System and method for ax / dx competency diagnosis and growth management platform based on item response theory

The system addresses the limitations of conventional AI/DX competency assessment by using IRT and LLMs for adaptive testing and profiling, offering precise, continuous, and personalized competency management.

KR103047115B1Active Publication Date: 2026-11-10김형태
View PDF 4 Cites 0 Cited by

Patent Information

Application Number
KR1020250189536
Authority / Receiving Office
KR · KR
Patent Type
Patents
Current Assignee / Owner
Filing Date
2025-12-03
Publication Date
2026-11-10
Estimated Expiration
2045-12-03

AI Technical Summary

Technical Problem

Conventional AI/DX competency assessment tools lack accuracy and multidimensional analysis, provide one-time assessments, and fail to track long-term competency changes, leading to ineffective learning plans and management.

Method used

A system utilizing Item Response Theory (IRT) for adaptive testing, type-specific profiling, and long-term growth curve tracking, combined with Large Language Models (LLM) for personalized question generation, to dynamically assess and manage AI/DX capabilities.

Benefits of technology

Enhances diagnostic accuracy, provides intuitive multidimensional profiling, supports continuous growth management, and enables data-driven decision-making through real-time competency tracking and personalized learning experiences.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 112025136475388-PAT00001_ABST
    Figure 112025136475388-PAT00001_ABST
Patent Text Reader

Abstract

The present invention relates to an AX / DX competency diagnosis and growth management system and method applying Item Response Theory (IRT). The system according to the present invention includes: an adaptive evaluation unit that dynamically selects and provides the next evaluation item based on the user's response; a profiling unit that analyzes the evaluation results and maps the user to one of a plurality of predefined competency types; and a growth management unit that stores evaluation results obtained at multiple different points in time as time-series data and generates a visualized growth curve representing the user's competency changes based thereon. Through this, the limitations of existing static and uniform diagnostic methods are overcome, and by diagnosing and managing an individual's AX / DX competencies more precisely, three-dimensionally, and continuously, the system has the effect of supporting effective personalized competency development and data-driven human resource management for organizations.
Need to check novelty before this filing date? Find Prior Art

Description

Technology Field

[0001] The present invention relates to the field of artificial intelligence (AI) and digital transformation (DX) capability diagnosis technology. More specifically, it relates to a system and a method of operation that organically combines a Computerized Adaptive Testing (CAT) method applying Item Response Theory (IRT), type-specific profiling based on multidimensional capability analysis, and a long-term growth curve tracking function based on time-series data to precisely diagnose the AX / DX capabilities of individuals and organizations and support customized growth management. Background Technology

[0003] As digital transformation accelerates, securing AX / DX (AI Transformation / Digital Transformation) capabilities for individuals and organizations has emerged as a core competitive advantage. Accordingly, various capability assessment tools have been developed, but conventional technologies have the following obvious limitations. First, most assessment tools rely on a static survey method that presents the same items to all subjects.

[0004] This approach presents problems such as reduced discriminatory power by presenting questions that are too easy or too difficult regardless of the subject's actual competency level, and undermined measurement accuracy by increasing respondent fatigue. Second, since diagnostic results are provided only as fragmentary scores or grades, it is difficult to understand the subject's competency characteristics in a multidimensional manner. In other words, there was a lack of multidimensional analysis and profiling functions capable of comprehensively identifying an individual's strengths, weaknesses, and potential. Third, most diagnoses were limited to one-time assessments, which restricted the ability to track and manage competency changes over time. This makes it difficult to establish long-term learning plans or systematically verify the effectiveness of educational investments.

[0005] As prior art, there exists a 'server and method for generating and providing customized problems based on personalized learning feedback and learning data' (Korean Registered Patent 10-2841906). However, this aims at generating individualized problems and does not target the complex and multidimensional competencies known as AX / DX that the present invention intends to address. Furthermore, the aforementioned prior art lacks the functionality to profile diagnostic results into intuitive types or to visually track and manage long-term growth curves, which is clearly different from the problem that the present invention aims to solve. Therefore, there is an urgent technical need for a competency diagnostic platform of a new paradigm that can overcome the limitations of existing technologies and diagnose and manage users' competencies more precisely, comprehensively, and continuously. The problem to be solved

[0007] The present invention aims to overcome the limitations of static and uniform evaluation methods inherent in conventional technology. Specifically, the first objective is to maximize the accuracy and efficiency of diagnosis by dynamically providing items optimized for an individual's actual competency level. Furthermore, the second objective is to move away from methods that provide results as fragmentary scores or grades, and instead provide type-specific profiles that allow for intuitive understanding by multidimensionally analyzing the user's AX / DX competency characteristics. Moreover, the third objective is to support continuous growth management and enable rational data-driven decision-making by systematically tracking and visualizing competency changes over time, rather than limiting the process to a one-time diagnosis. means of solving the problem

[0009] To solve the above-mentioned problems, an AX / DX competency diagnostic system according to one aspect of the present invention is configured to multidimensionally estimate a user's AX / DX competency based on Item Response Theory (IRT) and to manage the results in conjunction with type-specific profiles and growth curves.

[0010] More specifically, the system of the present invention includes a communication unit that receives response data for AX / DX competency assessment items from a user terminal, and a database that maintains an item pool in which assessment items mapped to a plurality of AX / DX sub-competency dimensions are stored along with IRT parameters including difficulty, discrimination, and guessability. The system comprises an adaptive evaluation unit that estimates a multidimensional competency vector representing a user's potential ability for at least two AX / DX sub-competency dimensions based on the response data and the IRT parameters stored in the item pool, and dynamically selects and provides the next assessment item from the item pool so that the amount of information in the multidimensional competency vector is maximized.

[0011] The adaptive evaluation unit selects and provides one or more starting items based on initial ability estimates, updates the multidimensional ability vector using Bayesian estimation or Maximum Likelihood Estimation techniques whenever a user response to each item is received, and repeatedly selects items from the item pool that maximize Fisher information in the updated multidimensional ability vector. Additionally, by terminating the evaluation when a pre-set measurement standard error converges below a threshold or reaches the maximum number of items, it ensures that measurement reliability above a predetermined level is secured with a minimum number of items.

