34 results about "Automated reasoning" patented technology

The invention discloses an electric power big data automatic reasoning platform and an electric power distribution system, and relates to the technical field of electric power systems, and the platform comprises a knowledge acquisition unit which collects and preprocesses electric power internal and external data, and converts the data into structured data through a knowledge processing unit; organizing the triple data by the dynamic knowledge graph construction unit to form an initial graph; the updating unit monitors a new dynamic state in real time and corrects or adds atlas content; the reasoning unit is based on the dynamic graph, power fault accurate reasoning and future demand trend prediction are achieved through a fault diagnosis and demand prediction algorithm, efficient integration and dynamic evolution of power knowledge are achieved through construction and real-time updating of the dynamic knowledge graph, timeliness and accuracy of the graph are ensured, and reliable data are provided for fault diagnosis; through a closed-loop architecture of data acquisition-inference analysis-scheme making-execution control, an intelligent sensor and an optimization algorithm are combined to generate a distribution scheme adaptive to a scene, and the power distribution adaptability and economy are improved.

The invention relates to the technical field of bidding document auditing, and particularly discloses a bidding document compliance auditing method based on a multi-modal knowledge graph, which comprises the following steps: extracting a standard reference file from an authoritative data source, and constructing the multi-modal knowledge graph in combination with natural language processing and image recognition technologies; through modal separation and structured processing, texts, tables and images in the bidding document file are converted into structural features, and standardized expression of an unstructured document is achieved; through semantic mapping and node matching, a one-to-one correspondence relationship is established between text modal features of a bidding document file and standard term nodes in a knowledge graph, and text difference features are extracted, so that the accuracy and pertinence of an auditing result are improved; based on an automatic reasoning mechanism of a knowledge graph logic rule, automatic compliance judgment is completed, and subjectivity and omission risks of manual auditing are reduced; the bidding document file is subjected to hierarchical labeling through cross-modal consistency verification, a compliance evaluation result is formed, and the intelligent level of auditing is improved.

The invention discloses a forest management data processing method driven by a large language model, and belongs to the technical field of forest resource management and artificial intelligence data processing crossing. The method comprises the steps of multi-source data acquisition and preprocessing, forest management knowledge graph construction, large language model fine adjustment and retrieval enhancement generation, data semantic fusion and understanding and management strategy reasoning, wherein the large language model performs automatic reasoning based on fused data, management intention and industry knowledge to form a management strategy conforming to a forest growth law and multi-target balance; operation plan text generation: on the basis of completing strategy and space matching, compiling a forest operation plan text meeting forestry industry specifications and management requirements, and man-machine interaction and continuous optimization: collecting feedback information of forestry workers on forest operation plans through a man-machine interaction interface, and systematic evaluation is carried out on the generated operation scheme from forest resource sustainability, ecological function improvement, operation goal achievement degree, risk controllability and policy compliance.

The invention discloses an agent training and evaluation method and system under typhoon disasters. Constructing a typhoon disaster dispatching knowledge graph according to the power grid typhoon emergency plan and the historical dispatching log; and a large language model is selected and is optimized through double-stage training of supervised learning and reinforcement learning. Generating a disposal flow reasoning plan by using the optimized model, and inserting an original sample to form a complete reasoning sample; analyzing the disaster description, extracting an inference target, generating a multi-hop inference chain according to the disposal priority and the logic constraint, and mapping the multi-hop inference chain to a knowledge graph; and recognizing subject entities in the reasoning process, and outputting answer entities in combination with parallel and sequential sampling and semantic similarity expansion reasoning paths. And finally, a question and answer simulation platform and a question library are built, the reasoning agent is questioned, and the reasoning ability of the reasoning agent is evaluated by adopting multi-angle similarity. According to the method, the automatic reasoning and decision-making level of power grid emergency disposal in a typhoon scene can be improved.

