KR-20260065482-A - A system for comparing and contrasting facts and evidence, configured to identify inconsistencies between a client’s statement data and evidentiary data, and to generate inquiries regarding the identified inconsistencies

KR20260065482AKR 20260065482 AKR20260065482 AKR 20260065482AKR-20260065482-A

Abstract

The present invention relates to a factual relationship and evidence comparison system comprising: a user terminal for collecting statement data and evidence data; and a server device for processing statement data and evidence data received from the user terminal; wherein the server device comprises: a database; a data collection tool configured to store the statement data and evidence data in the database; a natural language processing tool configured to extract factual relationship extraction data from the stored statement data; a data collection tool for extracting structured evidence data from the stored evidence data; a comparison and contrast tool configured to compare the factual relationship data and the structured evidence data item by item to determine whether there is a discrepancy and to generate a comparison and contrast table; and a question generation tool configured to generate a natural language-based question based on the discrepancy items included in the comparison and contrast table.

Inventors

천상현

Assignees

천상현

Dates

Publication Date: 20260508
Application Date: 20250701
Priority Date: 20241101

Claims (13)

In a factual and evidence comparison system configured to identify factual discrepancies between client statement data and evidence data, and to generate queries regarding discrepancies, A user terminal collecting the above statement data and the above evidence data; and A server device for processing statement data and evidence data received from the above user terminal; comprising The above server device is, database; A data collection tool configured to store the above statement data and the above evidence data in the above database; A natural language processing tool configured to extract factual relationship data from the above-mentioned stored statement data; A data collection tool for extracting structured evidence data from the above-mentioned stored evidence data; A comparison and contrast tool configured to determine whether there is a discrepancy by comparing the above factual data and the above structured evidence data item by item and to generate a comparison and contrast table; and A question generation tool configured to generate natural language-based questions based on discrepancies included in the above comparison table; comprising System for comparing facts and evidence.
In Article 1, The above natural language processing tool is, Configured to generate factual extraction data including at least one of the time of the event, the person involved in the event, the amount involved in the event, and an act related to the event from the statement data based on a factual item definition table containing item information defined by event type, System for comparing facts and evidence.
In Article 2, The above natural language processing tool is, Generate an internal query corresponding to each item in the above item definition table, and Configured to extract response values for the internal query from the statement data using a pre-trained language model, System for comparing facts and evidence.
In Article 1, The above comparison and contrast tool is, For each item in the above item definition table, the above factual relationship extraction data and the above structured evidence data are compared by corresponding them one-to-one, and Configured to generate a comparison table containing at least one of information regarding consistency and inconsistency, System for comparing facts and evidence.
In Article 4, The above comparison and contrast tool is, For the discrepancies included in the above comparison table, Calculate a importance score based on at least one of the legal impact of each of the above items, contextual importance, potential conflict between statements and evidence, issue of the case, and legal importance, and Configured to generate a mismatch item importance evaluation table based on the importance scores calculated above, System for comparing facts and evidence.
In a server device configured to generate a query by collecting client statement data and evidence data from a user terminal, database; A data collection tool configured to store the above statement data and the above evidence data in the above database; A natural language processing tool configured to extract factual relationship data from the above-mentioned stored statement data; A data collection tool for extracting structured evidence data from the above-mentioned stored evidence data; A comparison and contrast tool configured to determine whether there is a discrepancy by comparing the above factual data and the above structured evidence data item by item and to generate a comparison and contrast table; and A question generation tool configured to generate natural language-based questions based on discrepancies included in the above comparison table; comprising Server device.
In Article 6, The above natural language processing tool is, Configured to generate factual extraction data including at least one of the time of the event, the person involved in the event, the amount involved in the event, and an act related to the event from the statement data based on a factual item definition table containing item information defined by event type, Server device.
In Article 7, The above natural language processing tool is, Generate an internal query corresponding to each item in the above item definition table, and Configured to extract response values for the internal query from the statement data using a pre-trained language model, Server device.
In Article 6, The above comparison and contrast tool is, For each item in the above item definition table, the above factual relationship extraction data and the above structured evidence data are compared by corresponding them one-to-one, and Configured to generate a comparison table containing at least one of information regarding consistency and inconsistency, Server device.
In Article 9, The above comparison and contrast tool is, For the discrepancies included in the above comparison table, Calculate a importance score based on at least one of the legal impact of each of the above items, contextual importance, potential conflict between statements and evidence, issue of the case, and legal importance, and Configured to generate a mismatch item importance evaluation table based on the importance scores calculated above, Server device.
In Article 10, The above comparison and contrast tool is, Configured to calculate the above importance score based on a predefined rule-based algorithm or a pre-trained language model, Server device.
In Article 10, The above comparison and contrast tool is, Configured to prioritize displaying items among the items included in the above discrepancy item importance evaluation table whose importance score is above a threshold, Server device.
In Article 6, The above question generation tool is, In addition to the discrepancies included in the above comparison table, Configured to generate additional natural language-based questions even for information items not explicitly included in the above statement data, Server device.