[0012] The system of the present invention includes a profiling unit that receives the multidimensional capability vector finally estimated by the adaptive evaluation unit as input, and maps the user to one of the plurality of AX / DX capability types by applying a reference profile corresponding to each of the predefined plurality of AX / DX capability types and a weighting parameter reflecting the organizational culture characteristics of the organization to which the user belongs. Here, the AX / DX capability types can be defined as 16 types corresponding to combinations of four dichotomous axes (innovation orientation: EC, execution method: SA, data utilization: DI, organizational role: LP), and the profiling unit places the user's response pattern and multidimensional capability vector into a multidimensional space with axes for each AX / DX sub-competency, and then selects one type that is the optimal fit for the user by considering the distance from the reference vector corresponding to each type and the weighting parameter reflecting the organizational culture characteristics of the organization to which the user belongs.

[0013] In addition, the profiling department can automatically generate a detailed text-based profile for selected AX / DX competency types, including representative characteristics of the type, sub-competencies identified as strengths, sub-competencies requiring improvement, and recommended short-term, medium-term, and long-term action plans, and provide this to the user terminal, thereby presenting the diagnostic results as a three-dimensional profile accompanied by action guidelines rather than simple scores or grades.

[0014] The system of the present invention further includes a growth management unit that stores multidimensional capability vectors and mapped AX / DX capability types corresponding to multiple diagnoses performed on a user at different points in time as time-series data along with timestamps, and generates and provides a visualized growth curve representing the change in AX / DX capability over time for the user based on the time-series data. The growth management unit generates individual growth curves for scores by each AX / DX sub-competency dimension and type transition diagrams representing the change in AX / DX capability types over time from the diagnosis results at multiple points in time, and visualizes them as superimposed or parallel graphs on a user interface, thereby enabling an intuitive understanding of long-term growth trajectories at the individual and organizational levels.

[0015] In addition, the system of the present invention may further include a simulation unit that, for one or more competency development scenarios input by a user or manager, calculates an expected growth curve after scenario execution using investment variables such as training time, training costs, and project resources for each scenario, as well as time series data and a pre-trained prediction model, and calculates and outputs a return on investment (ROI) as the ratio of the competency enhancement effect to the investment variables. Through this, individuals and organizations can compare and evaluate the expected effects of various training and investment strategies in advance and select the most efficient competency development strategy.

[0016] Meanwhile, the database of the present invention further includes a reference data dataset containing reference materials regarding the organizational culture of a specific organization, a specific industry domain, a job group, or a rank system, in addition to the evaluation questions for each sub-competency dimension of AX / DX. The adaptive evaluation unit searches for and selects one or more reference materials highly relevant to the user from the reference data dataset based on the user's responses to self-report questions and multidimensional competency vectors, and provides these reference materials and user responses together to a massive language model (LLM) to automatically generate subsequent adaptive questions using a Retrieval-Augmented Generation (RAG) method to verify the reliability of user responses or confirm the user's actual competencies. The generated subsequent adaptive questions may include situation-based scenario questions to indirectly verify the truthfulness of self-report responses or knowledge- and case-based questions to confirm actual work performance abilities, and the system collects additional responses to these and reflects them in the estimation and updating of multidimensional competency vectors and the mapping of AX / DX competency types. Thus, the present invention combines IRT-based ability estimation with LLM·RAG-based item generation functions to compensate for bias in self-report responses and improve the validity of the diagnosis.

[0017] In another aspect, the present invention provides an AX / DX competency diagnosis method that implements the system. The method comprises the steps of: receiving user responses to AX / DX competency assessment items through a user terminal; estimating a multidimensional competency vector based on the user responses and a stored item pool containing IRT parameters including difficulty, discrimination, and guessing ability; repeatedly performing adaptive assessments by selecting items that maximize the amount of information; a profiling step of mapping a user to one of a plurality of AX / DX competency types based on the final multidimensional competency vector; and a growth management step of generating a visualized growth curve based on diagnosis results from multiple time points. Additionally, after the profiling step, the method may further include the steps of generating a structured prompt including the multidimensional competency vector, the mapped AX / DX competency type, the growth curve, and user metadata, providing it to a large language model, and receiving and outputting a growth solution from an LLM that includes a customized training course, online content, practical project, mentoring program, or career development proposal for the user. Furthermore, after the completion of the training course or project-based learning based on the growth solution, the curriculum of the training course or project, the results of the pre- and post-AX / DX competency assessment of the user, and the time-series growth curve are used as inputs to automatically generate highlight content by the LLM summarizing enhanced core competencies, representative cases, and future maintenance and reinforcement strategies, and provide this to the user terminal and / or organization manager terminal so that it can be utilized for training history management and subsequent growth management.

[0018] In another aspect, the present invention provides a server for performing the AX / DX competency diagnosis. The server includes a communication interface that communicates with a plurality of user terminals and an administrator terminal via a network, a processor, and a memory in which instructions executable by the processor are stored. When the instructions are executed by the processor, the server performs the functions described in the system claim, namely, IRT-based adaptive assessment, multidimensional competency vector estimation, AX / DX type mapping, growth curve generation, simulation, social learning hub integration, LLM·RAG-based follow-up question generation, and customized solution proposal functions. In particular, the server has a multi-tenant structure that manages user accounts belonging to multiple different organizations by tenant, and may further provide an organizational analysis API that calculates the organizational-level AX / DX competency distribution, personnel distribution by type, and organizational average growth curve for each tenant and provides them to the administrator terminal. In addition, the server is deployed on a server cluster composed of multiple physical or virtual server nodes, and by distributing adaptive evaluation, profiling, growth curve generation, and LLM integration processing among different nodes within the cluster and performing them in parallel, it can process real-time AX / DX capability diagnosis requests for multiple users.

[0019] In another aspect, the present invention provides a computer-readable recording medium having a program recorded thereon that enables the AX / DX capability diagnosis method to be performed when executed by a processor.

[0020] The computer-readable recording medium above stores code for executing each step defined in the method claim in combination with hardware, thereby enabling the AX / DX capability diagnostic system to implement the same technical concept in various forms of computing environments, such as servers, cloud environments, or individual terminal devices. Effects of the invention

[0023] According to the present invention, through an adaptive assessment based on item response theory, the ability of a user is measured quickly and precisely with a minimum number of items, thereby dramatically improving the accuracy and efficiency of the diagnosis.

[0024] Second, it has the effect of helping users intuitively and deeply understand their AX / DX competency characteristics through multidimensional competency analysis and specific types of three-dimensional profiling.

[0025] Third, through the function of tracking and managing long-term growth curves, it is effective in systematically monitoring the individual's competency development process and supporting the establishment of data-driven, customized growth roadmaps.

[0026] Fourth, the What-if simulation and ROI calculation functions predict the expected effects of capability development investment in advance, thereby supporting strategic decision-making for individuals and organizations.

