KR-20260064860-A - DATA INTEGRATION ANALYSIS SYSTEM BASED ON AI AND METHOD PERFROMING THEREOF

KR20260064860AKR 20260064860 AKR20260064860 AKR 20260064860AKR-20260064860-A

Abstract

The AI-based data integration analysis method according to the present invention comprises the steps of: providing medical data having different structures and formats to an AI-based data integration analysis server from a plurality of medical institution servers; collecting and processing the medical data received from the plurality of medical institution servers and storing it in a database; when a user terminal accesses the AI-based data integration analysis server and inputs a command for the database, the AI-based data integration analysis server generates a query that links to a table in the database or extracts data with conditions defined by the user; displaying the process of the user terminal generating the query by the AI-based data integration analysis server; modeling the data extracted according to the execution result of the query and training an AI model, and providing the AI-based data integration analysis server; and displaying the modeling process of the data extracted according to the execution result of the query and the AI analysis results to the user terminal.

Inventors

김상태
김기수
김대화
김연규
신용석

Assignees

주식회사 에이다루트

Dates

Publication Date: 20260508
Application Date: 20241029

Claims (10)

Multiple medical institution servers providing medical data with different structures and formats; An AI-based data integration analysis server that collects and processes medical data received from the aforementioned multiple medical institution servers and stores it in a database, generates a query to connect tables in the database or extract data with user-defined conditions, and provides an AI model trained after modeling using the extracted data based on the execution results of the query; Characterized by including a user terminal that displays the process of generating a query by accessing the AI-based data integration analysis server and inputting commands for the database, and displays the modeling process of extracted data and AI analysis results according to the execution result of the query. AI-based data integration analysis system.
In paragraph 1, The above user terminal When accessing the AI-based data integration analysis server, a user interface is displayed, and when the generation of a first query is requested, fields associated with the table are displayed together for selection, and when the generation of a second query is requested, if specific conditions are entered by the user, the process of grouping data matching those conditions is displayed. AI-based data integration analysis system.
In paragraph 1, The above user terminal Characterized by displaying data extracted according to the above query through a visualization tool, displaying a dataset prepared as training data to be used for AI analysis, and allowing an AI model to start learning through it. AI-based data integration analysis system.
In paragraph 1, The above AI-based data integration analysis server is Characterized by generating and executing an integrated query through join conditions and relationship definitions via a graphical user interface that can visually define relationships between multiple tables. AI-based data integration analysis system.
In paragraph 1, The above AI-based data integration analysis server is Characterized by generating and executing a query based on the search conditions after the user selects a desired column in the above database and the search conditions required for data retrieval are set. AI-based data integration analysis system.
A step of providing medical data with different structures and formats from multiple medical institution servers to an AI-based data integration analysis server; A step in which an AI-based data integration analysis server collects medical data received from the plurality of medical institution servers, processes it, and stores it in a database; When a user terminal connects to the AI-based data integration analysis server and inputs a command for the database, the AI-based data integration analysis server generates a query that connects to tables in the database or extracts data with user-defined conditions; A step of displaying the process in which the user terminal generates a query by the AI-based data integration analysis server; A step in which the AI-based data integration analysis server models using data extracted according to the execution result of the query, and then trains and provides an AI model; and The above user terminal is characterized by including a step of displaying the modeling process of the extracted data and the AI analysis results according to the execution result of the above query. AI-based data integration analysis method.
In paragraph 6, A step of displaying a user interface when the above user terminal connects to an AI-based data integration analysis server; A step in which, when the user terminal requests the generation of a first query, fields associated with the table are displayed together for selection; The above user terminal is characterized by including a step of displaying a process of grouping data that matches a specific condition when a user inputs a specific condition upon request for the generation of a second query. AI-based data integration analysis method.
In paragraph 6, The step of the above user terminal displaying the data extracted according to the above query through a visualization tool; and The above user terminal is characterized by including a step of displaying a dataset prepared as training data to be used for AI analysis and initiating training of an AI model through it. AI-based data integration analysis method.
In paragraph 6, The above AI-based data integration analysis server is characterized by including the step of generating and executing an integrated query through join conditions and relationship definitions via a graphical user interface that can visually define relationships between multiple tables. AI-based data integration analysis method.
In paragraph 6, The above AI-based data integration analysis server is characterized by including a step of generating and executing a query according to the search conditions when the user selects a desired column in the above database and the search conditions required for data retrieval are set. AI-based data integration analysis method.

