reference value stored in a database (110) or internal memory. The electronic device (110) can perform a comparison operation between the key performance indicator and the reference value. The electronic device (110) can evaluate the state based on the comparison result. The electronic device (110) can record the evaluation result as a basis for judgment.

[0320] In at least one embodiment of the present disclosure, the electronic device (110) may classify a state based on a comparison result with a reference value if the reference value is predefined. The electronic device (110) may apply a state classification criterion by referring to the predefined reference value. The electronic device (110) may classify a state based on whether a key performance indicator is greater than, less than, or equal to the reference value. The electronic device (110) may generate a state-specific classification result as a result code or message. The electronic device (110) may transmit the classification result to a subsequent step.

[0321] In at least one embodiment of the present disclosure, the electronic device (110) may set a threshold value and determine a separate state when a key performance indicator exceeds or falls below the threshold value. The electronic device (110) may store the threshold value in a database (110) or in internal memory. The electronic device (110) may classify a specific state when the key performance indicator exceeds the threshold value. The electronic device (110) may determine a separate state when it falls below the threshold value. The electronic device (110) may record whether the threshold value is exceeded or falls below as a basis for judgment.

[0322] In at least one embodiment of the present disclosure, the electronic device (110) may dynamically adjust the range of threshold values ​​or apply different threshold values ​​depending on the domain. The electronic device (110) may dynamically change the range of threshold values ​​according to the characteristics of the input data or the domain. The electronic device (110) may apply different threshold values ​​depending on the domain, such as industry, technology, or market. The electronic device (110) may record the history of threshold value changes in the database (110). The electronic device (110) may store the results of the dynamic threshold value application as a basis for judgment.

[0323] In one embodiment, the electronic device (110) may compare the calculated key performance indicators with external data (e.g., industry average, competitor indicators, etc.) by referring to an external comparison target. The electronic device (110) may receive the external comparison target data from a database (110) or an external system. The electronic device (110) may perform a comparison operation between the key performance indicators and the external comparison target (e.g., industry average, competitor indicators, etc.). The electronic device (110) may further evaluate the state based on the comparison result. The electronic device (110) may record the external comparison result as a basis for judgment.

[0324] In at least one embodiment of the present disclosure, the electronic device (110) may further subdivide the state according to its relative position with respect to an external comparison target. The electronic device (110) may subdivide the state according to whether a key performance indicator is higher, lower, or equal to the external comparison target. The electronic device (110) may generate a result code or message for each subdivided state. The electronic device (110) may transmit the subdivision results to a subsequent explanatory information generation step. The electronic device (110) may improve the precision of the state determination by utilizing the external comparison target.

[0325] In one embodiment, the judgment result may be expressed as one of a plurality of discrete states, such as a stable state, a dangerous state, a growth state, or a slowdown state. The electronic device (110) may predefine a plurality of discrete states (e.g., growth, stable, dangerous, slowdown, etc.). The electronic device (110) may select one of the corresponding states based on input key performance indicators and comparison conditions. The electronic device (110) may convert the selected state into a result code or message. The electronic device (110) may transmit the result code to a subsequent explanation information generation step.

[0326] In at least one embodiment of the present disclosure, the electronic device (110) may provide clear definitions and judgment criteria for each state. The electronic device (110) may clearly define a reference value, threshold value, or comparison target applied for each state. The electronic device (110) may store the judgment criteria for each state in a database (110) or internal memory. The electronic device (110) may update the reference value when the judgment criteria are changed. The electronic device (110) may ensure the reliability of the state judgment through the clear judgment criteria.

[0327] In one embodiment, the electronic device (110) can generate a result code or message for each state and transmit it to a subsequent step. The electronic device (110) can generate a unique result code or message for each state. The electronic device (110) can transmit the generated result code or message to the description information generation module (271). The electronic device (110) can transmit the result code or message to a user terminal (120). The electronic device (110) can record the result code transmission history.

[0328] In at least one embodiment of the present disclosure, the electronic device (110) can implement the entire flow of the state determination step (350) as a flowchart. The electronic device (110) can visualize the input, processing, and output processes of the state determination step (350) in the form of a flowchart. The electronic device (110) can clearly indicate the operation and branching conditions for each step within the flowchart. The electronic device (110) can ensure structural consistency of the determination logic through the implementation of the flowchart. The electronic device (110) can utilize the flowchart information in the explanation information generation step.