[0027] Fifth, it has the effect of enhancing the effectiveness of competency development by providing personalized learning experiences and promoting a collaborative learning culture through the proposal of customized solutions based on Large Language Models (LLM) and a social learning hub. Brief explanation of the drawing

[0029] FIG. 1 is a block diagram showing the overall configuration of an AX / DX capability diagnosis system according to one embodiment of the present invention. FIG. 2 is a flowchart illustrating an IRT-based adaptive evaluation process performed in the adaptive evaluation unit of the present invention. Figure 3 is a conceptual diagram illustrating the process of mapping multidimensional competency scores into 16 types in the profiling unit of the present invention. Figure 4 is a user interface (UI) screen showing an example of visualization of a user capability growth curve generated by the growth management unit of the present invention. FIG. 5 is a flowchart showing the overall flow of an AX / DX capability diagnosis method according to one embodiment of the present invention. Specific details for implementing the invention

[0030] Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the attached drawings. In describing the present invention, if it is determined that a detailed description of related known technologies may unnecessarily obscure the essence of the present invention, such detailed description will be omitted. Furthermore, the embodiments described below are merely examples for understanding the technical concept of the present invention as described in the claims, and the scope of the rights of the present invention is not limited thereto.

[0031] First, FIG. 1 shows an overall configuration diagram of an AX / DX competency diagnostic system (100) according to an embodiment of the present invention. The system (100) includes an adaptive evaluation unit (110) that presents optimal questions based on user responses, a profiling unit (120) that classifies users by multidimensionally analyzing evaluation results, and a growth management unit (130) that tracks and visualizes competency changes over time. Additionally, the system (100) may further include an AI solution proposal unit (140) that generates customized education design plans, a simulation unit (150) that predicts future growth and return on investment based on competency development scenarios, and a social learning hub (160) that supports interaction between users of the same or similar types. Each of the above components operates in conjunction with a database (170) that stores AX / DX evaluation questions and IRT parameters, reference data datasets, time-series diagnostic data, organizational unit metadata, etc., and a communication unit (180) that communicates with a plurality of user terminals and administrator terminals.

[0032] The database (170) includes an item pool (171) in which evaluation items for each AX / DX sub-competency dimension and IRT parameters such as difficulty (b), discrimination (a), and guessability (c) of each item are stored. Additionally, it may further include a reference data dataset (175) in which reference materials regarding the organizational culture of a specific organization, a specific industry domain, a job group, or a rank system are stored; time-series diagnostic data (176) in which multidimensional competency vectors and competency type history are stored by user and time point; and organizational metadata (177) in which information such as the distribution of AX / DX competencies at the organizational unit level, personnel distribution by type, and the organizational average growth curve is stored.

[0033] Additionally, the system (100) may be deployed on a server cluster (190) composed of multiple physical servers or virtual server nodes in actual implementation. The server cluster (190) may have a multi-tenant structure that manages user accounts belonging to multiple different organizations by tenant, and may be configured to logically separate, store, and process data by tenant.

[0034] Next, the operation of the adaptive evaluation unit (110) is described. The adaptive evaluation unit (110) diagnoses the user's AX / DX competency based on Item Response Theory (IRT). A number of evaluation items mapped to multiple AX / DX sub-competency dimensions are stored in the item pool (171) of the database (170), and each item is assigned an IRT parameter in a pre-estimated form that includes difficulty, discrimination, and guessability. In an embodiment of the present invention, the user's competency is expressed as a multidimensional competency vector θ = (θ₁, θ₂, …, θ) composed of at least two AX / DX sub-competency dimensions, where θ represents the potential competency value for the i-th sub-competency dimension and k is the number of AX / DX sub-competency dimensions.

[0035] More specifically, the Item Response Theory (IRT) model used by the adaptive evaluation unit (110) of the present invention is described as follows. In the present invention, each item is modeled as a logistic function in which the probability P(θ) of getting the correct answer is determined according to the respondent's ability level (θ). For a single item, at least one of a 1-parameter logistic model (1PL) that considers only the difficulty (b), a 2-parameter logistic model (2PL) that considers both the difficulty (b) and the discriminability (a), and a 3-parameter logistic model (3PL) that considers the difficulty (b), the discriminability (a), and the guessing ability (c) may be used. In 1PL, the probability of getting the correct answer increases in an S-shape depending on the difference between ability θ and item difficulty b, and in 2PL, as the discriminability a increases, the slope of the correct answer probability curve becomes steeper when the ability level is near the difficulty level, so that the item can better distinguish the ability level. In 3PL, a guessing probability c is added here, reflecting the characteristics of multiple-choice items (e.g., 5-option multiple choice) where even respondents with very low ability can guess the correct answer by chance with a probability of c or higher.

[0036] The curve visualized with the ability axis (θ) as the horizontal axis and P(θ) as the vertical axis, representing the correct answer probability P(θ) defined in this manner, is called the Item Characteristic Curve (ICC). In the system of the present invention, the server internally calculates the corresponding ICC for each item and manages item characteristics by interpreting the difficulty (b) as the point where P(θ) reaches the median value, the discrimination (a) as the slope of the curve near the difficulty level, and the guessing probability (c) as the lower limit of the correct answer probability when the ability is very low. For example, items with high difficulty show an ICC shifted to the right of the ability axis, while items with high discrimination show a curve that rises sharply around the difficulty level. Considering these shapes of the ICC, the system can dynamically correct the quality of the item pool (171) by additionally designing and inserting items with a large ICC slope in a specific ability range when the amount of information is insufficient in that range.

[0037] The amount of information provided by individual items at a specific ability level is quantified by an Item Information Function (IIF). In one embodiment of the present invention, the server calculates the IIF for a 2PL or 3PL model using the correct answer probability P(θ) and the derivative of θ therewith. Generally, the amount of information I(θ) forms a bell-shaped curve that increases as the discrimination (a) increases and as the correct answer probability P(θ) is in the middle region rather than leaning toward 0 or 1. The present system implements an adaptive evaluation strategy to present the most useful items for estimating the respondent's ability by pre-calculating and storing the IIF for each item for multiple points on the ability axis or calculating it in real-time during evaluation, and by preferentially selecting items with a large IIF value from the currently estimated ability vector θ.