Description

AI-based data integration analysis system and method for executing the same The present invention relates to an AI-based data integration analysis system and a method for executing the same. More specifically, it relates to an AI-based data integration analysis system and a method for executing the same that allows even non-experts to easily analyze data and generate data necessary for AI learning, thereby reducing the time required for data analysis and enabling complex data analysis tasks to be performed without relying on experts. In the 21st century, data has become a central element of digital innovation, and effectively analyzing and utilizing it is directly linked to the competitiveness of companies and organizations. As most business activities have become digitized, massive amounts of data are generated in real time, and its scope and forms have become highly diverse, including text, images, videos, and sensor data. Particularly in various industries such as finance, healthcare, retail, and manufacturing, the importance of data has surged, making the effective collection, processing, and analysis of it an essential task. By analyzing this massive amount of data, companies can understand consumer behavior patterns, optimize products and services, and explore new market opportunities. However, as the volume of data has increased, effectively processing and analyzing it has become more difficult. Data does not exist in a single format; instead, structured data (structured data retrieved from SQL databases) and unstructured data (social media posts, log files, image data, etc.) are mixed together. Data must be processed appropriately according to the characteristics of each dataset, and the process of connecting and linking various data sources for integrated analysis can be a very complex task. Furthermore, if this data is not managed properly, analysis accuracy may deteriorate or the reliability of the analysis results may be compromised. The importance of data collection and processing modules is highlighted in the process of integrating and analyzing various data sources. These modules are responsible for functions such as collecting data from diverse sources, transforming it into a format suitable for analysis, and cleaning up unnecessary duplicates or missing data. The traditional method of extracting data from databases was to write queries directly using programming languages such as SQL (Structured Query Language). However, this is a very complex and time-consuming task for non-experts. Although many companies recognize the need for data analysis, a problem has arisen where users without professional SQL knowledge find it difficult to utilize. Furthermore, high-quality datasets are required for data analysis and AI training. It is essential not only to simply collect data but also to process and preprocess it into a structure suitable for AI training. The performance of an AI training model depends heavily on the quality of the training data, and if the data is not properly labeled and structured, the model's performance may degrade. Data modeling tools play a crucial role in transforming collected data into an analyzable and learnable form. The goal of data modeling is to analyze complex data structures, discover meaningful relationships, and represent them visually. Additionally, data visualization tools effectively represent data, helping users understand and analyze it more easily. Existing data analysis and AI training platforms have faced the problem of difficulty in effectively integrating and analyzing various data sources. Building high-quality datasets necessary for AI training required significant time and effort, and the lack of automated data processing and analysis tools was cited as a limitation. These limitations are acting as obstacles for companies in achieving business innovation utilizing AI technology. FIG. 1 is a flowchart illustrating an AI-based data integration analysis system according to one embodiment of the present invention. FIGS. 2 and FIGS. 3 are exemplary diagrams for explaining the first query generation process during AI-based data integration analysis according to the present invention. FIG. 4 is an example diagram illustrating the first query generation process during AI-based data integration analysis according to the present invention. FIGS. 5 to 9 are exemplary diagrams for explaining the data modeling process during AI-based data integration analysis according to the present invention. FIGS. 10 to 13 are exemplary diagrams for explaining the AI learning process during AI-based data integration analysis according to the present invention. The aforementioned objectives, signatures, and advantages are described in detail below with reference to the attached drawings, thereby enabling those skilled in the art to easily implement the technical concept of the present invention. In describing the present invention, detailed descriptions of known technologies r