Corporate lifecycle-based customized b2b matching system

Abstract

The present invention relates to a B2B matching system tailored to the corporate lifecycle, comprising: a collection unit that receives corporate information from a service consumer terminal and receives and stores service provider information, which is information regarding a service provider, from a plurality of service provider terminals; a classification unit that classifies the corporate lifecycle stage based on the corporate information to generate lifecycle information; and a matching unit that selects important characteristics based on the corporate information, the service provider information, and the lifecycle information, and dynamically adjusts a first weight assigned to the important characteristics to perform matching between a service consumer and a service provider. The system includes an output unit that transmits the matching result to the service consumer terminal, wherein the classification unit generates life cycle information by classifying the life cycle stage of a company into one of the following stages: startup, growth, maturity, or decline, based on the company information, and the matching unit includes a preprocessing module that converts the company information, the service provider information, and the life cycle information into analyzable data, a feature selection module that generates feature data by selecting important features based on the preprocessed data, a learning module that trains a machine learning model based on the feature data, and an evaluation module that evaluates the suitability of the service provider through the learned model and performs matching between the service consumer and the service provider, wherein the feature selection module analyzes the features of the preprocessed data to select important features using the RFECV (Recursive Feature Elimination with Cross-Validation) technique, generates feature data by assigning a first weight to the importance of the features based on the life cycle information, and the first weight is dynamically assigned based on the correlation between the life cycle stage and the features, wherein if the life cycle stage is the growth stage, the first weight is recruitment support features and marketing 0.8 is assigned to the solution characteristic, and 0.0 to the cost reduction characteristic.When 2 is assigned and the life cycle stage is the maturity stage, the first weight is assigned 0.9 to the supply chain optimization service characteristic, 0.8 to the cost analysis service characteristic, and 0.2 to the recruitment support characteristic; when multiple conditions are included, the evaluation module sets a priority by integrating the correlation between the multiple conditions and the importance of the characteristics, calculates a second weight according to the formula expressed by Formula 1 below, calculates a matching score based on the second weight, and performs matching between service consumers and service providers; when the life cycle stage is the growth stage, the evaluation module performs matching between cloud solution service providers and information security consulting service providers by calculating the matching score based on the correlation between the condition for introducing cloud-based data analysis solutions and the condition for information security consulting; and when the life cycle stage is the maturity stage, the evaluation module calculates the matching score for the supply chain management service provider higher than the matching score for the quality management consulting service provider based on the correlation between the condition for supply chain optimization, the condition for raw material cost reduction, and the condition for productivity improvement, thereby supply chain management It is characterized by performing matching between service providers and quality management consulting service providers. [Formula 1] (where is the final weight of the characteristic, is the basic importance score of the characteristic, is the correlation score between the condition and another condition, and is an adjustment coefficient that determines the relative contribution of importance and correlation) According to the present invention, by integrally analyzing the life cycle stage and requirements of a company and matching the optimal service provider, the efficiency and reliability of business-to-business (B2B) transactions can be significantly improved.Furthermore, the present invention enhances the suitability of matching results by precisely reflecting the status and needs of companies and strengthens inter-company competitiveness by supporting decision-making based on highly reliable information. By providing a matching system that reflects the status of companies, it enables flexible and accurate matching even in dynamic business environments. Through this, companies can continuously discover new opportunities and promote long-term business growth by establishing stable and reliable trading partnerships.

Description

Technology Field

[0001] The present invention relates to a business lifecycle-tailored B2B (Business-to-Business) matching system. Background Technology

[0002] In the modern business environment, efficiently connecting suitable partners in business-to-business (B2B) transactions has established itself as a key factor in enhancing competitiveness. Companies are increasingly recognizing the need for data analysis and automation technologies to reduce time and costs in the process of finding reliable partners based on operational conditions and requirements, while simultaneously increasing the likelihood of transaction success.

[0003] Currently, commonly used B2B matching methods primarily involve recommending providers through simple data searches and the setting of limited conditions. As this approach fails to comprehensively reflect the status and characteristics of companies, matching results frequently do not align with actual needs. In particular, it has limitations, such as leading to inefficient transactions due to low suitability with providers or failing to satisfy corporate requirements.

[0004] Furthermore, conventional methods have a structural limitation in that they process corporate operational data and provider information separately, failing to analyze them integrally. Consequently, the matching process fails to adequately reflect the company's status and needs, and the lack of real-time analysis and automated matching algorithms hinders a rapid response to market changes. Such inefficient matching methods cause companies to consume excessive time and resources in finding suitable providers.

[0005] Therefore, there is a call for the development of a B2B matching system that comprehensively analyzes the status and needs of companies and automatically connects them with reliable business providers based on this analysis. Meanwhile, prior art related to this is disclosed in Korean Registered Patent Publication No. 10-2606162 (November 28, 2023). The problem to be solved

[0006] The present invention aims to solve the problems of the aforementioned prior art by providing a corporate lifecycle-customized B2B matching system that integrates and analyzes corporate information and service provider information based on the corporate lifecycle stages, automatically selects key characteristics using RFECV techniques and machine learning algorithms, and dynamically assigns weights for each lifecycle stage to quickly and accurately recommend suitable partners in business-to-business (B2B) transactions. means of solving the problem