The invention relates to the technical field of large language model optimization, and discloses a thinking chain generation method, device and equipment and a storage medium, and the method comprises the steps: constructing a parameter difference matrix and extracting a parameter change direction vector through comparing the parameter difference between a basic model for generating a long thinking chain and a target model for generating a short thinking chain; establishing a controllable continuous reasoning space; a low-rank matrix aligned with a direction vector is generated by utilizing a low-rank fine tuning technology, and the updating weight of the low-rank matrix is controlled by adjusting a normalization factor, so that the basic model can dynamically generate a thinking chain with adaptive length under different task complexity, and the optimal matching of the reasoning depth and the task requirement is realized. According to the method, the problems of redundant calculation and inference insufficiency caused by fixed threshold truncation and parameter stiffness in the prior art are solved, and the efficiency and correctness balance capability of the model in scenes such as intelligent question answering and automatic inference is remarkably improved.

The invention relates to the technical field of block chains, and provides an EVM instruction set abstract semantic description method and device based on a theorem prover, electronic equipment and a computer program product. The method comprises the steps of obtaining a to-be-processed EVM instruction set; and converting the EVM instruction set into a corresponding abstract semantic description based on an inductive definition method in the theorem prover. According to the method, a theorem prover with high credibility is utilized, and instruction behaviors during operation of the EVM instruction set can be accurately and completely described mathematically, so that the accuracy of abstract semantic description for the EVM instruction set is improved, and a solid mathematical basis is provided for automatic reasoning and formalized verification of the smart contract.

The invention belongs to the technical field of ecological data processing. The invention provides an ecological environment admission intelligent analysis risk assessment method and system based on a large model. An integrated scheme of multi-source data acquisition and targeted preprocessing, large model driven knowledge extraction, dynamic ecological environment access knowledge graph construction, project information intelligent analysis, multi-dimensional intelligent matching, risk inference engine iterative reasoning, structured report generation and knowledge and rule base continuous optimization is adopted. A full-process automatic ecological environment admission analysis and risk assessment system is constructed, the core problems that traditional ecological environment admission assessment highly depends on manpower, is low in efficiency and is non-uniform in standard are solved, and the technical defects that an existing system lacks deep text understanding and automatic reasoning ability, knowledge updating lags behind and complex correlation analysis is difficult are overcome. Through close connection and data circulation of all links, the automation degree and operation efficiency of the assessment process are greatly improved, the consistency and objectivity of assessment standards of similar projects are ensured, and the comprehensiveness and accuracy of risk identification are improved.

The invention discloses a traffic law enforcement virtual simulation training method based on a dynamic scene and cognitive evaluation, and the method comprises the steps: generating an initialization parameter according to a trainee portrait and a training target, and driving an emergence type scene generation engine to synthesize a dynamic three-dimensional virtual training scene containing composite illegal elements; in scene operation, multi-modal interaction behaviors and physiological response data of the trainees are synchronously collected to form a data stream; through real-time analysis of the data flow, a cognitive behavior analysis report for evaluating attention, cognitive load and decision intention is generated; constructing a space-time knowledge graph based on the student operation sequence, performing automatic reasoning and contradiction detection in combination with formalized legal rules, and outputting an evidence chain logic verification result; and fusing the report and the result to carry out multi-dimensional weighted evaluation, and generating a comprehensive score and personalized feedback. According to the method, normal form transformation from preset process training to open cognitive evaluation is realized, and the authenticity, pertinence and evaluation depth of training are improved.

Methods, systems, and computer program products for generating computer source code products. According to one method, an engineer creating or modifying an implementation of computer source code selects design patterns from a catalog of pre-existing design patterns. The design patterns are known abstractions that solve specific problems of computer program design and implementation. The selected design patterns are then bound to elements within a representation of the implementation known as a model. The model is then converted into a representation using elemental design patterns, and this abstraction model is used to generate the computer source code in any programming language. The abstraction model is also suitable as input for an automated inference engine for validation or verification purposes prior to or after the generation of the computer source code.

