KR-20260066669-A - System and Method for Automatically Converting Document Formatting Using an Artificial Intelligence Agent

Abstract

The present invention relates to a system and method for automatic document format conversion using an artificial intelligence agent. The method of the present invention comprises the steps of: receiving a first document to be converted into a format and a sample document serving as a standard for the format; an artificial intelligence agent extracting and analyzing respective format components from both documents; an artificial intelligence agent mapping the format features of the sample document to the corresponding components of the first document to generate format conversion data; and outputting a final document based on the format conversion data in which only the format is replaced according to the sample document while preserving the content of the first document. According to the present invention, non-destructive format conversion is realized without altering the text content of the first document at all, thereby fundamentally blocking information distortion caused by AI illusions, while also having the effect of intelligently extracting and automatically applying a format from a sample document of a heterogeneous format.

Inventors

• 안범주

Assignees

• 안범주

Dates

Publication Date

20260512

Application Date

20260330

Claims (1)

1. In a method for automatically converting document formats using an artificial intelligence agent, A step of receiving a first document to be converted into a format and a sample document serving as the standard for the format; A step in which an artificial intelligence agent extracts and analyzes each format component from the first document and the sample document; The step of generating format conversion data by mapping the format features of the analyzed sample document to the corresponding components of the first document by the artificial intelligence agent; and A step of outputting a final document in which the format of the first document is modified to be identical or similar to the format of the sample document based on the format conversion data. Method for converting document formats using an AI agent.

Description

System and Method for Automatically Converting Document Formatting Using an Artificial Intelligence Agent System and Method for Automatically Converting Document Formatting Using an Artificial Intelligence Agent The present invention relates to a document processing technology using an artificial intelligence (AI) agent, and more specifically, to a system and method that simultaneously receives a first document to be converted into a format and a sample document serving as a standard for the format, wherein the AI agent extracts and analyzes the format components of the two documents respectively and automatically maps the format characteristics of the sample document to the corresponding components of the first document, thereby converting only the format (Layout & Style) to be identical or similar to the sample document while preserving 100% of the content of the first document, and outputting a final document. In the modern corporate environment, various types of documents, such as contracts, reports, proposals, technical documents, and academic papers, are generated and distributed in large quantities. These documents frequently require meeting different formatting standards depending on the requirements of each organization or customer. For example, reports submitted to public institutions must adhere to specific fonts, margins, paragraph spacing, and heading hierarchy according to the institution's standard document format, while global companies require standardization of formatting between English and Korean documents. Previously, this formatting conversion was performed manually by personnel or processed using Desktop Publishing (DTP) software such as Microsoft Word or Adobe InDesign; however, this was an extremely inefficient process that took hours to days, especially as the volume of the documents increased. To address these issues, various document conversion automation technologies have been proposed. An example of conventional technology is a method that automatically inserts document content into a predefined fixed template. However, this method is applicable only when the structure of the template perfectly matches that of the original document; there were issues where automatic conversion failed or the output was distorted if components of the original document were absent from the template, or conversely, if components were present in the template but not in the original. Another known conventional technique involves rewriting documents using Generative AI, particularly Large Language Models (LLMs). Services such as IronCrow AI and eety.ai generate new drafts with a similar writing style when an example document is uploaded. However, since this method essentially involves the AI newly 'generating' the document's 'content,' there is a persistent risk that information requiring accuracy—such as original figures, proper nouns, and legal terms—will be distorted or altered due to the hallucination phenomenon inherent to AI. Furthermore, because this method alters the original text content itself, it has a fundamental limitation in that it cannot be applied to legally binding contracts or official documents. In addition, conventional document format conversion technologies were mostly limited to specific file formats such as word processor formats (docx, hwp) or PDF, and could not extract format information when the sample document was a scanned image or a document of a different format. Furthermore, formatting of non-text elements within the document, such as tables, charts, formulas, and images, was insufficient, resulting in low utility in actual work environments. Therefore, there is an urgent need for new technology that can intelligently extract formatting from sample documents in heterogeneous formats and automatically apply it to the original document without altering the content of the original document at all. FIG. 1 : An overall configuration diagram of a document format automatic conversion system using an artificial intelligence agent according to one embodiment of the present invention, showing the interconnection relationships between a user terminal (10), a cloud processing server (600), a data input unit (100), an AI format analysis engine (200), a semantic mapping agent (300), a format conversion rendering engine (400), and a user interface (500). FIG. 2: An internal configuration diagram of an AI format analysis engine (200) according to one embodiment of the present invention, showing the detailed configuration and data flow of a layout analyzer (210), an attribute extractor (220), and a structure analysis model (230). FIG. 3: A diagram illustrating an example of a Semantic Format Mapping mechanism in which a Semantic Mapping Agent (300) according to one embodiment of the present invention determines a component correspondence relationship between a first document (A) and a sample document (B), showing the operation process of a mapping table generator (330) and a semantic