[0007] The above objective is achieved, according to the present invention, by a collection unit that receives enterprise information from a service consumer terminal and receives and stores service provider information, which is information regarding a service provider, from a plurality of service provider terminals; a classification unit that classifies the life cycle stages of an enterprise based on the enterprise information and generates life cycle information; and a matching unit that selects important characteristics based on the enterprise information, the service provider information, and the life cycle information, and dynamically adjusts a first weight assigned to the important characteristics to perform matching between a service consumer and a service provider. The system includes an output unit that transmits the matching result to the service consumer terminal, wherein the classification unit generates life cycle information by classifying the life cycle stage of a company into one of the following stages: startup, growth, maturity, or decline, based on the company information, and the matching unit includes a preprocessing module that converts the company information, the service provider information, and the life cycle information into analyzable data, a feature selection module that generates feature data by selecting important features based on the preprocessed data, a learning module that trains a machine learning model based on the feature data, and an evaluation module that evaluates the suitability of the service provider through the learned model and performs matching between the service consumer and the service provider, wherein the feature selection module analyzes the features of the preprocessed data to select important features using the RFECV (Recursive Feature Elimination with Cross-Validation) technique, generates feature data by assigning a first weight to the importance of the features based on the life cycle information, and the first weight is dynamically assigned based on the correlation between the life cycle stage and the features, wherein if the life cycle stage is the growth stage, the first weight is recruitment support features and marketing 0.8 is assigned to the solution characteristic, and 0.0 to the cost reduction characteristic.When 2 is assigned and the life cycle stage is the maturity stage, the first weight is assigned 0.9 to the supply chain optimization service characteristic, 0.8 to the cost analysis service characteristic, and 0.2 to the recruitment support characteristic; when multiple conditions are included, the evaluation module sets a priority by integrating the correlation between the multiple conditions and the importance of the characteristics, calculates a second weight according to the formula expressed by Formula 1 below, calculates a matching score based on the second weight, and performs matching between service consumers and service providers; when the life cycle stage is the growth stage, the evaluation module performs matching between cloud solution service providers and information security consulting service providers by calculating the matching score based on the correlation between the condition for introducing cloud-based data analysis solutions and the condition for information security consulting; and when the life cycle stage is the maturity stage, the evaluation module calculates the matching score for the supply chain management service provider higher than the matching score for the quality management consulting service provider based on the correlation between the condition for supply chain optimization, the condition for raw material cost reduction, and the condition for productivity improvement, thereby supply chain management This is achieved by an enterprise lifecycle customized B2B matching system characterized by performing matching between service providers and quality management consulting service providers.[Formula 1]. ( is a characteristic The final weight of, is the basic importance score of the above characteristic, is a condition and other conditions Correlation score between, and (an adjustment coefficient that determines the relative contribution of importance and correlation) In addition, the above-mentioned corporate information includes corporate establishment date information, sales volume information, employee information, industry information, business location information, credit rating information, patent information, certification information, business activity area information, and project execution history information.

[0008] delete

[0009] delete

[0010] delete

[0011] delete Effects of the invention

[0012] According to the present invention, by integrally analyzing the life cycle stages and requirements of a company and matching the optimal service provider, the efficiency and reliability of business-to-business (B2B) transactions can be significantly improved.

[0013] Furthermore, the present invention enhances the suitability of matching results by precisely reflecting the status and needs of companies and strengthens inter-company competitiveness by supporting decision-making based on highly reliable information. By providing a matching system that reflects the status of companies, it enables flexible and accurate matching even in dynamic business environments. Through this, companies can continuously discover new opportunities and promote long-term business growth by establishing stable and reliable trading partnerships.

[0014] Meanwhile, the effects of the present invention are not limited to those mentioned above, and various effects may be included within the scope obvious to a person skilled in the art from the contents described below. Brief explanation of the drawing

[0015] FIG. 1 illustrates an overall corporate lifecycle customized B2B matching system according to an embodiment of the present invention, and FIG. 2 illustrates the connections between the components of a matching unit of a corporate lifecycle-customized B2B matching system according to an embodiment of the present invention, and FIG. 3 illustrates an example of a corporate lifecycle-tailored B2B matching system according to an embodiment of the present invention. Specific details for implementing the invention

[0016] Hereinafter, some embodiments of the present invention will be described in detail with reference to exemplary drawings. It should be noted that in assigning reference numerals to the components of each drawing, the same components are given the same reference numeral whenever possible, even if they are shown in different drawings.

[0017] Furthermore, in describing the embodiments of the present invention, if it is determined that a detailed description of related known configurations or functions would hinder understanding of the embodiments of the present invention, such detailed description is omitted.

[0018] In addition, terms such as first, second, A, B, (a), (b), etc., may be used when describing the components of the embodiments of the present invention. These terms are used merely to distinguish the components from other components, and the essence, order, or sequence of the components is not limited by the terms used.

[0020] From now on, a business lifecycle customized B2B matching system (100) according to an embodiment of the present invention will be described in detail with reference to the attached drawings.

[0021] In the following description, the corporate life cycle customized B2B matching system (100) is abbreviated as the matching system (100).

[0022] As illustrated in FIG. 1, a matching system (100) according to one embodiment of the present invention includes a service consumer terminal (110), a service provider terminal (120), a collection unit (130), a classification unit (140), a matching unit (150), and / or an output unit (160). Alternatively, the matching system (100) may be composed of a collection unit (130), a classification unit (140), a matching unit (150), and / or an output unit (160), excluding the service consumer terminal (110) and / or the service provider terminal (120). In this case, the matching system (100) may receive necessary information from the service consumer terminal (110) and / or the service provider terminal (120) and perform matching between the service consumer and the service provider.