To provide a device, a dataset and a computer-implemented method for testing automated reasoning, in particular, automated tabular data reasoning, for determining an answer to a question based on a table including contents of cells in a tabular structure.SOLUTION: The method includes: providing a table, a question, and an answer (300); determining a prediction for an answer by automated reasoning based on contents of cells of the table and the question (304-2, 306-2, 308-2); and determining, depending on a difference between the prediction and the answer, at least two of a robustness metric of automated reasoning, in particular, automated tabular data reasoning, with respect to changes in the tabular structure, an attention metric of automated reasoning, in particular, automated tabular data reasoning, with respect to at least one cell of the table, and a robustness metric with respect to changes in numerical contents of at least one cell of the table (304-3, 306-3, 308-3).SELECTED DRAWING: Figure 3

The invention provides an advanced instruction understanding and long-range navigation method based on a visual language model, and the method comprises the steps: carrying out the semantic object reasoning and spatial position prediction based on an advanced instruction and a global three-dimensional scene with annotations through a visual language model, and obtaining a target region; determining a target waypoint in the local boundary of the target area through a random sampling mode; and based on the panoramic view of the target waypoint, through the intelligent model with the body, carrying out detection and positioning to obtain the position information of the target object. Compared with the large difference between a navigation task and a real world demand in the prior art, the method comprises the following steps: performing semantic object reasoning and spatial position prediction through a visual language model, and determining the position of a target object; based on the panoramic view of the target waypoint, through the intelligent model with the body, detection and positioning are carried out, and a target object is determined; understanding of a high-level abstract instruction is realized, and a target object and a possible position are automatically reasoned; and the navigation capability of the robot in a complex real environment is obviously improved.

A large language model driven forest management data processing method belongs to the cross technical field of forest resource management and artificial intelligence data processing. Multi-source data collection and preprocessing, forest management knowledge graph construction, large language model fine tuning and retrieval enhancement generation, data semantic fusion and understanding, management strategy reasoning: based on the fusion data, management intention and industry knowledge, the large language model automatically reasons to form the management strategy conforming to the forest growth law and the multi-objective balance, management scheme text generation: on the basis of completing the strategy and space matching, the forest management scheme text conforming to the forestry industry specification and management requirement is compiled, human-computer interaction and continuous optimization: through the human-computer interaction interface, the feedback information of the forest management scheme of the forestry workers is collected, and the generated management scheme is systematically evaluated from the forest resource sustainability, ecological function improvement, management target achievement degree, risk controllability and policy compliance.

Techniques are described for providing a SAT-based solver for a quantifier-free theory of strings and bit vectors. The solver can be used by an automated reasoning service of a cloud provider network to analyze policies and policy consequences. The solver reduces an input formula to a Boolean satisfiability problem by encoding the input formula into an equivalently satisfiable propositional formula, where the satisfiability of the equivalently satisfiable propositional formula is determined by a SAT solver. Instead of using a traditional DPLL(T)-style algorithm, the solver described here limits the length of variables in an input formula and reduces the problem to a single formula, which can then be solved using incremental SAT solving.The solver can be used independently or as part of a portfolio of solvers to determine the satisfiability or unsatisfiability of certain formulas corresponding, for example, to questions about user policies within a cloud provider network.

The application relates to the technical field of ship process design, in particular to a ship complex structure three-dimensional process design method, a computer storage medium and equipment. The application extracts key process features and processes them by interpreting the three-dimensional model of the ship complex structure to be subjected to process design, then obtains process design knowledge rules of the model process features based on complex structure process knowledge system reasoning, then performs process intelligent design based on the process design indication rules, realizes process automatic reasoning and design, improves the knowledge reuse rate, reduces repetitive labor, shortens the manufacturing cycle of products, and solves the technical problem of low process design efficiency of the ship complex structure.