[0329] In one embodiment, the flowchart may include a sequential comparison process of input key performance indicators, reference values, threshold values, and external comparison targets. The electronic device (110) may specify the key performance indicator input step within the flowchart. The electronic device (110) may include the sequential comparison process with reference values, threshold values, and external comparison targets in the flowchart. The electronic device (110) may display branching conditions for each comparison step in the flowchart. The electronic device (110) may include the step-by-step flow of the flowchart in the explanatory information.

[0330] In one embodiment, each branching point of the flowchart may branch according to a state-specific judgment criterion. The electronic device (110) may apply the state-specific judgment criterion at each branching point within the flowchart. The electronic device (110) may branch to the next step according to the condition satisfied at the branching point. The electronic device (110) may record the branching result as a basis for judgment. The electronic device (110) may support various state judgment scenarios through the branching structure.

[0331] In one embodiment, the flowchart may include the fact that the order of connecting steps may be changed according to various embodiments. The electronic device (110) may design the order of connecting steps in the flowchart differently for each embodiment. The electronic device (110) may record a history of changes in the order of connecting steps. The electronic device (110) may support multiple state determination methods through various flowchart structures. The electronic device (110) may include flowchart information for each embodiment in the description information.

[0332] In at least one embodiment of the present disclosure, the electronic device (110) may provide a separate processing path for edge cases (e.g., insufficient data, failure to meet criteria, conflicting results, etc.) that may occur during the state determination process. The electronic device (110) may pre-verify whether edge cases, such as insufficient input data, failure to meet criteria, or conflicting results, have occurred. The electronic device (110) may select a separate processing path when an edge case occurs. The electronic device (110) may record the result of the processing path selection as a basis for judgment. The electronic device (110) may store the edge case processing history as a log.

[0333] In at least one embodiment of the present disclosure, the electronic device (110) may set the judgment result to an undefined state when an edge case occurs, or request additional input from the user. The electronic device (110) may designate the judgment result to an undefined state when an edge case occurs. The electronic device (110) may transmit an additional input request message to the user terminal (120) when the undefined state occurs. The electronic device (110) may re-perform the state judgment after receiving the additional input. The electronic device (110) may record the undefined state and the additional input request history.

[0334] In at least one embodiment of the present disclosure, the electronic device (110) may include edge case processing results along with judgment basis in the description information. The electronic device (110) may transmit the edge case processing results along with judgment basis data to the description information generation module (271). The electronic device (110) may include an edge case processing path and results in the description information. The electronic device (110) may transmit the edge case processing history included in the description information to a user terminal (120). The electronic device (110) may ensure transparency and traceability of the judgment results through the edge case processing history.

[0335] In at least one embodiment of the present disclosure, the electronic device (110) may output explanatory information regarding the judgment basis and judgment path corresponding to the judgment result to a user terminal based on the generated judgment result. The electronic device (110) may receive the judgment result generated by the Rule-based judgment engine (255) as an input value. The electronic device (110) may query the judgment basis data corresponding to the input judgment result from the database (110) or internal memory. The electronic device (110) may extract the path information from which the judgment result was derived from the execution history of the judgment logic. The electronic device (110) may transmit the judgment basis and judgment path information to the explanatory information generation module (271). The explanatory information generation module (271) may generate explanatory information by integrating the received judgment result, judgment basis, and judgment path information. The electronic device (110) may transmit the generated explanatory information to the user terminal (120) via a network. By including the judgment result, judgment basis, and judgment path in the explanatory information, the electronic device may enable the user to verify the reliability and transparency of the result.

[0336] In at least one embodiment of the present disclosure, the purpose of generating explanatory information is to enable the user to ensure the reliability and transparency of the result by providing the judgment basis and judgment path together with the judgment result. The explanatory information generation module (271) can integrate the judgment result, the judgment basis, and the judgment path into a single data structure. The explanatory information can enhance the reliability of the result by clearly recording the process and basis for deriving the judgment result. The explanatory information can ensure transparency by explicitly including each step of the judgment path. The explanatory information may also include related information such as the data source, feature vector, KPI, and condition used in the judgment. The explanatory information can be designed to be intuitively verifiable on the user terminal (120).