[0023] The service consumer terminal (110) transmits corporate information to the collection unit (130) and can receive service provider matching results from the output unit (160), and is connected to the collection unit (130) and the output unit (160) via a network.

[0024] Specifically, the service consumer terminal (110) can transmit corporate information entered by the service consumer to the collection unit (130) through an interface that can be linked with a web browser-based platform, a dedicated application, or an IoT device, and receive matched service provider information from the output unit (160) and provide it to the service consumer.

[0025] The service consumer terminal (110) can be implemented through electronic devices such as computers, smartphones, tablets, etc., but is not limited to these, and any device that allows the service consumer to input corporate information and check the service provider matching results in real time can be used.

[0026] In this case, corporate information refers to information regarding the general status of a company used to match the optimal service provider by analyzing the company's status and requirements, and may include information on the company's establishment date, sales volume, employees, industry, business location, credit rating, patents held, certifications held, business activity regions, and / or project execution history.

[0027] Information on the date of establishment refers to the year a company was founded and can be used as an indicator to understand the company's history and current status.

[0028] For example, companies established within the last 1 to 3 years may be classified as being in the startup stage, and services such as fundraising, initial market entry consulting, and branding support can be matched with them. On the other hand, companies established for more than 10 years may be classified as being in the mature stage, and advanced services such as organizational management, cost optimization, and support for global expansion can be matched with them.

[0029] Sales volume information, such as annual sales, is information about a company's financial status and can be used as an indicator to evaluate the company's growth potential and stability.

[0030] For example, a company with rapidly increasing sales can be classified as being in the growth stage, and services such as marketing support, productivity improvement solutions, and new market entry strategies can be matched. On the other hand, a company with declining sales can be classified as being in the decline stage, and solution services such as restructuring, cost reduction, and business transformation support can be matched.

[0031] Employee information refers to data regarding a company's workforce size and organizational structure, and can be used as an indicator to evaluate the company's internal operational status and human resource requirements.

[0032] For example, a company with a rapidly increasing number of employees can be classified as being in the growth stage, and services such as recruitment support, employee training, and human resource management solutions can be matched. On the other hand, a company with a decreasing number of employees can be considered to be in the decline stage, and services such as organizational restructuring, retirement management, and efficiency measures can be matched.

[0033] Industry information refers to information about the industrial sector to which a company belongs and its major business activities, and can be used as an indicator to evaluate the company's industrial characteristics and requirements.

[0034] For example, companies in the manufacturing industry can be matched with services related to production process improvement, cost reduction, and equipment maintenance, while companies in the IT industry can be matched with technology-centric services such as software development, data security, and cloud solutions.

[0035] Business location information refers to information regarding a company's physical location and activity area, and can be used as an indicator to evaluate suitability with the service provider's activity area.

[0036] For example, companies operating in a specific region can be matched with services that are well-versed in the market environment and regulatory requirements of that region; furthermore, for services where physical accessibility is critical (e.g., on-site training, facility inspection, etc.), the accuracy and efficiency of the matching can be significantly improved when the company's location matches the service provider's area of ​​activity.

[0037] Credit rating information is data regarding a company's credit status and financial reliability, and can be used as an indicator to assess the possibility of stable transactions.

[0038] Companies with high credit ratings are more likely to establish strategic partnerships or participate in large-scale projects through trust-based collaboration with various service providers, which can lead to matching with relevant service providers. Conversely, companies with low credit ratings may be matched with specialized service providers related to financing, financial structure improvement, and credit recovery support.

[0039] Patent ownership information refers to information regarding patents held by a company and can be used as an indicator to evaluate the company's technological capabilities and innovativeness.

[0040] By analyzing the number and scope of patents and technical originality, it is possible to identify a company's technology commercialization potential, licensing opportunities, and the need for R&D cooperation; furthermore, companies holding numerous patents can be matched with service providers specializing in technology licensing and joint R&D, or receive support for advanced technology commercialization strategies.

[0041] Possessed certification information refers to certifications related to quality, safety, environment, management systems, etc., and can be used as an indicator to evaluate a company's reliability and quality assurance capabilities.

[0042] For example, if a company lacks additional certifications required by a specific industry among the certifications it holds (e.g., ISO 27001 Information Security Management System certification, ISO 45001 Occupational Safety and Health Management System certification), a certification specialist service provider capable of providing or supporting such certifications can be matched.

[0043] Information on business activity regions refers to details regarding a company's primary activity areas and geographical scope, and can be used as an indicator to evaluate market accessibility and regional requirements.

[0044] For example, companies classified as multinational corporations can be matched with service providers capable of supporting entry into global markets, while companies classified as regionally based can be matched with service providers specialized in the characteristics of that region.

[0045] Project execution history information refers to details regarding the types, scale, and performance of projects previously undertaken by a company, and can be used as an indicator to evaluate the company's expertise and experience.

[0046] For example, a company with a history of executing large-scale projects can be matched with a service provider possessing the expertise and capabilities to successfully carry out large projects, while a company with experience in executing multiple projects can be matched with a service provider equipped with a high level of expertise to meet complex requirements or coordinate multiple stakeholders.

[0047] The service provider terminal (120) transmits service provider information to the collection unit (130) and is connected to the collection unit (130) via a network.