[0038] Furthermore, the present invention manages the measurement precision of the overall test level by using a Test Information Function (TIF) that sums the information amounts of individual items. The TIF is defined as the sum of the information amounts of multiple items having an item information function IIF for the same ability level θ, and it means that the larger the TIF in a specific θ range, the smaller the standard error of the test result and the higher the reliability in that range. The server can calculate the TIF for the items included in the item pool (171) and design an item composition strategy such that the TIF is maximized, for example, near the average ability level within the organization or the target ability level. In addition, the shape of the TIF is estimated based on the amount of information accumulated during the evaluation process, and the test efficiency can be dynamically adjusted, such as by suggesting additional items to supplement a specific ability range if the amount of information is insufficient in that range, or by reducing the number of items by satisfying an early termination condition if sufficient information is secured.

[0039] In one embodiment of the present invention, the calculation of the IRT model, ICC, IIF, and TIF may be performed by a machine learning / numerical analysis module within the server. For example, the server may be configured to efficiently calculate the correct answer probability P(θ) and its derivative, the item information function IIF, and the test information function TIF in the form of vector-matrix operations for a predetermined interval of the competency axis θ using a Python-based numerical computation library and matrix computation functions, and to maintain high responsiveness even in real-time adaptive evaluations by caching the calculation results and reusing them during repeated evaluations. Through this, the system of the present invention can stably implement the principle of information quantity optimization based on statistical test theory in an actual AX / DX competency diagnosis service environment.

[0040] When a user starts the AX / DX competency assessment, the adaptive evaluation unit (110) selects one or more starting questions based on an initial competency estimate (e.g., a vector centered at 0 or an estimate based on a preliminary survey and basic information) and presents them to the user terminal. Whenever a response to each question is received from the user terminal, the adaptive evaluation unit (110) updates the user's multidimensional competency vector θ in real time using a statistical estimation technique such as Bayesian Estimation or Maximum Likelihood Estimation.

[0041] FIG. 2 is a flowchart illustrating an IRT-based adaptive evaluation process performed in the adaptive evaluation unit of the present invention.

[0042] For an updated ability vector θ, the adaptive evaluation unit (110) calculates the amount of Fisher information provided by each item and selects an item from the item pool (171) that maximizes the amount of information at the current ability level, and provides it to the user as the next item. This process is repeated as shown in the flowchart in FIG. 2, and the evaluation is terminated when at least one of a pre-set termination condition is satisfied, for example, whether the maximum number of items has been reached or whether the maximum value of the measurement standard error calculated from the covariance matrix of the ability vector estimate converges to a threshold value or lower. By doing so, unlike the conventional method of presenting the same items to all users, the user's AX / DX ability can be measured quickly and accurately with a minimum number of items.

[0043] Another important feature of the present invention is that the adaptive evaluation unit (110) does not simply use existing questions stored in the question pool (171), but combines user responses to self-report questions with reference data datasets (175) and dynamically generates subsequent adaptive questions using a search augmented generation (RAG) method with a large language model (LLM).

[0044] More specifically, when a user responds positively to a self-report question such as “I have experience leading digital collaboration within a team” or “I am familiar with data-driven decision-making,” the adaptive evaluation unit (110) first searches for organizational culture data such as relationship orientation, hierarchy sensitivity, and collectivist tendencies of a specific organization of the user, such as South Korea, and AX / DX job competency data required in the relevant industry and job, taking into account the user profile (organization, industry, job, rank, etc.) and the ability vector θ estimated so far, in the reference data dataset (175), and selects one or more reference materials that are highly relevant to the user.

[0045] Next, the adaptive evaluation unit (110) generates a structured prompt including the selected reference materials, the user's self-reported response content, the currently estimated ability vector θ, and a list of previously presented and answered questions, and provides it to the LLM. The LLM uses the prompt as input to generate a subsequent adaptive question that includes at least one of a situation-based scenario question to indirectly verify the truthfulness of the self-reported response or a knowledge- and case-based question to verify the user's actual work performance ability.

[0046] The adaptive evaluation unit (110) presents a follow-up question generated from the LLM to the user terminal and receives a response to it. Subsequently, the response to the follow-up question is analyzed integrally with the existing IRT-based response data and is reflected in the updating of the multidimensional ability vector θ and the reliability correction for a specific sub-competency dimension. For example, if a user repeatedly makes inconsistent choices in a related situation-based question despite responding with “very high” in the self-report response, this can be reflected by adjusting the estimated ability value or reliability weight of the corresponding sub-competency dimension. In this way, the adaptive evaluation unit (110) of the present invention can significantly improve the reliability and realism of the AX / DX competency diagnosis compared to the existing static self-report survey.

[0047] Next, the configuration and operation of the profiling unit (120) will be described. The profiling unit (120) receives the multidimensional capability vector θ and related metadata finally estimated from the adaptive evaluation unit (110) and maps the user to one of a plurality of predefined AX / DX capability types. FIG. 3 conceptually illustrates this profiling process.

[0048] In an embodiment of the present invention, AX / DX capability types can be defined, for example, as 16 combinations based on four dichotomous axes. More specifically, the system configures diagnostic questions according to the following four axes: (1) Innovation orientation axis: E (Early Adopter) / C (Cautious Validator), (2) Execution style axis: S (Strategic Planner) / A (Agile Executor), (3) Data utilization axis: D (Data Builder) / I (Insight Driver), and (4) Organizational role axis: L (Leader) / P (Participant). A total of 40 diagnostic questions are prepared, with 10 questions for each axis, and the options for each question are pre-tagged with a code (E / C, S / A, D / I, L / P) indicating the polarity value of the corresponding axis. For example, the option “Sign up immediately and test it myself when a new AI tool is released” is tagged as E (Early Adopter) on the innovation orientation axis, while the option “Decision after thoroughly reviewing reviews and user feedback” is tagged as C (Cautious Validator).

[0049] When user responses are collected, the system aggregates the number of responses for the E-tagging option and the C-tagging option for Axis 1 (Innovation Tendency). If the number of E selections is greater than the number of C selections, it assigns E as the code for the Innovation Tendency axis; otherwise, it assigns C. Similarly, for Axis 2 (Execution Method), it compares the number of S responses with the number of A responses to select S or A, for Axis 3 (Data Utilization), it selects D and I, and for Axis 4 (Organizational Role), it selects L and P. By sequentially combining the codes determined for the four axes in this manner, the user is represented as an AX / DX competency type in the form of a 4-character code such as “ESDL”, “CADP”, or “EAIP”, and there are a total of 2⁴=16 possible combinations.