[0337] In at least one embodiment of the present disclosure, the descriptive information may be configured to identify at least one of the data source used in the judgment, the feature vector, the key performance indicator, and the applied condition, together with the judgment result, thereby ensuring the reproducibility and traceability of the judgment result. The descriptive information generation module (271) may include the identifier of the data source used in the judgment, the value of the feature vector, the key performance indicator (KPI), and the applied condition information in the descriptive information. The descriptive information may clarify the source of the data by recording the unique ID and collection time of each data source. The descriptive information may verify whether the same result is derived for the same input by recording the method of generating the feature vector and the input value in detail. The descriptive information may include detailed information used in the calculation, such as the KPI calculation formula, weights, and parameters. The descriptive information may log condition information, such as the reference value, threshold value, and comparison target applied in the judgment. The descriptive information may guarantee reproducibility and traceability by recording all input, processing, and output steps of the judgment process as metadata.

[0338] In one embodiment, the electronic device (110) can ensure reproducibility by including the input data used for judgment and the processing path in the description information, thereby ensuring that the same result is derived for the same input. The electronic device (110) can record the original value of the input data and the data source ID in the description information. The electronic device (110) can include the processing result of each step of the processing path and the applied logic in the description information. The electronic device (110) can configure the description information to verify whether the same judgment result is generated when the same input data and the same processing path are given. The electronic device (110) can support reproducibility verification by adding metadata such as the processing time, data version, and application rule to the description information.

[0339] In one embodiment, the electronic device (110) can output explanatory information to the user terminal (120) in a visual or other manner. The electronic device (110) can convert the explanatory information into various visual formats, such as text, tables, graphs, and flowcharts. The electronic device (110) can automatically adjust the method of outputting explanatory information according to the display size, resolution, and interface environment of the user terminal (120). The electronic device (110) can distinguish each item of the explanatory information and guide the user step-by-step. In addition to the visual output of explanatory information, the electronic device (110) can provide information in other ways, such as voice guidance and notification messages.

[0340] In at least one embodiment of the present disclosure, explanatory information may be provided by integrating various information, such as result messages, judgment paths, and data sources, so that it can be intuitively understood by the user. An explanatory information generation module (271) may integrate multiple pieces of information, such as result messages, judgment paths, data sources, feature vectors, KPIs, and condition information, into a single piece of explanatory information. The explanatory information may be intuitively conveyed to the user by clearly distinguishing each piece of information. The explanatory information may enhance the user's understanding by visually emphasizing the meaning of the judgment result, the basis for derivation, the data source, the processing path, etc. Considering user experience, the explanatory information may omit unnecessary information and provide a summary of only the core information.

[0341] In at least one embodiment of the present disclosure, the explanation information generation module (271) can generate explanation information regarding the judgment basis and judgment path corresponding to the judgment result based on the judgment result generated by the rule-based judgment engine (255). The explanation information generation module (271) can receive the judgment result received from the rule-based judgment engine (255) as an input value. The explanation information generation module (271) can query the judgment basis data corresponding to the judgment result from the database (110) or internal memory. The explanation information generation module (271) can extract path information from which the judgment result was derived from the execution history of the judgment logic. The explanation information generation module (271) can generate explanation information by integrating the judgment result, the judgment basis, and the judgment path. The explanation information generation module (271) can transmit the generated explanation information to the output module (273).

[0342] In one embodiment, the descriptive information may include a result code or a status message that identifies a judgment result. The descriptive information generation module (271) may generate a result code or a status message corresponding to the judgment result. The result code may be a unique value that can identify the state of the judgment target (e.g., growth, stability, risk, slowdown, etc.). The status message may be in the form of text that can clearly convey the meaning of the judgment result to the user. The descriptive information may be provided to the user terminal (120) by including both the result code and the status message.

[0343] In at least one embodiment of the present disclosure, the explanatory information may include data that visually represents the decision path (e.g., flowchart, step-by-step path information, etc.). The explanatory information generation module (271) may convert each step of the decision path into visual data such as a flowchart, tree structure, or step-by-step list. The explanatory information may diagram the flow of the decision path so that the user can grasp the entire decision process at a glance. The explanatory information may display the input values, processing results, and applied rules for each step as visual elements. The explanatory information may clearly distinguish and visualize the branching, conditions, and results of the decision path.

[0344] In at least one embodiment of the present disclosure, the electronic device (110) can visualize each step of a decision path and output it to a user terminal (120). The electronic device (110) can convert step information of the decision path into various visualization formats, such as graphs, flowcharts, and tables. The electronic device (110) can visually distinguish and display input, processing, and output information for each step of the decision path. The electronic device (110) can output the decision path visualization data to the screen of the user terminal (120). The electronic device (110) can transmit the decision path visualization data in the form of an API response or a message.