The invention discloses a task scheduling method and system for SMT solving, and relates to the technical field of automatic reasoning, and the method comprises the steps: carrying out the feature extraction of an SMT problem instance training set, and obtaining the feature vector of each SMT problem instance; clustering a feature vector set formed by all feature vectors by adopting an unsupervised clustering algorithm to generate a plurality of instance families, and storing a clustering center of each instance family; aiming at each instance family, constructing a hyper-parameter space of a solver sequence and time distribution; for each instance family, searching in a hyper-parameter space through a Bayesian optimization algorithm of a hybrid agent model to obtain an optimal hyper-parameter corresponding to the instance family, the optimal hyper-parameter being used for generating a solver scheduling scheme of the instance family; and for an input SMT problem instance, extracting features and matching the features to the nearest clustering center, and calling a solver scheduling scheme corresponding to the optimal hyper-parameter of the instance family to execute SMT solving, so that the overall performance and the resource utilization rate of an SMT solving system are improved.

This invention relates to a multimodal intelligent review method for communication documents, specifically in the field of artificial intelligence. By performing deep fusion analysis of visual and linguistic elements on images of communication bidding documents, a document knowledge network rich in semantic associations is constructed. It then simulates expert thinking for automated reasoning and verification, accurately understanding complex content such as mixed text and images, and stamped annotations. This transforms unstructured information into a computable knowledge graph and supports rule-based and vector-based intelligent retrieval, significantly improving the objectivity, accuracy, and efficiency of the review process. It achieves automatic extraction of review elements, verification of logical consistency, and deep semantic querying, providing reliable and interpretable intelligent decision support for communication project review.

The invention provides an automatic reasoning method and system for use specifications of a reusable component of a smart contract, and belongs to the technical field of security analysis of the reusable component of the smart contract. An automatic crawling framework is used for collecting source codes of reusable components of mainstream smart contracts to be analyzed and document information of the reusable components of the mainstream smart contracts to be analyzed. And the crawled information of the reusable component of the smart contract is processed by using the task planning intermediate, and the planning of the reusable component use specification automatic analysis working process is completed according to a two-layer planning principle of'process task-subtask '. And constructing and using an intelligent analysis model oriented to reusable component code semantic understanding to process the code understanding class subtasks. And constructing and using a contract analysis tool set to process the program analysis class subtasks. And constructing and using an auxiliary tool set to process the message conversion type subtasks. And integrating the analysis result of each process task, and outputting a usage specification report of the reusable component of the smart contract, thereby realizing efficient and accurate automatic reasoning of usage specification information of the reusable component of the smart contract.

This invention discloses an embedded system vulnerability detection method based on intelligent agent collaboration, aiming to address the shortcomings of existing technologies in vulnerability detection in complex embedded environments, including insufficient capability, lack of autonomy, and lack of adaptability. This invention establishes a structured representation of system state evolution by constructing a causal graph including code, hardware state, and physical feedback. It uses an autonomous hunting agent driven by a large language model as its core, autonomously generating a hunting plan based on the causal model and coordinating the use of various tools such as fuzz testing, symbolic execution, and static analysis for vulnerability discovery and verification. After discovering an initial vulnerability, it automatically infers and synthesizes a cross-domain exploit chain to assess the complete attack surface. Finally, it uses a reinforcement learning framework to enable the system to continuously learn from hunting experience and optimize its decision-making strategy. This invention upgrades vulnerability detection from passive pattern matching to an active exploration and reasoning process, achieving efficient and intelligent detection and exploit verification of unknown vulnerabilities in embedded systems.

This invention relates to a method and system for generating data quality rules and producing datasets. The method includes the following steps: S1, performing semantic analysis and mapping on the patterns of the data table to be governed based on an industry knowledge graph, and automatically inferring and generating candidate data quality rules with confidence assessment; S2, receiving specific business task requirements, selecting a matching subset of rules from the candidate rules, and using this subset of rules to progressively govern the original data, producing a high-quality dataset. This application automatically generates candidate data quality rules based on industry knowledge graphs, transforming rule formulation from manual to automatic generation, and transforming business personnel from rule writers to reviewers, significantly reducing the threshold and labor costs for rule formulation, and solving the problem of low efficiency in rule formulation relying on human experts; by selecting rules according to specific business task requirements and producing a high-quality dataset, targeted data generation is achieved, solving the problem of low efficiency in constructing high-quality datasets.