[0048] Specifically, the service provider terminal (120) can transmit service provider information entered by the service provider to the collection unit (130) through an interface that can be linked with a web-based platform, a dedicated application, or an IoT device.

[0049] The service provider terminal (120) can be implemented as an electronic device such as a computer, smartphone, or tablet, and is not limited thereto, and any device capable of inputting service provider information can be used.

[0050] In this case, service provider information refers to information regarding the status and capabilities of a service provider used to perform matching between a service provider and a service consumer, and may include service item information, technology information, regional information, project execution history information, evaluation information, certification information, financial information, pricing information, etc.

[0051] Service item information is information regarding the specific types of tasks that a service provider can perform and the services that can be provided, and may include the industry sectors that the service provider can support, technical expertise, and the scope and limitations of the services that can be provided.

[0052] For example, in the manufacturing sector, this may include component design, production and assembly, quality control, and maintenance services, while in the IT sector, it may include software development, cloud infrastructure management, and network security services.

[0053] Possessed technology information is information indicating the technical capabilities possessed by a service provider and their applicability within a specific industry, and may include the service provider's core technologies, research and development capabilities, patent ownership, and technology certification status; it can be used to determine suitability with the technical requirements demanded by service consumers or to perform matching for projects requiring advanced technology.

[0054] Regional information is information specifying the geographical scope within which a service provider can provide services, and may include the service provider's operational area, experience in a specific region, and ability to comply with regulations in that region; it can be used for services where physical accessibility is important (e.g., on-site inspection, facility maintenance, etc.) or for matching to meet the specific requirements of a particular region.

[0055] Project execution history information includes the types, scale, and performance of projects performed by the service provider, and can be used as information to evaluate the service provider's expertise and experience.

[0056] Specifically, project execution history information may include information on the history of large-scale projects in a specific industry, performance indicators after project completion, customer satisfaction, etc., and past experience with similar projects can be used to determine the likelihood of meeting the expectations of service consumers.

[0057] Evaluation information is information regarding the quality level of services previously provided by the service provider and customer feedback, which can be used by service consumers to evaluate the reliability and performance capabilities of the service provider, and may include information such as customer survey results, whether quality certifications have been obtained, and performance indicators during the service provision period.

[0058] Certification information includes information regarding certificates and credentials held by the service provider and can be used to determine compliance with industry standards and international credibility.

[0059] For example, it may include international standards such as ISO 9001 (quality management), ISO 27001 (information security), and CE certification, and can be used to prove the technical capabilities and reliability of the service provider.

[0060] Financial information indicates the financial status and scale of business operations of a service provider and can be used to assess the service provider's ability to maintain long-term cooperative relationships or support large-scale projects.

[0061] Financial information includes annual sales, asset size, debt ratio, and financial stability indicators, and can be used as a basis to guarantee reliability in business-to-business transactions.

[0062] Price information is information indicating the pricing method and cost structure of services provided by a service provider, and can serve as a standard during the negotiation process with service consumers.

[0063] Price information may include the basic cost of the service, price variations based on additional options, and discounts based on service provision conditions.

[0064] The collection unit (130) receives corporate information from a service consumer terminal (110) and can receive and store service provider information, which is information about a service provider, from a plurality of service provider terminals (120). It is connected to the service consumer terminal (110) and the service provider terminal (120) via a network and is electrically connected to the classification unit (140) and the matching unit (150).

[0065] The collection unit (130) receives corporate information in real time from the service consumer terminal (110) and service provider information from the service provider terminal (120), and can store it in a structured data format so that the classification unit (140) and the matching unit (150) can analyze and process it.

[0066] In addition, the collection unit (130) can ensure the security and integrity of the data by undergoing encryption and authentication procedures to safely transmit the collected data through the network.

[0067] Specifically, the collection unit (130) encrypts the data using the Transport Layer Security (TLS) protocol, thereby preventing the data from being exposed to the outside or tampered with during transmission. Additionally, by performing a mutual authentication procedure using digital certificates, it can be ensured that the data is exchanged only within a trusted network.

[0068] The collection unit (130) applies a checksum or Message Authentication Code (MAC) based on a hash algorithm to verify the integrity of the transmitted data, and can support stable data transmission through a data retransmission and error recovery mechanism when an anomaly is detected.

[0069] The classification unit (140) can generate life cycle information by classifying the life cycle stages of a company based on company information received from the collection unit (130), and is electrically connected to the collection unit (130) and the matching unit (150).

[0070] Here, the stages of a company's life cycle are criteria for distinguishing the state of a company from its establishment through growth, maturity, and decline, and may include the startup stage, growth stage, maturity stage, and / or decline stage.

[0071] The startup phase refers to the early stage of a company's establishment, where revenue is unstable and market entry, fundraising, and brand building are the main challenges. For example, a startup established within the last three years may fall into this category.

[0072] The growth stage refers to a state in which a company's revenue and number of employees increase rapidly, market share expands, and business scope expands. For example, it may include companies with annual revenue growing by more than 20% and a rapid increase in the number of employees.

[0073] The maturity stage refers to a state in which a company occupies a stable position in the market and generates continuous profits, even though the growth rates of sales and the number of employees have slowed. For example, a company that has been established for more than 10 years and maintains annual sales above a certain level may fall into this category.

