LOGICAL INFERENCE-BASED ARTIFICIAL INTELLIGENCE SYSTEM

Abstract

This invention relates to a system (1) that enables the creation of a logical information model using transaction records, user information, transaction history and network connection data used in the fields of finance, health, logistics and security, and the performance of logical inference and analysis operations on the created model.

Description

1 TARIFF LOGICAL INFERENCE-BASED ARTIFICIAL INTELLIGENCE SYSTEM Technical Area This invention utilizes transaction records in the fields of finance, healthcare, logistics, and security. A logical structure is created using user information, transaction history, and network connection data. creating an information model and making logical inferences from the created model It relates to a system that enables analysis processes to be carried out. Previous Technique Today, artificial intelligence systems analyze large datasets, tasks are distributed among multiple AI agents and given to the user. Topological Data Analysis (TDA-15) for providing automated decision results. Topological Data Analysis), Multi-Agent System (MAS), and Axiomatic rule structures are used in current systems for data relationships. Connections are being analyzed, each agent is performing specific tasks, and Automated process outputs are generated. However, mathematical analysis is performed. Using these methods to examine the connections between datasets, 20 The decisions made by artificial intelligence agents are controlled through common rules. and the result of a process produced by one agent should not be transmitted by other agents The inability to prevent erroneous transaction outputs due to the inability to verify the transaction. the inability to reliably track the steps and by multiple agents Deficiencies in the form of contradictory decisions in complex tasks performed 25 It is emerging. Therefore, considering the studies and shortcomings in the current technique... when available, user actions, transaction histories, user behavior 30 different types of data in the form of records, network connections and corporate data records Access rules, transaction rules and security for information obtained from sources. 2 evaluation according to criteria, analysis of connections between data, inconsistent records, unexpected transaction behavior, and unusual data patterns a system that enables identification and the performance of logical inference operations It is understood that it is needed. In the Chinese patent document numbered CN118132762, which is included in the known state of the art... a multi-agent intelligent search engine and human-machine interaction system and The method is discussed. The diagram of the application is a human-machine interaction. It is applied to the system, creating an intelligent receiving engine based on multiple factors, The key information of the query data is controlled by a central control component. 10 Multiple factors are considered in the analysis, and the necessary candidate for the response is derived through collaboration. Information is gathered and generated based on key data; intelligent agent, artificial intelligence model It performs deep semantic understanding on the query information through this means and query Key information can be extracted accurately and comprehensively; intelligent factor, 15 to generate the answer based on key information using an artificial intelligence model. It retrieves and recalls the necessary candidate information, thus providing the required candidate to generate the response. information can be retrieved more comprehensively and accurately, candidate information can be retrieved. The likelihood of callback omissions, recall errors, or failures is reduced, candidate The accuracy of data retrieval is improved and the user experience is enhanced. This improves the response information generated based on candidate data. The quality can be improved, and a more satisfactory answer can be obtained from the user's perspective. Brief Description of the Invention The purpose of this invention is to provide data analysis for finance, healthcare, logistics, security, public sector and scientific fields. 25 User actions, transaction history, and network connections are listed in these areas. using artificial intelligence agents to make inferences between pieces of information the analysis of logical and structural relationships using mathematical methods, Identifying inconsistencies, abnormal patterns, and structural connections and analyzing the results The goal is to create a system that enables this to be presented to the user. 30 3 Detailed Description of the Invention The "Logical Inference Based" approach was implemented to achieve the purpose of this invention. The "Artificial Intelligence System" is shown in the attached figure; Figure 1. Schematic view of the system described in the invention. The parts shown in the figure are individually numbered, and the corresponding numbers correspond to these numbers. given below: 1. System 2. Application 3. Database 4. Server Transaction logs used in finance, healthcare, logistics, and security, user logical information is formed using data such as transaction history and network connection data. creating a model and performing logical inference and analysis on the created model. The system described in the invention was developed to enable the performance of these operations. (1); 20 - the user's smartphone, tablet computer, desktop computer or portable an electronic device in the form of a computer that automatically processes data. running applications and / or software that function and produce meaningful results, users' transaction information, user account information, transaction history, user movements, access logs, connection information and queries to be analyzed 25 entering the target system and analysis results, correlation results, risk assessments, inferences, and anomaly reports at least one application configured to enable display (2), - financial transaction rules obtained through application (2), health assessment rules, logistics route restrictions, security access requirements, transaction records, 