[0345] In at least one embodiment of the present disclosure, the descriptive information may be configured to identify at least one of the data source used for judgment, the feature vector, the key performance indicator, and the applied condition. The descriptive information generation module (271) may include the identifier of the data source used for judgment, the value of the feature vector, the key performance indicator (KPI), and the applied condition information in the descriptive information. The descriptive information may record metadata such as the unique ID of each data source, the time of collection, and the data version. The descriptive information may include the method of generating the feature vector, the input value, and the dimension information as metadata. The descriptive information may record detailed information used in calculation, such as the KPI calculation formula, weights, and parameters, as metadata. The descriptive information may leave condition information, such as the reference value, threshold value, and comparison target applied for judgment, as metadata.

[0346] In one embodiment, the electronic device (110) may include metadata of the judgment process (e.g., processing time, data version, applied rule, etc.) in the description information. The electronic device (110) may record information of the time at which the judgment was performed in the description information. The electronic device (110) may include version information for each stage, such as the version of the data source, the time of creation of the feature vector, and the time of KPI calculation, in the description information. The electronic device (110) may record information such as rules, logic, and parameters applied to the judgment as metadata. The electronic device (110) can ensure the reproducibility and traceability of the judgment result through the metadata.

[0347] In at least one embodiment of the present disclosure, the descriptive information may include an identifier (ID) and a timestamp of the data source used for judgment. The descriptive information generation module (271) may record a unique identifier (ID) for each data source in the descriptive information. The descriptive information generation module (271) may record the collection time or creation time of the data source as a timestamp. The descriptive information may include the ID and timestamp for each data source together to clarify the source and time of the data. If multiple data sources are used, the descriptive information may record the ID and timestamp separately for each source.

[0348] In at least one embodiment of the present disclosure, the electronic device (110) may record a unique identifier and collection time information for each data source in the description information. The electronic device (110) may look up the unique ID of the data source in the database (110) or internal memory. The electronic device (110) may extract the collection time or creation time of the data source from the system clock or data metadata. The electronic device (110) may include the ID and timestamp for each data source in the description information to increase the reliability and traceability of the data.

[0349] In at least one embodiment of the present disclosure, the descriptive information may include reference tokens for feature vectors and key performance indicators (KPIs). The descriptive information generation module (271) may record the identifier or reference token of the feature vector used in the judgment in the descriptive information. The descriptive information generation module (271) may include the identifier or reference token of the KPI in the descriptive information. The descriptive information may clarify the association of data used in the judgment process through the feature vector and the reference token of the KPI. If multiple feature vectors or KPIs are used, the descriptive information may record reference tokens separately for each item.

[0350] In at least one embodiment of the present disclosure, the electronic device (110) may include identification information of the feature vector used for judgment and the KPI in the description information. The electronic device (110) may generate a unique identifier or hash value of the feature vector and record it in the description information. The electronic device (110) may include a unique identifier or reference value of the KPI in the description information. The electronic device (110) may ensure the reproducibility and traceability of the judgment result through the identification information of the feature vector and the KPI.

[0351] In at least one embodiment of the present disclosure, the explanatory information may include a log of a comparison condition (e.g., reference value, threshold value, comparison target, etc.) applied to the judgment. The explanatory information generation module (271) may record detailed information of the comparison conditions, such as the reference value, threshold value, and comparison target applied to the judgment, in the explanatory information. The explanatory information may include the value, application time, change history, etc. of each comparison condition in the form of a log. If multiple comparison conditions are used, the explanatory information may record a separate log for each condition. The explanatory information may clarify the basis for deriving the judgment result through the comparison condition log.

[0352] In one embodiment, the electronic device (110) may record detailed information on the conditions used in the judgment process in the description information. The electronic device (110) may include the values ​​of all conditions applied in the judgment process and the application times in the description information. The electronic device (110) may record in detail the history of changes to the condition values, the order of application, and the results for each condition. The electronic device (110) may increase the reliability and reproducibility of the judgment results through condition-specific logs.