[0050] Meanwhile, the code calculation based on the number of responses as described above is merely one embodiment, and in other embodiments, it may be implemented by calculating the polarities (E / C, S / A, D / I, L / P) of each axis as continuous values ​​using capability vector components for each sub-capability dimension estimated based on IRT, and then performing binary classification according to a predetermined threshold value. For example, the capability dimension θ corresponding to the innovation orientation axis i Regarding this, it is possible to determine E if it is above a certain threshold value and C if it is below, or to calculate the lagging probability of E / C using a Bayesian method by synthesizing the information amount and response patterns of the item group associated with θi, and then determine the axis code by taking the maximum value. By doing so, the simple summation method and the IRT-based continuous estimation method can be selectively applied or combined, allowing for flexible response to various organizational environments and data quality levels.

[0051] For each type, a standard capability vector (prototype vector) for the AX / DX sub-competency dimension and weight parameters reflecting the specific organizational culture characteristics of the organization to which the user belongs, such as South Korea, are pre-set. For example, the ESDL type may be named as “Innovation Leader” or “AI Evangelist,” and a descriptive description such as “a leader-type talent who leads the organization’s AI transformation by strategically introducing new technologies, and executes a systematic innovation roadmap by combining data-driven decision-making and a long-term vision” may be stored as metadata. In this case, for the corresponding type, emphasized strengths such as rapid market responsiveness, strategic leadership, data-driven persuasiveness, and the ability to drive organizational change, as well as cautionary factors such as excessive optimism, a gap in pace with the team, a lack of execution in detail, and the risk of burnout, can be defined together; additionally, information regarding recommended learning paths (e.g., organizational change management, AI governance and ethics, data strategy formulation, innovation leadership programs), suitable job roles (e.g., CDO, CTO, Innovation Executive, AI Transformation PM), and recommended growth strategies (e.g., pace control, supplementing execution, risk management, etc.) can also be pre-registered by type.

[0052] The profiling unit (120) maintains a reference profile for each of the 16 types defined as above, which includes metadata such as (a) a reference capability vector for each AX / DX sub-competency dimension, (b) a combination of four axis codes (e.g., ESDL, CADP, etc.), (c) weighting parameters reflecting specific organizational culture characteristics of the organization to which the user belongs, such as South Korea, (d) a list of strengths and weaknesses, (e) recommended learning and career development paths and suitable job roles, and (f) descriptions of representative behavioral characteristics. When the user's final multidimensional capability vector and four axis codes are calculated, the profiling unit (120) determines the type that best fits the user by considering the vector similarity with the reference profile and the culture weighting parameters together, and automatically generates and provides a text-based detailed profile that includes a detailed description of the type, a summary of strengths and weaknesses, a recommended action plan, and recommended education and career paths. By doing so, the user can intuitively understand a three-dimensional AX / DX capability identity that combines their innovation tendencies, execution methods, data utilization styles, and roles within the organization, rather than a simple score or grade.

[0053] The profiling unit (120) first places the final multidimensional ability vector θ in a k-dimensional space with axes for each AX / DX sub-competency, and then calculates the distance (e.g., Euclidean distance or cosine similarity) from the reference ability vector corresponding to each type. Subsequently, cultural parameters quantified as relationship orientation, hierarchy sensitivity, and collectivist tendencies are applied as weights to select the type that best fits the user by considering not only simple mathematical distance but also suitability in the domestic organizational environment. In cases where there is a slight difference in similarity between multiple types, the final type can be determined by referring to additional information such as job group, rank, and characteristics of the affiliated organization.

[0054] The profiling unit (120) automatically generates a text-based detailed profile for the AX / DX capability type selected as above, including the representative characteristics of the type, sub-capabilities identified as strengths, sub-capabilities requiring improvement, and recommended short-term, medium-term, and long-term action plans, and provides this to the user terminal. This detailed profile can be automatically generated through a predefined type-specific rule-based template or text generation logic using LLM.

[0055] Next, the composition and operation of the Growth Management Department (130) will be described. To support the long-term AX / DX competency development of the user, the Growth Management Department (130) manages and visualizes the results of multiple diagnostic tests performed at different points in time as time-series data. When the user repeatedly performs diagnostic tests at multiple points in time, such as at the time of hiring, before and after promotion, before and after completing training, and at the time of annual evaluation, the multidimensional competency vector θ and the mapped AX / DX competency type for each point in time are accumulated and stored in the time-series diagnostic data (176) along with timestamps.

[0056] The growth management unit (130) retrieves the time series data in response to a request from a user or a request from an administrator terminal and generates individual growth curves that show changes in ability over time for each AX / DX sub-competency. For example, as shown in FIG. 4, time is placed on the horizontal axis and ability on the vertical axis, and growth curves corresponding to each sub-competency, such as data literacy, AI utilization ability, digital collaboration, and strategic thinking, can be superimposed or displayed in parallel.

[0057] Additionally, the Growth Management Department (130) can generate a type transition diagram that intuitively shows the direction in which a user has developed their capabilities, such as “Data-driven Strategist → Digital Collaboration Facilitator → AI Technology Leader,” by connecting the AX / DX capability type codes at each diagnostic point in chronological order. Furthermore, by overlaying the user’s personal growth curve with the growth curve of the same type group or the same organization average stored in the Social Learning Hub (160) or organizational metadata (177), a benchmark function can be provided that allows the user to compare their growth rate and level with the peer group or organizational average.

[0058] Next, the operation of the AI ​​solution proposal unit (140) is described. The AI ​​solution proposal unit (140) generates a customized training design plan for individual users or organizational units using a large language model (LLM). To this end, the AI ​​solution proposal unit (140) generates a structured prompt including user identification information (anonymized if necessary), information on affiliated organization, job, and rank, a final multidimensional capability vector θ, AX / DX capability types and detailed profiles determined by the profiling unit (120), time-series growth curve data generated by the growth management unit (130), and goals of the organization or user (e.g., upgrading specific sub-competencies within one year, preparing for a specific job transition, etc.) and provides it to the LLM.

[0059] Based on the aforementioned structured prompt, LLM generates a customized training plan that includes courses, online content, practical projects, mentoring and coaching programs, and self-study materials suitable for the user. The generated plan may include a list of recommended courses, a recommended sequence, an estimated duration, concurrent practical projects and study activities, and, if necessary, the target and method of mentoring. This customized training plan is provided on the user terminal or an administrator dashboard, and the person in charge can review it and finalize it as an actual training implementation plan after making some modifications or supplements.