30 user information, transaction history, network connection data, analytics queries, 4 records of the generated logical information model, inference results, inconsistencies to store records, anomaly detections, and explanatory outputs. at least one database structured (3) and - Financial transaction rule, health assessment rule obtained via application (2), The rules regarding logistics route restrictions and security access requirements, along with transaction logging, are 5. processing data such as user information, transaction history, and network connection data, etc. The subject is logical knowledge, established by creating logical connections between rules and data. creating a model, making logical inferences from the created information model, The relationship between transaction time and transaction amount, user information and transaction history. The relationship between them, analyzing the relationship between access conditions and network connection data, 10 conflicting transaction records, user who does not comply with access terms their movements, unusual transaction sequences, unexpected network connections, and data The most structured system designed to identify abnormal connection patterns in its structure It contains a small server (4). The application (2) in the system (1) which is the subject of the invention, is defined by the user as an axiom, Entering data into the target system in the form of rules, facts, queries, and job descriptions. the input field provided and the logical inference generated by the server (4) results, analysis results, explanatory information, inference chains, category The outputs in the form of diagrams, persistence diagrams and topological network maps are 20 The result is displayed to the user as text, table, or image output. It is configured to provide at least one interface that includes the display area. The database (3) in the system (1) which is the subject of the invention, target through the application (2) axiom, rule, fact, query and job description data transmitted to the system; financial risk 25 modeling, regulatory compliance, complex system simulation, scientific exploration Transaction logs and user information used in process and logistics planning operations, to keep records of transaction histories and network connection data is structured. The database (3) is logically created by the server (4). information model records, logical inference results, mathematical analysis 30 results, inference chains, evidence trees, anomaly detection records, category diagrams, persistence diagrams, topological network maps, and the explanatory outputs for later viewing or reuse. It is configured for storage. The server (4) in the system (1) which is the subject of the invention, any remote communication 5 to communicate with the application (2) and the database (3) using the protocol and In order to exchange data with the application (3) through this established communication is configured. The server (4) is configured by the user via the application (2). Receiving data entered in the form of axioms, rules, facts, queries, and job descriptions, Analyzing input statements using controlled natural language, data relationships, processing 10 logical terms in the form of conditions, access definitions, process rules, or decision criteria Ensuring that definitions are entered into the target system, first-order logic, Modal logic, definitional logic, OWL (Web Ontology Language) Ontological Language that Defines Conceptual Relationships), RDF (Resource Description) Framework - 15 that displays data by relating it in a node and link structure. Data Model), JSON (JavaScript Object Notation - Structured Data Representation) Format) and XML (Extensible Markup Language) obtaining structured data in a specific format, the relationship between that data, Connection, order of operations, conditional and rule structures can be analyzed and correlated. and to transform it into a data structure on which logical inferences can be performed 20 is configured. The server (4) receives data from the user or external data sources. the axioms, rules, facts, transaction records, user information, transactions received data including their history, network connection data, security logs, and query data. past transaction records, past user records stored on the base (3), Evaluating with past inference results and past analysis records, the same 25 the user shares records belonging to the same process and / or the same data structure in a common link. grouping data under structures such as context information, processing space, data type, and operation. Organizing time according to data source and connection structure, for the same topic or the same axioms and rules of the processing field under micro-theory structures axiom 30 using grouping, versioning and historical tracking methods When were the rule, inference, and analysis records created, and which data were used? 6 to record how it was produced and by what process it was modified It is configured accordingly. The server (4) is registered on the database (3). whether axioms, rules, and facts are logically consistent with each other to check if there are any conflicting rules for the same data. Determining that it is not present, table methods within the scope of consistency management, 5 Resolution method and SMT solvers (Satisfiability Modulo Theories Solver - Checking whether both mathematical and logical constraints are satisfied simultaneously. Using the Eden Analysis System, we identify and resolve conflicting records. It is configured to mark and analyze. The server (4) is created Resolution method, natural derivation, induction on the logical information model 10 learning algorithms based on and abductive reasoning method (Current Data and An inference method that allows for the identification of the most likely explanation based on relationships. making logical inferences using user-entered queries axioms, rules, facts, transaction logs, user information, transaction histories, and Generating responses based on relationships between network connection data, rule 15. Updates, terms of operation, and potential consequences of