[0074] The decline stage refers to a state where declining sales, increasing costs, and organizational downsizing occur, requiring restructuring or business transformation. For example, this may include companies that have experienced a continuous decline in sales or a reduction in the number of employees over the past three years.

[0075] The classification unit (140) can quantitatively evaluate the state of the company by using a machine learning-based internal algorithm to determine the life cycle stage and comparing it with a pre-set standard for each life cycle stage.

[0076] Here, the machine learning-based internal algorithm may include the processes of feature extraction, quantitative analysis, and / or step classification.

[0077] The feature extraction process is a process of extracting data necessary for classifying life cycle stages, such as the year of establishment, sales growth rate, change rate in the number of employees, and average growth rate by industry. The quantitative analysis process is a process of evaluating the state of the company step by step using statistical criteria and machine learning models such as decision trees and Support Vector Machines (SVM) based on the extracted data. The stage classification process is a process of deriving the most appropriate stage among startup, growth, maturity, and decline based on the analysis results.

[0078] For example, a company in its second year of establishment with annual revenue of less than 100 million won and 10 or fewer employees may be classified as being in the startup stage. On the other hand, a company in its seventh year of establishment with annual revenue of 3 billion won or more, a revenue growth rate of 25%, and an increase in the number of employees to 50 or more may be classified as being in the growth stage. Additionally, a service company in its 15th year of establishment may be classified as being in the decline stage if its revenue has decreased by 10% annually over the past three years and its number of employees has decreased by 20%.

[0079] Life cycle information generated in the classification unit (140) is transmitted to the matching unit (150) and can be used as data to match service providers suitable for each life cycle stage, thereby increasing the accuracy of the matching and supporting the derivation of matching results optimized for the requirements of the company.

[0080] The matching unit (150) can match service consumers and service providers based on corporate information, service provider information and life cycle information, and is electrically connected to the collection unit (130), classification unit (140) and output unit (160).

[0081] As illustrated in FIG. 2, the matching unit (150) includes a preprocessing module (151) that converts corporate information, service provider information, and life cycle information into analyzable data, a feature selection module (152) that selects important features based on the preprocessed data to generate feature data, a learning module (153) that trains a machine learning model based on the selected feature data, and an evaluation module (154) that evaluates the suitability of the service provider through the trained model and performs matching between the service consumer and the service provider.

[0082] The preprocessing module (151) can input transmitted corporate information, service provider information, and lifecycle information as data, convert them into analyzable data through data refinement, and is electrically connected to the collection unit (130), the classification unit (140), and / or the feature selection module (152).

[0083] More specifically, the preprocessing module (151) can normalize the input data to consistently adjust the format of the data. For example, if sales information is input in a mixture of Korean Won (\) and US Dollar ($), it can be converted into a single unified currency unit based on exchange rate information, and the year of establishment input as text can be converted into integer data and stored.

[0084] The preprocessing module (151) can detect missing information within the dataset and compensate for it through a missing value processing process based on a statistical-based compensation method using the mean or median, or a method of generating an estimate through pattern analysis with adjacent data. For example, if sales volume information of a specific company is missing, the data can be replaced by utilizing the average value within the same industry and region.

[0085] Outlier detection and processing are important parts of the preprocessing process, and statistical criteria (e.g., standard deviation, IQR (Interquartile Range), etc.) can be used to identify, correct, or remove abnormal data. For instance, if a value more than 10 times greater than the typical sales level is entered, it can be verified and appropriately corrected to prevent distortion during the analysis process.

[0086] The preprocessing module (151) can increase efficiency by removing unnecessary data from the matching process through filtering. For example, detailed notes or additional information that do not affect the matching can be removed, and only the main data can be processed.

[0087] Format conversion and encoding are other key functions of the preprocessing module (151) that can convert text data or categorical data into numeric data or into an analyzable format. For example, if industry information is input as “manufacturing,” “IT,” and “service,” it is converted into the numbers 1, 2, and 3, respectively, to be prepared for processing by a machine learning model.

[0088] The preprocessing module (151) continuously monitors the input data through real-time data update and verification procedures, and maintains the latest state by immediately reflecting the modified information. Through the verification process, the accuracy and completeness of the data are guaranteed, and errors that may occur during the matching process can be prevented.

[0089] The feature selection module (152) generates feature data by selecting features that play an important role in the matching process based on the preprocessed data received from the preprocessing module (151), and is electrically connected to the preprocessing module (151) and the learning module (153).

[0090] Specifically, the feature selection module (152) can analyze all features of the data received from the preprocessing module (151) and use the RFECV technique to separate the main features that affect the matching result from the features that do not, thereby selecting important features.

[0091] Here, RFECV (Recursive Feature Elimination with Cross-Validation) refers to a method that selects the optimal combination of features by repeatedly removing unimportant features based on the performance of a machine learning model, and derives the optimal result by evaluating generalization performance through cross-validation.

[0092] The RFECV technique evaluates the importance of each feature on the matching result, allowing for the iterative removal of features with low importance or low correlation and the identification of the optimal feature combination.

[0093] The feature selection module (152) evaluates the correlation between corporate information and service provider information and can generate feature data by assigning weights to the importance of the features based on life cycle information.

[0094] Additionally, the feature selection module (152) performs weighting to increase data analysis and matching accuracy, thereby allowing processing that reflects the relative importance of each feature to the matching result.

[0095] The weights are set based on the association between a specific life cycle stage and the characteristics of the company, and by giving high priority to important characteristics, it is possible to derive matching results with a higher degree of fit in the learning module (153) and evaluation module (154) described later.