Description

A system for comparing and contrasting facts and evidence, configured to identify inconsistencies between a client’s statement data and evidentiary data, and to generate inquiries regarding the identified inconsistencies. The present invention relates to an artificial intelligence-based factual analysis and evidence comparison technology, and more specifically, to a technology for a legal case support method that structures and compares statements entered in natural language with documented evidence, and enables the derivation and refinement of key issues of a case through automatic querying and feedback on discrepancies. In traditional legal practice, lawyers or legal experts have primarily used a method of manually reviewing and comparing natural language-based statements made by clients regarding a case with documented evidence, such as contracts and promissory notes, to determine consistency. This approach has limitations in quickly and accurately identifying the facts when the volume of data is large or statements are not specific, and there is also a possibility of information omission or errors occurring due to the reliance on subjective interpretation. In addition, existing technologies do not sufficiently provide functions to systematically analyze discrepancies between statements and evidence, or to generate semantic-based queries on discrepancies to collect feedback, which can delay the clarification of issues in a case and lead to inefficiency in establishing response strategies. In particular, although some attempts were made to introduce Natural Language Processing (NLP) technology, they often remained at the level of simple keyword extraction or standardized question-and-answer, which limited the implementation of structured analysis and dynamic query generation that reflect the actual context of each event. Figure 1 is a configuration diagram of a factual relationship and evidence comparison and contrast system (1) according to the present invention. FIG. 2a is a drawing for explaining an embodiment of a factual relationship item definition table according to the present invention. FIG. 2b is a drawing showing a client's statement in a loan repayment case according to one embodiment of the present invention. FIG. 3a is a drawing for explaining a promissory note, which is evidence according to one embodiment of the present invention. FIG. 3b is a drawing showing a comparison table according to one embodiment of the present invention. FIG. 3c is a drawing showing an inconsistency item importance evaluation table according to one embodiment of the present invention. FIG. 3d is a drawing showing an updated comparison table according to one embodiment of the present invention. FIG. 4 is a diagram showing a user interface for generating a query on a mismatch item on a comparison table according to the present invention. FIG. 5a is a diagram showing natural language-based statements entered by a client through a user terminal in a contract dispute litigation situation according to one embodiment of the present invention. FIG. 5b is a drawing showing an example of a contract submitted in a contract dispute case according to one embodiment of the present invention. FIG. 5c is an example drawing of applying a comparison table (231) according to one embodiment of the present invention to a type of contract dispute case. FIG. 5d is a drawing illustrating an inconsistency item importance evaluation table (232) based on the comparison contrast table of FIG. 5c according to one embodiment of the present invention. FIG. 5e is a drawing illustrating a query providing user interface (UI) in which the result of operation of a question generation tool (240) according to one embodiment of the present invention is visually displayed through an output unit (150) of a user terminal (100). FIG. 6a shows an example of a user interface screen in which a result providing tool according to the present invention provides a summary list of mismatched items. FIG. 6b shows an example in which a result providing tool according to the present invention visualizes and provides review priority or items requiring response based on color, grade, and priority. Specific details of the embodiments are included in the detailed description and drawings. The advantages and features of the present invention and the methods for achieving them will become clear by referring to the embodiments described below in detail together with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below but may be implemented in various different forms. These embodiments are provided merely to ensure that the disclosure of the present invention is complete and to fully inform those skilled in the art of the scope of the invention, and the present invention is defined only by the scope of the claims. Throughout the specification, the same reference numerals refer to the same components. Figure 1 is a conf