[0353] In at least one embodiment of the present disclosure, the output module (273) can output the explanation information generated by the explanation information generation module (271) by transmitting it to the user terminal (120) via a network. The output module (273) can transmit the explanation information received from the explanation information generation module (271) to the user terminal (120) via a network interface. The output module (273) can transmit the explanation information using various network protocols such as TCP / IP, HTTP, and WebSocket. The output module (273) can provide the explanation information to the user terminal (120) in real-time or in batch mode. The output module (273) can monitor whether the transmission of the explanation information is successful and perform retransmission in the event of an error.

[0354] In one embodiment, the electronic device (110) may provide descriptive information to a user terminal (120) in the form of an API response or a message. The electronic device (110) may convert the descriptive information into various response formats such as RESTful API, GraphQL, and message queue. The electronic device (110) may serialize the descriptive information into a standard data format such as JSON, XML, or CSV and transmit it. The electronic device (110) may include result codes, status messages, decision paths, metadata, etc., in the response structure of the descriptive information. The electronic device (110) may provide the descriptive information to the user terminal (120) in real time in the form of an API response or a message.

[0355] In at least one embodiment of the present disclosure, the electronic device (110) can output explanatory information to a user terminal (120) in a dashboard, report, or other visualization format. The electronic device (110) can convert the explanatory information into various visualization formats such as a dashboard screen, report document, graph, table, etc. The electronic device (110) can automatically select the output format of the explanatory information according to the environment of the user terminal (120). The electronic device (110) can visually distinguish and display key information such as judgment results, judgment paths, data sources, and KPIs on a dashboard. The electronic device (110) can configure the explanatory information in a report format with various layouts such as summary, detail, and comparison.

[0356] In at least one embodiment of the present disclosure, the explanatory information may be configured in various formats, such as tables, graphs, and step-by-step paths. The explanatory information generation module (271) may configure the explanatory information into various visual elements, such as tables, graphs, flowcharts, and step-by-step lists. The explanatory information may visualize the trend of change in judgment results, state distribution, path flow, etc., as graphs or tables. The explanatory information may express step-by-step path information as a list or flowchart to aid the user's understanding. The explanatory information may integrate visualization data of various formats and output it to the user terminal (120).

[0357] In one embodiment, the explanatory information may include input data, processing paths, condition information, etc., for verifying the reproducibility of a judgment result. The explanatory information generation module (271) may include the original value of the input data, data source ID, processing path, and application condition information in the explanatory information for verifying the reproducibility of the judgment result. The explanatory information may be used for verifying reproducibility by recording the input, processing, and output results of each step in detail. The explanatory information may be configured to verify whether the same judgment result is derived under the same input and conditions. The explanatory information may also provide test scenarios, input values, expected results, etc., for verifying reproducibility.

[0358] In at least one embodiment of the present disclosure, the electronic device (110) may configure explanatory information to verify whether the same judgment result is derived under the same input and conditions. The electronic device (110) may include input data, processing path, condition information, and judgment result in the explanatory information to enable reproducibility verification. The electronic device (110) may re-execute the judgment logic when the same input data and the same conditions are given to verify whether the result matches. The electronic device (110) may record the reproducibility verification result as a log or output it to a user terminal (120). The electronic device (110) may enhance the reliability and transparency of the system through the reproducibility verification scenario.

[0360] FIG. 4 shows an example of an electronic device according to one embodiment.

[0361] Referring to FIG. 4, the electronic device (130) may include memory (410) and a processor (430).

[0362] The memory (410) can store instructions (e.g., programs) executable by the processor (430). For example, the instructions may include instructions for executing the operation of the processor (430) and / or the operation of each component of the processor (430).

[0363] The memory (410) can be implemented as a volatile memory device or a non-volatile memory device.

[0364] Volatile memory devices can be implemented as DRAM (dynamic random access memory), SRAM (static random access memory), T-RAM (thyristor RAM), Z-RAM (zero capacitor RAM), or TTRAM (Twin Transistor RAM).

[0365] Non-volatile memory devices can be implemented as EEPROM (Electrically Erasable Programmable Read-Only Memory), flash memory, MRAM (Magnetic RAM), Spin-Transfer Torque (STT)-MRAM, Conductive Bridging RAM (CBRAM), FeRAM (Ferroelectric RAM), PRAM (Phase change RAM), Resistive RAM (RRAM), Nanotube RRAM, Polymer RAM (PoRAM), Nano Floating Gate Memory (NFGM), holographic memory, Molecular Electronic Memory Device, or Insulator Resistance Change Memory.

[0366] The processor (430) can process data stored in memory (410). The processor (430) can execute computer-readable code (e.g., software) stored in memory (410) and instructions triggered by the processor (430).