access changes. to conduct scenario simulations for evaluation, performed which axioms, rules, data connections, and transaction relationships are used in inference operations inference chains and evidence trees to show that it was produced using It is configured to create the server (4), axioms, data schemas, 20 Ontologies, processes, and the relationships between inference outcomes are discussed in category theory. modeling using, object, morphism, functor used in category theory Financial using (Building Preservation Mapping) and natural transformation methods. data, security logs, transaction histories, network connection data, and process logs. matching data relationships between them, information from different data fields 25 to combine, analyze compositional relationships and relationships between data structures It is structured to evaluate consistency. The server (4) uses axioms. related datasets, transaction histories, network connection data, user their movements and inference results using topological data analysis / homology to analyze, transform datasets into point cloud or graphical data structures, 30 Vietoris-Rips (Connection Based on Proximity Relationships Between Data Points) 7 (Filtering Method that Creates Structures) and Čech (Common Intersection of Data Points) (Filtering method that creates link and cluster structures according to their fields) Examining connections within data using filtering methods, persistent homology Using this method to identify loops, gaps, and anomalous data patterns, and to detect persistence. Generating diagrams, barcode structures, and topological summaries, and extracting features from data. 5 the strongest connections within, the longest-lasting cycles, and the unusual Identifying connection clusters and performing noise-tolerant anomaly detection. It is structured as follows: The server (4) performs inference operations. their agents, mathematical analysts who perform mathematical analysis operations its agents, planning agents who perform task sequencing and process planning, 10 data collection agents who perform data collection and data processing operations, financial risk analysis, regulatory compliance check, scientific data analysis, security evaluation, complex system simulation or logistics planning processes a field expert who conducts assessments related to specific transaction areas. managing agents, ensuring each agent adheres to the axioms and rules specific to their area of ​​responsibility, 15 transaction logs, user information, transaction histories, and analysis results to enable access to BDI (Belief-Desire-Intention) Decision-making models, rule-based decision structures, and learning-based methods. using FIPA ACL (Foundation for Determining Agents' Processing Behavior) Intelligent Physical Agents Agent Communication Language - Artificial Intelligence Agents 20 Messages between agents via communication protocols based on Inter-Agent Communication Language (IAL) axiom, inference, result of mathematical analysis, task information, and plan. Sharing, task allocation, result aggregation, contract networking, voting and agents working together using market mechanism methods to provide and the actions that agents will perform axioms, rules, inferences 25 to guide according to the results and mathematical analysis results It is structured. The server (4) performs logical operations carried out by the agents. inference processes, category theory analyses, topological data analyses, scenario planning Intermediate data created as a result of simulations and field expert assessments to gather the results, compare evaluations from different agents, the same 30 data, same transaction log, same user activity, same network connection, or same process 8 contradictions, inconsistencies, or differences among the results obtained. to determine whether an evaluation exists, axioms, rules, procedure logically based on records, past analysis results, and data relationships. to create the most consistent result, financial risk based on the results created. assessment, regulatory compliance check, scientific hypothesis testing, safety 5 ultimate control for complex system simulation and logistics planning processes. Generating analysis outputs, identifying risk situations, and abnormal data patterns, structural inconsistencies, transactional relationships, scenario outcomes, and inferences their evaluations can be reused later on the database (3), 10 to be recorded for questioning or viewing is configured. The server (4) displays the generated logical inference results. mathematical analysis results, explanatory information, inference chains, evidence trees, category diagrams, persistence diagrams, topological networks maps, anomaly notifications, structural consistency assessments, and scenarios Simulation outputs are presented in a user-understandable format such as text, tables, graphs, or 15 converting to visual output format, determining which axioms the generated results are based on using rules, data relationships, inference processes, and mathematical analyses Generating outputs that explain what was produced, graphs, data links, inter-agent visually present communication, category diagrams, and topological analysis results. It is configured to prepare and deliver to the user via the application (2). 20 Industrial Application of the Invention Thanks to the system (1) which is the subject of the invention, finance, logistics, health, security, public, energy and User operations in fields such as scientific data analysis, electronic device 25 AI agents using records, transaction histories, and network connections inferences are made through mathematical relationships in data structures. analysis using various methods, abnormal conditions and structural inconsistencies This ensures the identification and presentation of the results obtained. 9 Around these fundamental concepts, the invention is called "Logical Inference-Based Artificial Intelligence". It is possible to develop a wide variety of applications related to the System (1)”, and the invention This cannot be limited to the examples described here, but is primarily stated in the claims. It is like that.