[0096] More specifically, the characteristic selection module (152) can evaluate the importance of each characteristic and then dynamically assign weights according to the life cycle stage.

[0097] For example, in the case of an IT company that has been established for four years and is recently seeing a rapid increase in the number of employees and is aiming to enter a new market, the characteristic selection module (152) analyzes the company's life cycle information and corporate information and assigns a high weight (0.8) to the characteristics of recruitment support and marketing solutions, while assigning a low weight (0.2) to the characteristics of cost reduction, which are relatively less important. As a result, recruitment-related service providers and digital marketing service providers can be recommended preferentially in the matching results.

[0098] As another example, in the case where a manufacturer established for more than 15 years maintains stable sales but profitability is declining due to a recent increase in raw material costs, the characteristic selection module (152) determines that the company is in a mature stage and that characteristics related to cost reduction are most important, and may assign high weights of 0.9 and 0.8 to supply chain optimization and cost analysis service characteristics, respectively. On the other hand, a low weight of 0.2 may be assigned to recruitment support characteristics, which are less important in the mature stage.

[0099] The weighting process adjusts the relative importance of characteristics based on life cycle information and corporate information, and can be designed so that the matching result closely matches the actual requirements of the company, thereby allowing the characteristic selection module (152) to maximize the efficiency and accuracy of the data and contribute to deriving the optimal matching result.

[0100] Additionally, the feature selection module (152) can perform processing of categorical data. Values ​​input as categorical data, such as industry, region, and service type, are converted into a form that can be analyzed by a machine learning model, and the impact of each category on the matching result can be independently evaluated.

[0101] For example, characteristics considered important in manufacturing companies (e.g., cost reduction, productivity improvement) may not be important in IT companies, and characteristics can be selected to reflect these differences.

[0102] Meanwhile, the feature selection module (152) can improve data efficiency by analyzing multicollinearity among features in the data to identify features with high correlation and removing redundant information.

[0103] Multicollinearity is a phenomenon in which one independent variable in a multiple regression model has a linearly predictable relationship with one or more other independent variables. It occurs when two or more characteristics have a high correlation with each other, and if such characteristics are included in a machine learning model, it can reduce the reliability of the analysis results.

[0104] For example, the sales growth rate and the rate of change in the number of employees are characteristics closely related to corporate growth, and it is highly likely that the two characteristics will have a high correlation simultaneously. In this case, if both characteristics are maintained, unnecessary redundant information may be included in the analysis, and the machine learning model may overfit the data. To prevent this, the feature selection module (152) can evaluate the correlation between the two characteristics using a correlation coefficient.

[0105] Specifically, if the correlation coefficient is measured to a high value of 0.9 or higher, it can be designed to remove one of the two characteristics, and in this process, the more meaningful characteristic can be selected by reflecting the purpose of the model and the requirements of the company.

[0106] For example, if the annual sales growth rate of a manufacturer established five years ago is 25% and the change rate in the number of employees is 20%, and the correlation coefficient between the two characteristics is analyzed to be 0.93, the characteristic selection module (152) can reduce data redundancy by maintaining the sales growth rate and removing the change rate in the number of employees.

[0107] The learning module (153) trains a machine learning model based on feature data received from the feature selection module (152) and is electrically connected to the feature selection module (152) and the evaluation module (153).

[0108] The learning module (153) can analyze the interaction between corporate information, service provider information and life cycle information, and generate a machine learning model that can predict the suitability between the service consumer and the service provider.

[0109] The learning module (153) uses selected feature data as input values ​​to train a machine learning algorithm, and the algorithms used at this time may include a Decision Tree, Random Forest, Support Vector Machine, and a deep learning-based Neural Network model, and can perform hyperparameter tuning and cross-validation to optimize the prediction performance of the model.

[0110] Additionally, the learning module (153) can verify the performance of the learned model using test data along with the learning data.

[0111] Test data can be used to calculate evaluation metrics such as model accuracy, precision, recall, and F1 score by comparing actual matching results with prediction results; based on these metrics, the trained model is optimized and can ultimately be used to evaluate the matching fit.

[0112] For example, if a company in the growth stage is analyzed to have a sales growth rate of 20% and an employee growth rate of 15%, and receives characteristics that have high importance regarding marketing support and recruitment-related services, the learning module (153) can use these characteristic data as input data to learn the matching records between companies and service providers with similar conditions in the past. For instance, if a service provider that has provided digital marketing services in the past has recorded high satisfaction with similar companies, the learned model can be trained to recommend that service provider preferentially.

[0113] In addition, the learning module (153) can update the model in real time when new data is input to reflect the latest requirements and market trends. If there are newly emerging requirements or trends in a specific industry, the suitability of the matching results can be continuously improved through a learning process that reflects them.

[0114] The evaluation module (154) can evaluate the suitability of a service provider through a machine learning model learned in the learning module (153), perform matching between a service consumer and a service provider, and is electrically connected to the learning module (153) and the output unit (160).

[0115] The evaluation module (154) analyzes the prediction results of the learned model based on the input data, and through this, can calculate the correlation between each characteristic quantitatively to calculate the matching score.

[0116] Specifically, the evaluation module (154) compares all characteristics of the data to generate results that accurately reflect the current state and potential needs of the company, and performs the role of increasing the suitability of the matching by dynamically adjusting the importance of specific characteristics. In this process, the evaluation module (154) can evaluate whether specific conditions match the needs and life cycle stages of the company by comparing all characteristics of the data.