[0060] The simulation unit (150) predicts the expected growth curve and return on investment (ROI) after the execution of a scenario based on an educational design plan selected by the user or a separate competency development scenario defined by the user. For example, if the user inputs a scenario such as “completion of a 3-month advanced data analysis course and execution of 2 related projects,” the simulation unit (150) estimates the expected multidimensional competency vector at a specific point in time (e.g., 6 months later, 1 year later, 2 years later) using time-series diagnostic data (176) of existing users who have completed the same or similar education, standard growth patterns by industry, job function, and rank, and statistical values ​​stored in organizational metadata (177), and displays this as an expected growth curve in the form of a graph. In addition, by quantitatively comparing the degree of competency improvement with investment costs such as education costs, education time, resources and labor costs required for project execution included in the scenario, the ROI of the scenario is calculated, and the efficiency between multiple scenarios is provided.

[0061] The social learning hub (160) matches multiple users classified by the profiling department (120) into groups based on the same or similar AX / DX competency types, and provides an online community space within each group where collaborative learning can be performed, such as Q&A, material sharing, and joint study schedule management. For example, users classified as “AI technology leaders” can share the latest AI technology trends or plan joint projects in a group dedicated to that type. The growth management department (130) can calculate the average growth curve and representative success stories for each group and provide them to the social learning hub (160), thereby allowing individual users to be motivated by comparing their own growth curve with the group average.

[0062] FIG. 5 is a flowchart showing the overall flow of the AX / DX competency assessment method of the present invention implemented through the components described above. First, the system receives a response to an AX / DX competency assessment question from a user terminal and performs an adaptive assessment step in which the system repeatedly selects and provides the question that maximizes the amount of information while estimating and updating a multidimensional competency vector θ using a question pool (171) to which IRT parameters are assigned. If a self-report question is included, the system searches a reference data dataset (175), provides the selected reference data, user response, and current competency vector to the LLM to generate subsequent adaptive questions in a RAG manner, and additionally collects responses to this to reflect in the competency vector and reliability correction.

[0063] Next, once the adaptive assessment is completed, the system compares the final multidimensional competency vector θ with the reference profiles and culture weight parameters of the 16 predefined AX / DX competency types to map the user to a specific type and generate and provide a detailed profile. Subsequently, it saves diagnostic results from multiple time points as time-series data and generates growth curves by sub-competency, type transition diagrams, and comparison graphs between individual and group means to provide visual information to users and managers.

[0064] In addition, the system can automatically generate customized training plans and growth solutions for the user by generating a structured prompt containing the final competency vector, competency type, growth curve, user metadata, etc., and providing it to the LLM. For competency development scenarios selected or defined by the user, the simulation unit (150) calculates the expected growth curve and return on investment (ROI) using time-series diagnostic data and a prediction model. After the training course or project-based learning is completed, the LLM automatically generates highlight content summarizing enhanced core competencies, representative cases, and future maintenance and reinforcement strategies using the relevant curriculum, pre- and post-diagnosis results, and growth curve as input, and provides this content to the user and organizational manager so that it can be utilized for training history management and subsequent growth management.

[0065] The AX / DX competency diagnostic system (100) according to the present invention may be implemented in a form including a server or server cluster (190) configured to perform the above-mentioned functions, and the server may include a communication interface that communicates with a plurality of user terminals and administrator terminals through a network, a processor that executes AX / DX diagnostic logic, and a memory in which instructions executable by the processor are stored. The instructions are configured to implement functions such as the above-mentioned adaptive evaluation, profiling, growth curve generation, AI solution proposal, simulation, social learning hub, and organizational analysis API.

[0066] Meanwhile, the AX / DX competency diagnostic system (100) of the present invention does not rely on simple self-report questionnaires or static evaluation items, but includes an experiment and practice-based quantitative and qualitative mixed evaluation structure to scientifically measure actual competencies required in the digital age, including cognition, psychology, behavior, and tool utilization. Specifically, the system (100) integrates the following evaluation modules to measure multidimensional AX / DX competencies, such as the user's cognitive ability (Cognition), psychological state (Psychology), behavioral patterns (Behavior), AI and digital tool utilization skills (Tool Literacy), digital work methods (DX Process), creativity and innovation (Innovation), and ethics and safety (Ethics & Governance).

[0067] 1. Cognitive Experiment

[0068] It provides tasks that evaluate actual thinking strategies, such as problem definition ability, pattern recognition, reasoning ability based on limited information, and hypothesis generation ability. It utilizes experimental techniques based on cognitive psychology and human-computer interaction (HCI), such as the Stroop Test, N-back Test, and problem interpretation scenarios, to collect users' reaction times, selection tendencies, and logical reasoning patterns in real time. Through this, it is possible to objectively evaluate actual thinking strategies and problem-solving methods that are difficult to identify through self-report items.

[0069] 2. Behavioral Task

[0070] It includes tasks that allow observation of the user's actual behavioral patterns, such as maintaining focus, prioritizing, and immediate execution ability. For example, behavioral consistency, habit strength, and environmental control ability are quantitatively evaluated through "5-minute focus test," "distraction removal practice," and "immediate-procrastination choice pattern analysis." The collected behavioral data is accumulated and managed in a time series by the growth management department (130) and used to analyze the trends of individual behavioral changes.

[0071] 3. AI Application Practice (AI Literacy Hands-on Test)

[0072] The user's practical AI utilization capabilities are measured through practice that includes actual AI utilization processes such as creating prompts, evaluating results, and creating improvement prompts. The user's ability to design questions, interpret generated results, and apply problem-solving skills are evaluated quantitatively and qualitatively, reflecting actual digital tool utilization skills that cannot be identified through simple survey questions. The evaluation results are integrated into the calculation of the user's Tool Literacy index and DX Process index in the profiling unit (120).

[0073] 4. Workflow Automation Testing

[0074] It includes practical tasks that measure the ability to recognize repetitive tasks and explore automation opportunities.

[0075] It evaluates users' digital problem-solving skills and ability to adapt to the DX environment through scenario-based automation design and inefficiency exploration practice, and presents a direction for long-term capability development in conjunction with a time-series growth curve.

[0076] 5. AI-based Behavioral Pattern Digital Twin (Digital Cognitive Twin)

[0077] Cognitive, behavioral, AI utilization, and automation design data collected from the above evaluation module are integrated to generate a thinking and behavioral model for individual users. Through this model, it is possible to predict the user's future behavior and problem-solving patterns, provide customized prescriptions and real-time feedback, and link with the customized education design of the profiling unit (120) and the AI ​​solution proposal unit (140).