Claims

REQUESTS 1. Transaction records used in finance, healthcare, logistics, and security, using user information, transaction history, and network connection data the creation of a logical information model and 5 tasks on the created model It enables the performance of analytical processes through logical inference; - the user's smartphone, tablet computer, desktop computer or run on an electronic device in the form of a portable computer, an application that automatically processes data and produces meaningful results and / or execute software, user transaction information, user 10 account information, transaction history, user activity, access logs, Connection information and queries to be analyzed are sent to the target system. input and analysis results, correlation results, risk assessments, inferences, and anomaly reports At least one application configured to enable display (2), 15 - financial transaction rules obtained through application (2), health evaluation rules, logistics route restrictions, security access requirements, transaction logs, user information, transaction history, network connection data, analysis queries, generated logical information model records, inference results, discrepancy records, anomaly detections, and explanation 20 containing at least one database (3) configured to store its outputs and - Financial transaction rule obtained via application (2), health assessment The process is governed by rules such as logistics route restrictions and security access requirements. records in the form of user information, transaction history, and network connection data. processing data, establishing logical connections between those rules and the data. 25 to create a logical information model by establishing connections, the information created making logical inferences on the model, transaction time and transaction cost the relationship between user information and transaction history, access Analyzing the relationship between the condition and network connection data, conflicting factors transaction logs, user activities that do not comply with access terms, the usual 30 off-process sequences, unexpected network connections, and data structure issues. 11 at least one structured to identify abnormal linkage patterns a payment system characterized by server (4) (1).

2. User-defined axioms, rules, facts, queries, and job descriptions. input field that enables data to be entered into the target system and server (4) 5 The logical inferences and analysis results generated by descriptive information, inference chains, category diagrams, persistence the outputs in the form of diagrams and topological network maps are given to the user Results display that allows the output to be shown as text, table, or visual output. The application is configured to provide at least one interface containing the area (2) and 10 A system like the one in Claim 1, characterized (1).

3. Axioms, rules, facts, queries transmitted to the target system via application (2) and job description data; financial risk modeling, regulatory compliance, Complex system simulation, scientific discovery process, and logistics planning 15 Transaction logs, user information, and transaction histories used in operations and network connection data, and data structured for record keeping. any of the above claims characterized by base (3) such a system (1).

4. Logical information model records generated by the server (4), logical inference results, mathematical analysis results, inference chains, evidence trees, anomaly detection records, categories diagrams, persistence diagrams, topological network maps, and The output descriptions can be viewed or reused later. 25 characterized by the database structured to store (3) a system like any of the above requests (1).

5. Using any remote communication protocol, the application (2) and data to communicate with the base (3) and through this communication the application (3) 30 server (4) configured to exchange data with 12 a system like any of the above characterized claims (1).

6. Axioms, rules, facts entered by the user via the application (2), Obtaining data in the form of queries and job descriptions, controlled natural language 5 Analyzing input statements using data relationships and transaction conditions. Logical elements in the form of access definitions, process rules, or decision criteria Ensuring that the definitions are entered into the target system is the primary objective. Logic, modal logic, definition logic, OWL, RDF, JSON, and XML. retrieving structured data in the format, 10 among the data in question Relationship, link, sequence of operations, conditional and rule structures can be analyzed. related data on which logical inferences can be performed. characterized by the server (4) configured to convert to its structure a system like any of the above requests (1).

7. Axioms and rules received from the user or external data sources, facts, transaction logs, user information, transaction history, network connection data, security logs and query data database (3) past transaction records, past user records, history stored on it Evaluating the inference results and past analysis records, the same 20 to the user, records belonging to the same process and / or the same data structure grouping data under link structures, context information, processing space, data to organize according to type, processing time, data source and connection structure, the same axioms and rules related to the subject or the same field of study are micro-theory. Grouping under structures, versioning and historical tracking methods 25 on what date the axiom, rule, inference, and analysis records were created using how it was created, what data was used to generate it, and what process Server configured to record that it has been changed as a result (4) as in any of the above claims characterized by system (1). 30 13 8. The axioms, rules and facts recorded on the database (3) to check if they are logically compatible with each other, the same To determine whether there are conflicting rules for the data. Within the scope of consistency management, table methods, resolution method and Using SMT solvers, we identify the conflicting records and resolve the conflicts. 5 Characterized by the server (4) configured to mark and parse. a system like any of the above-mentioned requests (1).