[0117] For example, if it is determined through an analysis of manufacturer A, for whom patent commercialization is important, that consulting services related to the commercialization of new production technology and compliance with local regulations are required, the learned model analyzes data that was successfully matched under similar conditions in the past to recommend a service provider with high suitability, and the evaluation module (154) uses this score to comprehensively compare regional accessibility and patent commercialization experience to adjust the matching score, and a service provider with extensive experience in patent commercialization consulting and regulatory compliance is selected as the matching result.

[0118] As another example, if it is determined through analysis based on the project execution history and credit rating information of Service B, a company with a low credit rating, that funding support for credit enhancement is needed along with an initial market entry strategy, the learned model can analyze the correlation between the credit rating and the funding success rate through past data and calculate the suitability of financial institutions and consulting firms. The evaluation module (154) adjusts the matching score by comparing the financial institution's past funding success cases with the consulting firm's initial market entry experience, and can derive a matching result with the financial institution by determining that credit enhancement is the most important requirement.

[0119] Meanwhile, the evaluation module (154) can derive the optimal matching result by integrating the correlation between each condition and the importance of each characteristic when multiple conditions are included to set priorities.

[0120] The evaluation module (154) can set weight-based priorities by integrating correlation data and importance by characteristic through Equation 1, which dynamically adjusts weights by condition.

[0121] {Formula 1}

[0122]

[0123] Here is a characteristic The final weight of, is the base importance score of the corresponding characteristic, is a condition and other conditions Correlation score between, and is an adjustment coefficient that determines the relative contribution of importance and correlation.

[0124] The evaluation module (154) can calculate the final matching score based on the weights of each priority-set characteristic and can apply a machine learning algorithm to reflect the influence of the interaction between conditions on the matching result.

[0125] Through this, the evaluation module (154) can not only perform independent calculations between the defined conditions but also analyze the overall effect of the interaction between the conditions on the matching suitability to derive the optimal result.

[0126] For example, if the analysis of patent information and industry information of startup company C in the growth stage reveals that the introduction of a cloud-based data analysis solution is a major task and that consulting services to strengthen information security are simultaneously required, the trained model calculates a suitability score for each of the cloud solution provider and the information security consulting provider, and the evaluation module (154) determines that the introduction of the cloud solution is a major task directly linked to improving data analysis efficiency and may assign a high score to the solution provider. However, in a situation where security regulations are strengthened, information security consulting is also considered important, so both the cloud solution provider and the information security consulting provider may be included in the matching result.

[0127] As another example, if manufacturer D, which has entered the maturity stage, is a company that satisfies the conditions of the maturity stage, such as stable sales and market share, but it is determined through the analysis of production process data that there is a lack of supply chain efficiency resulting in increased raw material costs and decreased productivity, the evaluation module (154) may analyze that supply chain optimization services would be more suitable for solving this company's problems than cost reduction solutions generally recommended for companies in the maturity stage. The learned model calculated the suitability of supply chain management experts and quality management consultants, and the evaluation module (154) may assign a high matching score to supply chain management experts, judging that supply chain optimization plays a more important role in solving raw material cost issues and improving production processes, while at the same time, quality management consultants may be additionally included in the matching results for complementary improvements to the production process.

[0128] The output unit (160) can transmit the matching result received from the matching unit (150) to the service consumer terminal (110) and is electrically connected to the service consumer terminal (110) and the matching unit (150).

[0129] The output unit (160) can convert the matching result received from the matching unit (150) into a form that is easy for the service consumer to understand and transmit it to the service consumer terminal (110), thereby delivering the derived matching result to the service consumer, and may include a score indicating the reliability and suitability of the matching result, the reason for the recommendation, and detailed information of the service provider.

[0130] Matching results can be output in various formats such as text, graphs, and charts, and may include detailed information or provide summarized information depending on the requirements of the service consumer.

[0131] For example, along with the suitability score of the matched service provider, a chart visualizing the key characteristics used in the matching (e.g., location information, technology information, cost information) and the extent to which those characteristics contributed to the matching result can be provided.

[0132] Additionally, the output unit (160) includes a real-time data transmission function, so that as soon as the matching result is obtained, it can be immediately transmitted to the service consumer terminal (110). Through this, the service consumer can check the matching result in real time and, if necessary, take immediate follow-up action.

[0133] Meanwhile, the output unit (160) ensures the integrity and security of the data through encryption and authentication procedures during the data transmission process, and can prevent the matching result from being exposed to the outside or tampered with.

[0134] Specifically, the output unit (160) applies a Transport Layer Security (TLS) protocol to maintain the confidentiality of the transmitted data and prevent eavesdropping and tampering on the network path through which the data is transmitted.

[0135] Additionally, the output unit (160) uses a digital certificate to authenticate the source and recipient of the data, and can strengthen security procedures so that data is transmitted only to authenticated service consumer terminals (110).

[0136] The output unit (160) may utilize a hash algorithm-based checksum or a Message Authentication Code (MAC) to verify the integrity of the transmitted data. Through this, it can quickly detect whether the transmitted data has been tampered with during reception and, if an anomaly occurs, perform data retransmission or error handling procedures.