[0078] Therefore, the AX / DX competency diagnostic system (100) proposed in the present invention, unlike existing self-report survey-based evaluation methods, can scientifically measure user-specific competencies based on actual performance data and support accurate growth prescriptions and education design based on this. This experiment and practice-oriented design is an essential component for objectively evaluating the true thinking ability, behavioral competency, and technology utilization of individuals and organizations in an AI and digital environment, and significantly improves the reliability and real-world suitability of the system.

[0079] Meanwhile, the AX / DX capability diagnosis method of the present invention may be implemented in the form of a computer program executed by a processor of the server or user terminal, and the computer program may be stored on various types of computer-readable recording media such as semiconductor memory, hard disk, SSD, optical disk, flash memory, etc.

[0080] Although preferred embodiments of the present invention have been described above, they are merely examples, and the scope of the present invention is not limited thereto. It is obvious that various modifications, alterations, and substitutions are possible by those skilled in the art within the scope of the technical concept of the present invention as described in the claims. Explanation of the symbols

[0082] 100, entire system 110, Adaptive Evaluation Department 120, Profiling Department 130, Growth Management Department 140, AI Solution Proposal Department 150, Simulation Department 160, Social Learning Hub 170, database 171, Question Pool 175, Reference Dataset 176, Time series diagnostic data 177, Organization Metadata 180, Communications Department 190, server cluster