[0367] The processor (430) may be a data processing device implemented in hardware having a circuit having a physical structure for executing desired operations. For example, the desired operations may include code or instructions included in a program.

[0368] For example, a data processing device implemented in hardware may include a microprocessor, a central processing unit, a processor core, a multi-core processor, a multiprocessor, an Application-Specific Integrated Circuit (ASIC), and a Field Programmable Gate Array (FPGA).

[0369] The processor (430) can cause the electronic device (130) to perform one or more operations by executing code and / or instructions stored in memory (410). The operations performed by the electronic device (130) have been described in detail with reference to FIGS. 1 to 3, so redundant descriptions will be omitted below.

[0371] The embodiments described above may be implemented as hardware components, software components, and / or combinations of hardware and software components. For example, the devices, methods, and components described in the embodiments may be implemented using a general-purpose computer or a special-purpose computer, such as, for example, a processor, a controller, an arithmetic logic unit (ALU), a digital signal processor, a microcomputer, a field programmable gate array (FPGA), a programmable logic unit (PLU), a microprocessor, or any other device capable of executing and responding to instructions. The processing unit may execute an operating system (OS) and software applications executed on said operating system. Additionally, the processing unit may access, store, manipulate, process, and generate data in response to the execution of the software. For ease of understanding, the processing unit may be described as being used as a single unit, but those skilled in the art will understand that the processing unit may include multiple processing elements and / or multiple types of processing elements. For example, the processing unit may include multiple processors or one processor and one controller. In addition, other processing configurations, such as parallel processors, are also possible.

[0372] Software may include computer programs, code, instructions, or a combination of one or more of these, and may configure a processing unit to operate as desired or instruct the processing unit independently or collectively. Software and / or data may be stored on any type of machine, component, physical device, virtual equipment, computer storage medium, or device so as to be interpreted by the processing unit or to provide instructions or data to the processing unit. Software may be distributed over networked computer systems and may be stored or executed in a distributed manner. Software and data may be stored on computer-readable recording media.

[0373] The method according to the embodiment may be implemented in the form of program instructions that can be executed through various computer means and recorded on a computer-readable medium. The computer-readable medium may store program instructions, data files, data structures, etc., either individually or in combination, and the program instructions recorded on the medium may be those specifically designed and configured for the embodiment or those known and available to those skilled in the art of computer software. Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks, and magnetic tapes; optical recording media such as CD-ROMs and DVDs; magneto-optical media such as floptical disks; and hardware devices specifically configured to store and execute program instructions, such as ROM, RAM, and flash memory. Examples of program instructions include machine code, such as that generated by a compiler, as well as high-level language code that can be executed by a computer using an interpreter, etc.

[0374] The hardware device described above may be configured to operate as one or more software modules to perform the operation of the embodiment, and vice versa.

[0375] Although the embodiments have been described above with reference to the limited drawings, those skilled in the art can apply various technical modifications and variations based thereon. For example, appropriate results may be achieved even if the described techniques are performed in a different order than described, and / or if the components of the described system, structure, device, circuit, etc. are combined or assembled in a form different from described, or replaced or substituted by other components or equivalents.

[0376] Therefore, other implementations, other embodiments, and equivalents to the claims also fall within the scope of the claims set forth below.