[0137] Additionally, the output unit (160) can encrypt the data by applying the Advanced Encryption Standard (AES) algorithm, which is an encryption standard, to protect the sensitive matching result data. Through this, security can be enhanced so that even if the matching result data is stolen due to an external attack, it cannot be decrypted.

[0138] As described above, according to the matching system (100) according to the embodiment of the present invention, by integrally analyzing the life cycle stage and requirements of a company and matching the optimal service provider, the efficiency and reliability of business-to-business (B2B) transactions can be significantly improved.

[0139] Furthermore, the present invention enhances the suitability of matching results by precisely reflecting the status and needs of companies and strengthens inter-company competitiveness by supporting decision-making based on highly reliable information. By providing a matching system that reflects the status of companies, it enables flexible and accurate matching even in dynamic business environments. Through this, companies can continuously discover new opportunities and promote long-term business growth by establishing stable and reliable trading partnerships.

[0141] Although all components constituting the embodiments of the present invention have been described above as being combined or operating together, the present invention is not necessarily limited to such embodiments. That is, within the scope of the purpose of the present invention, all components may be selectively combined and operated in one or more ways.

[0142] Furthermore, terms such as "include," "compose," or "have" described above, unless specifically stated otherwise, mean that the relevant component may be inherent; therefore, they should be interpreted as allowing for the inclusion of additional components rather than excluding them. All terms, including technical or scientific terms, have the same meaning as generally understood by those skilled in the art to which the present invention pertains, unless otherwise defined. Commonly used terms, such as those defined in advance, should be interpreted in accordance 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 the present invention.

[0143] Furthermore, the above description is merely an illustrative explanation of the technical concept of the present invention, and those skilled in the art to which the present invention pertains will be able to make various modifications and variations within the scope of the essential characteristics of the present invention.

[0144] Accordingly, the embodiments disclosed in this invention are intended to illustrate, not limit, the technical concept of the invention, and the scope of the technical concept of the invention is not limited by these embodiments. The scope of protection of this invention shall be interpreted by the claims below, and all technical concepts within an equivalent scope shall be interpreted as being included within the scope of rights of this invention. Explanation of the symbols

[0145] 100: Enterprise lifecycle customized B2B matching system according to an embodiment of the present invention 110: Service consumer terminal 120: Service provider terminal 130: Collection Department 140: Classification section 150: Matching section 151: Preprocessing Module 152: Attribute Selection Module 153: Learning Module 154: Evaluation Module 160: Output section

Claims

Claim 1 A collection unit that receives corporate information from a service consumer terminal and receives and stores service provider information, which is information about a service provider, from a plurality of service provider terminals; a classification unit that classifies the life cycle stages of a company based on the corporate information and generates life cycle information; and a matching unit that selects important characteristics based on the corporate information, the service provider information, and the life cycle information, and performs matching between a service consumer and a service provider by dynamically adjusting a first weight assigned to the important characteristics. The output unit transmits the matching result to the service consumer terminal, and the classification unit classifies the life cycle stage of a company into one of the following stages based on the company information: startup stage, growth stage, maturity stage, or decline stage to generate the life cycle information. The matching unit includes a preprocessing module that converts the company information, the service provider information, and the life cycle information into analyzable data; a feature selection module that selects important features based on the preprocessed data to generate feature data; a learning module that trains a machine learning model based on the feature data; and an evaluation module that evaluates the suitability of the service provider through the learned model and performs matching between the service consumer and the service provider. The feature selection module analyzes the features of the preprocessed data to select important features using the RFECV (Recursive Feature Elimination with Cross-Validation) technique, and generates the feature data by assigning the first weight to the importance of the feature based on the life cycle information. The first weight is dynamically assigned based on the correlation between the life cycle stage and the feature, wherein if the life cycle stage is the growth stage, the first weight is recruitment support feature and marketing 0.8 is assigned to the solution characteristic, and 0.0 to the cost reduction characteristic.When 2 is assigned and the above life cycle stage is the maturity stage, the above first weight is assigned 0.9 to the supply chain optimization service characteristic, 0.8 to the cost analysis service characteristic, and 0.2 to the recruitment support characteristic; when the above evaluation module includes multiple conditions, it sets a priority by integrating the correlation between the multiple conditions and the importance of the characteristics, calculates a second weight according to the formula expressed by Formula 1 below, calculates a matching score based on the second weight, and performs matching between service consumers and service providers; when the above life cycle stage is the growth stage, the above evaluation module performs matching between cloud solution service providers and information security consulting service providers by calculating the matching score based on the correlation between the condition for the introduction of cloud-based data analysis solutions and the condition for information security consulting; and when the above evaluation module is the maturity stage, it calculates the matching score for supply chain management service providers higher than the matching score for quality management consulting service providers based on the correlation between the condition for supply chain optimization, the condition for raw material cost reduction, and the condition for productivity improvement, thereby supply chain management A corporate lifecycle-tailored B2B matching system characterized by performing matching between service providers and quality management consulting service providers.[Formula 1]. ( is a characteristic The final weight of, is the basic importance score of the above characteristic, is a condition and other conditions Correlation score between, and (an adjustment coefficient that determines the relative contribution of importance and correlation) Claim 2 A B2B matching system tailored to the corporate lifecycle according to claim 1, wherein the corporate information includes corporate establishment date information, sales volume information, employee information, industry information, business location information, credit rating information, patent information, certification information, business activity region information, and project execution history information. Claim 3 delete Claim 4 delete Claim 5 delete

Images