Claims

Claim 1 A system for diagnosing AX competencies and DX competencies, including one or more of a user's cognition, psychology, behavior, and tool utilization based on Item Response Theory, comprising: a communication unit receiving response data for AX competency and DX competency assessment items from a user terminal; a database maintaining an item pool containing assessment items mapped to multiple AX competency and DX competency sub-competency dimensions, along with Item Response Theory parameters including difficulty, discrimination, and guessability, and a reference data dataset regarding a specific organization; an adaptive evaluation unit estimating a multidimensional competency vector representing the user's potential ability for at least two AX competency and DX competency sub-competency dimensions based on the response data and the Item Response Theory parameters stored in the item pool, and dynamically selecting the next assessment item from the item pool to provide to the user terminal such that the amount of information in the multidimensional competency vector is maximized; and receiving the multidimensional competency vector finally estimated by the adaptive evaluation unit as input, applying a reference profile corresponding to each of a plurality of predefined AX competency and DX competency types and a weighting parameter reflecting the organizational culture characteristics of the organization to which the user belongs, thereby A profiling unit that maps a user to any one of the plurality of AX competency and DX competency types; and a growth management unit that stores the multidimensional competency vector and the mapped AX competency and DX competency types corresponding to multiple diagnoses performed on the user at different points in time as time-series data along with timestamps, and generates and provides a visualized growth curve representing the change over time of the AX competency and DX competency for the user based on the time-series data.An AX competency and DX competency diagnostic system comprising, wherein the adaptive evaluation unit searches for and selects one or more reference materials highly relevant to the user from the reference material dataset based on the user's response to a self-report question and the multidimensional competency vector, provides the searched and selected reference materials and the response to the self-report question together to a large language model to automatically generate subsequent adaptive questions using a search-augmented generation method to verify the reliability of the response to the self-report question or to confirm the user's actual competency, provides the generated subsequent adaptive questions to the user terminal to collect additional responses, and is configured to reflect the analysis results of the additional responses in updating the multidimensional competency vector and mapping AX competency and DX competency types. Claim 2 An AX competency and DX competency diagnostic system according to claim 1, wherein the adaptive evaluation unit selects and provides one or more starting items based on an initial competency estimate, updates the multidimensional competency vector using one or more of Bayes estimation or Maximum Likelihood Estimation techniques whenever a user's response to each item is received, repeatedly selects items from the item pool that maximize Fisher information in the updated multidimensional competency vector, and terminates the evaluation when a pre-set measurement standard error converges to a threshold or lower or reaches the maximum number of items. Claim 3 An AX competency and DX competency diagnostic system according to claim 1 or 2, wherein the plurality of AX competency and DX competency types are defined as pre-set types, and the profiling unit places the multidimensional competency vector in a multidimensional space with axes for each AX competency and DX competency sub-competency, and then selects one type that is optimally suited to the user by considering together the distance from a reference vector corresponding to each of the pre-set types and a weight reflecting the organizational culture characteristics of the organization to which the user belongs. Claim 4 An AX competency and DX competency diagnostic system according to paragraph 3, wherein the profiling unit automatically generates and provides to a user terminal a text-based detailed profile for selected AX competency and DX competency types, including (a) representative characteristics of the type, (b) sub-competencies identified as strengths, (c) sub-competencies requiring improvement, and (d) recommended short-term, medium-term, and long-term action plans. Claim 5 An AX competency and DX competency diagnostic system according to claim 4, further comprising an AI solution proposal unit that generates a structured prompt using the multidimensional competency vector and AX competency and DX competency types derived by the profiling unit and the growth curve generated by the growth management unit as input, and provides the structured prompt to a large language model to generate and provide a customized educational design plan including at least one of an educational course, online content, practical project, and mentoring program suitable for the user. Claim 6 An AX competency and DX competency diagnostic system according to claim 5, wherein the growth management unit generates, together with individual growth curves for scores by sub-competency dimension of each AX competency and DX competency and type transition diagrams representing changes over time in the AX competency and DX competency types from diagnostic results at multiple different points in time, and visualizes them as superimposed or parallel graphs on a user interface. Claim 7 The AX capability and DX capability diagnostic system according to claim 6 further comprises a simulation unit that, for one or more capability development scenarios input by a user or manager, calculates an expected growth curve after the execution of the scenario using investment variables for each scenario, time series data, and a pre-trained prediction model, and calculates and outputs a return on investment (ROI) as the ratio of the capability enhancement effect with respect to the investment variables. Claim 8 An AX competency and DX competency diagnostic system according to claim 7, wherein the system matches multiple users classified by the profiling unit into groups of the same or similar AX competency and DX competency types, and provides a social learning hub that enables communication within each group, and provides a group-specific average growth curve and benchmark indicators linked to the growth curve generated by the growth management unit. Claim 9 delete Claim 10 An AX competency and DX competency diagnostic system according to claim 1, wherein the reference data dataset includes data regarding the characteristics of a specific organizational culture and data regarding AX competency and DX competency job competencies required for each specific industry domain, and the adaptive evaluation unit is configured to generate a subsequent adaptive item including at least one of (a) a situation-based scenario item for indirectly verifying the truthfulness of the self-report response, or (b) a knowledge and case-based item for verifying the user's actual work performance ability, depending on the selection of the reference data dataset. Claim 11 A method for diagnosing AX competencies and DX competencies executed by a computer comprises: a step of receiving response data including user responses to AX competency and DX competency assessment items through a user terminal; a step of estimating a multidimensional competency vector representing a user's potential ability for at least two AX competency and DX competency sub-dimensions based on the user responses and an item pool stored together with Item Response Theory parameters including difficulty, discrimination, and guessability; an adaptive evaluation step of selecting an item from the item pool to maximize the amount of information in the multidimensional competency vector and providing it to the user terminal, and repeating the reception of response data, estimation of the multidimensional competency vector, and selection of items until a pre-set termination condition is satisfied; and a profiling step of mapping the user to one of the plurality of AX competency and DX competency types by comparing the final multidimensional competency vector at the end of the adaptive evaluation step with a reference profile corresponding to each of a plurality of pre-defined AX competency and DX competency types and a weighting parameter reflecting the organizational culture characteristics of the organization to which the user belongs. A growth management step is performed by storing the multidimensional capability vector and mapped AX capability and DX capability types corresponding to multiple diagnoses performed on the user at different points in time as time-series data, and generating a visualized growth curve representing the change over time of the AX capability and DX capability of the user based on the time-series data, wherein the adaptive evaluation step comprises: a step of searching for and selecting one or more reference materials highly relevant to the user based on the response data and a pre-stored reference data dataset regarding a specific organization; and a step of automatically generating subsequent adaptive questions using a search-augmented generative method to verify the reliability of the user's response or confirm the user's actual capability by providing the searched and selected reference materials and the response data to a large language model.A method for diagnosing AX competency and DX competency, characterized by including the step of providing the generated subsequent adaptive items to the user terminal and receiving additional responses thereto to reflect them in the estimation and updating of the multidimensional competency vector. Claim 12 A method for diagnosing AX competency and DX competency according to claim 11, wherein the termination condition in the adaptive evaluation step comprises at least one of (1) a condition of reaching a pre-set maximum number of items, and (2) a condition in which the maximum value of the measurement standard error calculated from the covariance matrix of the estimate for the multidimensional competency vector becomes less than or equal to a predetermined threshold. Claim 13 A method for diagnosing AX competencies and DX competencies according to claim 11 or 12, wherein the method further comprises, after the profiling step, (f) generating a structured prompt including the multidimensional competency vector, mapped AX competency and DX competency types, a growth curve generated in the growth management step, and user metadata, and providing it to a macro-language model; and (g) receiving and outputting a growth solution from the macro-language model that includes at least one of a training course, online content, practical project, mentoring program, or career development proposal customized for the user. Claim 14 delete Claim 15 A method for diagnosing AX and DX competencies according to claim 13, further comprising: a step of automatically generating, by a massive language model, highlight content summarizing core competencies enhanced through said training or project, representative cases, and future maintenance and reinforcement strategies, using the curriculum of said training or project, pre- and post-AX competency and DX competency diagnosis results for said user, and time-series growth curve as inputs after the training or project-based learning according to said growth solution is completed; and a step of providing said highlight content to at least one of a user terminal and an organization manager terminal so that it can be utilized for training history management and subsequent growth management. Claim 16 A server that performs AX capability and DX capability diagnosis, comprising a communication interface that communicates with multiple user terminals and administrator terminals via a network; and a processor; and a memory in which instructions executable by the processor are stored; wherein, when the instructions stored in the memory are executed by the processor, the method comprises: receiving response data including user responses to AX competency and DX competency assessment items from a user terminal; maintaining an item pool in which a plurality of AX competency and DX competency assessment items are stored along with Item Response Theory parameters including difficulty, discrimination, and guessability, and a reference data dataset regarding a specific organization; estimating a multidimensional competency vector representing the user's potential ability for at least two AX competency and DX competency sub-dimensions based on the response data and the Item Response Theory parameters, selecting the next assessment item from the item pool to transmit to the user terminal such that the amount of information in the multidimensional competency vector is maximized, and repeating the process until a termination condition is satisfied to estimate a final multidimensional competency vector; and, based on the final multidimensional competency vector, applying a reference profile corresponding to each of a plurality of predefined AX competency and DX competency types and a weighting parameter reflecting the organizational culture characteristics of the organization to which the user belongs, thereby [determining] the user among the plurality of AX competency and DX competency types Step of mapping to one;A server characterized by being configured to perform the step of storing diagnostic results performed on the user at different points in time as time-series data, generating the user's AX competency and DX competency growth curves based on the time-series data and providing them to the user terminal or administrator terminal, wherein the step of estimating the final multidimensional competency vector includes the step of searching for and selecting one or more reference materials highly relevant to the user based on the response data and the reference material dataset, providing the searched and selected reference materials and the response data to a large language model to automatically generate subsequent adaptive questions using a search-augmented generation method to verify the reliability of the user's response or confirm the user's actual competency, and transmitting the generated subsequent adaptive questions to the user terminal to receive additional responses and reflecting the analysis results of the additional responses in the update of the multidimensional competency vector. Claim 17 In claim 16, the server is characterized by having a multi-tenant structure that manages user accounts belonging to multiple different organizations by tenant, and further providing an organizational analysis API that calculates and provides to an administrator terminal the distribution of AX capabilities and DX capabilities by organization, personnel distribution by type, and an organizational average growth curve for each tenant. Claim 18 In claim 17, the server is characterized by being configured to process real-time AX capability and DX capability diagnostic requests for multiple users by distributing adaptive evaluation, profiling, growth curve generation, and large language model integration processing across different nodes within the cluster and performing them in parallel. Claim 19 A computer-readable recording medium having a program stored therein that enables the execution of an AX capability and DX capability diagnostic method when executed by a processor, wherein the computer-readable recording medium has a computer program stored therein that is combined with hardware to execute the method of claim 11. Claim 20 A computer program stored on a computer-readable recording medium that enables the execution of an AX capability and DX capability diagnostic method when executed by a processor, wherein the computer program stored on the computer-readable recording medium is combined with hardware to execute the method of claim 11.

Citation Information

Patent Citations

  • Apparatus and system for performing adaptive test based item response theory

    KR1020200038046A

  • Computer adaptive multistage testing device and system

    KR1020220057422A

  • System and method for diagnosing digital transformation

    KR1020240105714A

  • Method for providing data analysis platform system using artificial intelligence and record media recorded program for implement thereof

    KR1020250056635A