Claims

Claim 1 A method for supporting corporate decision-making based on different types of public data, performed by an electronic device comprising at least one processor, comprising: a step of acquiring different types of industrial data from multiple public data sources and generating an integrity verification value for the acquired industrial data to verify the integrity of the industrial data; a step of normalizing the industrial data according to a data structure predefined based on an industrial domain by referring to an entity classification structure and data field configuration predefined by industrial domain, mapping entity names and field names of the industrial data to standard entities and standard field names respectively based on a field mapping table, converting data types and units, and, if values ​​acquired from two or more public data sources exist for the same data item, selecting one value based on a reliability indicator for each data source or merging the values; a step of sorting the normalized industrial data according to a time standard and generating a multidimensional feature vector to represent at least one state among the industry, technology, or market to which the company belongs in the corporate decision-making process by assigning indicators of the industrial domain, indicators of the technology domain, and indicators of the market domain respectively to different dimensions; a step of calculating the key performance indicators of the company through predefined calculation rules based on the feature vector; rule-based A step of generating a judgment result for the state by determining, through a judgment engine, whether the state corresponds to one of a plurality of discrete states by comparing the key performance indicator with a plurality of preset threshold intervals;A method comprising the step of generating explanatory information regarding a judgment basis and a judgment path corresponding to the judgment result based on a fixed rule or template, and outputting the generated explanatory information to a user terminal, wherein the explanatory information includes an identifier of a data source used to generate the judgment result, the feature vector, the key performance indicator, and the condition applied to the comparison, and is generated with the same content when the industrial data and the plurality of threshold value intervals are identical. Claim 2 The method according to claim 1, wherein the public data source is a publicly accessible data source provided by at least one of a government, an international organization, an industry association, a patent office, an academic institution, or a media company, and includes at least one of policy data, industrial statistical data, technical data, patent data, paper data, or news data. Claim 3 A method according to claim 1, wherein the previously defined data structure is defined based on at least one of an entity classification structure, a data field configuration, or a data schema that is pre-set to correspond to the industrial domain, and the normalization step further comprises the step of recording a schema version of the transformed industrial data. Claim 4 A method according to claim 1, wherein the step of generating the feature vector comprises: a step of calculating a rate of change indicator based on the difference of the normalized industrial data between multiple time points; and a step of further including a cross-domain combined value, which combines the rate of change indicator and indicators of different domains, as an element of the multidimensional feature vector. Claim 5 A method according to claim 1, wherein the step of calculating the key performance indicator comprises the step of applying a weight matrix to a plurality of elements constituting the feature vector and combining them according to the predefined calculation rules, and applying a pre-set normalization coefficient to each item of the key performance indicator to standardize the unit and range, thereby calculating a key performance indicator in the form of an index representing the state of the industry, technology, or market to which the company belongs. Claim 6 A method according to claim 1, wherein the step of generating the judgment result comprises: a step of setting at least one of the plurality of threshold intervals based on an external comparison target periodically updated from the public data source; and a step of classifying the state into any one of the plurality of discrete states according to the priority of the plurality of state classification criteria. Claim 7 A method according to claim 1, wherein the step of outputting to the user terminal comprises the step of configuring the explanatory information to enable verification of whether the same judgment result is derived for the same industrial data and the same conditions by further including in the explanatory information the original value of the industrial data, the identifier of the data source, the processing result and applied logic of each of the normalization step to the step of generating the judgment result, and the unique identifier or hash value of the feature vector. Claim 8 In an electronic device, at least one processor; The electronic device includes a memory for storing instructions, and when the instructions are executed individually or collectively by the at least one processor, the electronic device acquires different types of industrial data from multiple public data sources, generates an integrity verification value for the acquired industrial data to verify the integrity of the industrial data, refers to a predefined entity classification structure and data field configuration for each industrial domain, maps the entity names and field names of the industrial data to standard entities and standard field names respectively based on a field mapping table, and converts data types and units, and when values ​​acquired from two or more public data sources exist for the same data item, selects one value based on a reliability indicator for each data source or merges the values ​​to normalize the industrial data according to a predefined data structure based on the industrial domain, sorts the normalized industrial data according to a time standard, and assigns indicators of the industrial domain, indicators of the technology domain, and indicators of the market domain to different dimensions respectively, thereby generating a multidimensional feature vector to represent the state of at least one of the industry, technology, or market to which the company belongs in the decision-making process of the company, and based on the feature vector, through predefined calculation rules, the core performance of the company Calculate an indicator, and by a rule-based judgment engine, compare the key performance indicator with a plurality of preset threshold intervals to determine whether the state corresponds to one of a plurality of discrete states, thereby generating a judgment result for the state, and generate explanatory information regarding the judgment basis and judgment path corresponding to the judgment result based on a fixed rule or template, and output the generated explanatory information to a user terminal, wherein the explanatory information includes an identifier of the data source used to generate the judgment result, the feature vector,An electronic device comprising the above key performance indicators and the conditions applied to the above comparison, and which is generated with identical content when the above industrial data and the above multiple threshold value intervals are identical. Claim 9 A method according to claim 1, wherein the normalization step is performed in the data collection and normalization layer of the electronic device, the step of generating the feature vector to the step of generating the judgment result is performed in the analysis layer of the electronic device, the step of outputting to the user terminal is performed in the output layer of the electronic device, the analysis layer initiates processing in response to detecting a processing completion signal of the data collection and normalization layer, and the data transmitted between the layers includes version information of the data schema.

Images