9. The resolution method is applied to the generated logical information model, naturally. Derivation, inductive learning algorithms, and abductive reasoning 10 making logical inferences using the method, by the user Axioms, rules, facts, transaction logs, user inputs are used in the queries entered. based on the relationships between information, transaction histories, and network connection data. generating responses, rule updates, terms of service, access Scenario 15 to evaluate the potential consequences of the changes. performing simulations, which inference processes are performed using axioms, rules, data connections, and transaction relationships to create inference chains and evidence trees to show that it was produced the above requests are characterized by the server (4) configured to be so. a system like any other (1). 20 10. Axioms, data schemas, ontologies, processing procedures, and inference results. modeling the relationships between categories using category theory, category The objects, morphisms, functors, and natural transformations used in his theory using methods such as financial data, security records, transaction 25 data relationships between histories, network connection data, and process logs to match, to combine information from different data fields, to analyze compositional relationships and relationships between data structures characterized by the server (4) configured to evaluate consistency a system like any of the above requests (1). 30 14 11. Datasets, transaction histories, and network connections related to the axioms. topological data including user activity and inference results Analyzing datasets using point cloud or homology analysis Converting to a graphical data structure, Vietoris-Rips and Čech filtering. Examining the connections within the data using methods, persistent homology 5 using it to identify loops, gaps, and anomalous data patterns, Generating persistence diagrams, barcode structures, and topological summaries, features By extracting the strongest, longest-lasting connections within the data, Identifies loops and unusual connection sets and is noise-resistant. Server (4) configured to perform anomaly detection and 10 a system like any of the above characterized claims (1).

12. Inference agents that perform inference operations, mathematical analysis. mathematical analysis agents that perform operations, task sequencing and 15 planning agents who carry out process planning, data collection and data data collection agents who carry out regulatory processes, financial risk analysis, regulatory compliance check, scientific data analysis, security evaluation, complex system simulation, or logistics planning Assessments of specific transaction areas in the form of transactions 20 managing field specialist agents, each agent's own task axioms, rules, transaction logs, user information related to the field, To enable access to transaction histories and analysis results, BDI, rule using decision-making structures based on learning-based methods, agents Determining transaction behaviors, FIPA ACL-based communication protocols 25 through which agents exchange messages, axioms, inferences, and mathematical concepts. Analysis results, task information and plan sharing, task distribution, result aggregation, contract network, voting and market mechanism to enable agents to work together using methods and the agents The operations it will perform are based on axioms, rules, inferences, and 30 structured to guide according to the results of mathematical analysis any of the above requests characterized by the server (4) such a system (1).

13. Logical inference processes performed by agents, category theory analyses, topological data analyses, scenario simulations, and field 5 Gathering interim results based on expert assessments, Comparing assessments from different agents, the same data, the same transaction log, same user activity, same network connection, or same process contradictions, inconsistencies, or differences among the results obtained. to determine whether an evaluation exists, axioms, rules, 10 based on transaction records, past analysis results, and data relationships generating the most logically consistent result, the results generated financial risk assessment, regulatory compliance check, scientific hypothesis testing, safety checks, complex system simulation, and To produce final analysis outputs for logistics planning processes, identified risk 15 their status, abnormal data patterns, structural inconsistencies, processing their relationships, scenario outcomes, and inference evaluations in the database (3) to be used, questioned or reused later server (4) configured to record for viewing A system like any of the above-mentioned described requirements 20 (1).

14. The logical inference results and mathematical analysis results that have been generated, descriptive information, inference chains, evidence trees, categories diagrams, persistence diagrams, topological network maps, anomaly 25 notifications, structural consistency assessments, and scenario simulations outputs in text, tables, graphs, or formats that are understandable to the user. converting to visual output format, which of the generated results axioms, rules, data relationships, inference processes, and mathematics Generating outputs that explain how the data was produced using analyses, graphs, 30 data links, inter-agent communication, category diagrams, and topological structures. 16 Prepare the analysis results visually and through the application (2) characterized by the server (4) configured to transmit to the user a system like any of the above requests (1). 10 20 30

Images