The application discloses a data reasoning method and device, equipment and a storage medium, and relates to the technical field of data processing. The method comprises the following steps: obtaining offline data to be reasoned; creating a plurality of reasoning tasks for the offline data, and respectively configuring resources occupied by the reasoning tasks to obtain corresponding resource configuration information; taking the offline data as input and executing each reasoning task according to the resource configuration information, so as to perform reasoning analysis on the offline data through the reasoning task and obtain corresponding reasoning results. According to the application, the resources occupied by each reasoning task are configured when the reasoning task is created, and the reasoning task can be automatically scheduled according to the resource configuration information, so that automatic reasoning analysis of offline data is realized, the reasoning process of offline data is simplified, and the labor cost is saved.

The invention provides an automatic inference analysis method and system for a business technology atlas in the field of power marketing, and the method comprises the steps: constructing a power marketing business technology atlas comprising a business layer, a process layer and a data layer through fusing multi-source heterogeneous data, and achieving the automatic inference analysis through combining an inference rule base and a graph traversal algorithm, and outputting a visual result and iteratively optimizing the reasoning model. According to the method, automatic process verification, data traceability and cross-business association analysis of the power marketing business can be realized, the problems of low efficiency and difficulty in problem traceability of traditional manual verification are solved, and the method is suitable for power marketing scenes such as business expansion access and the like.

Techniques are described for providing a SAT-based solver for a quantifier-free theory of strings and bit vectors. The solver can be used by an automated reasoning service of a cloud provider network to analyze policies and the consequences of policies. The solver reduces an input formula to a Boolean satisfiability problem by encoding the input formula into an equisatisfiable propositional formula, where the satisfiability of the equisatisfiable propositional formula is determined by a SAT solver. Rather than using a traditional DPLL(T) style algorithm, the solver described herein bounds the length of variables in an input formula and reduces the problem to a single formula, which can then be solved using incremental SAT solving. The solver can be used independently or as part of a portfolio of solvers used to determine the satisfiability or unsatisfiability of certain formula corresponding, e.g., to questions about users' policies within a cloud provider network.

The invention discloses an aviation knowledge graph optimization system combining multi-confidence automatic reasoning and artificial collaborative verification. The system comprises a rule base screening module, an entity-level confidence evaluation module, a relation-level confidence evaluation module, a global confidence evaluation module, a semantic confidence evaluation module and a confidence fusion device. The invention discloses an efficient triple evaluation system for an aviation domain knowledge graph, and aims to provide an efficient triple evaluation system for the aviation domain knowledge graph, improve the evaluation accuracy and calculation efficiency by optimizing an embedding technology and a neural network structure, and provide reliable support for constructing a high-quality aviation domain knowledge graph.

A generative pre-trained large language model (LLM) may create domain-specific text answers in various formats such as JSON, XML, HTML, SQL, or programming language. However, LLM may "appear illusion", generating incorrect or meaningless answers that deviate from reality, thereby weakening trust on its output or causing more serious consequences. The disclosed techniques use a sampling-based method and an equivalence checker. Generating, by the LLM, a plurality of answers (samples) to the cue; if the plurality of answers are equivalent, the LLM is likely to answer correctly. If the samples are not consistent or contradictory, it is more likely that the LLM has illusion, or ambiguity is prompted. The functional equivalence of the sample is validated using an automatic inference equivalence verifier, which provides a method of detecting and possibly correcting hallucination questions in the LLM generated answers.

Disclosed are techniques for automated reasoning via natural intelligence of unstructured data in order to generate meaning from unstructured data using a human-based logical reasoning framework. The disclosure provides solutions for a situation where two or more potential options are selected but where only one option is permitted. For example, the present disclosure addresses a need in medical coding where two codes are selected based on an automated reading of a medical report.

A device and a computer implemented method for determining rules for determining an answer to a question about content, in particular content representing a digital image, a time series, for example of sensor data, audio, synthetic data, or data that can be represented as objects and their relations, by automated reasoning on the question, and the content. The method includes providing a set of rules for determining the answer to the question; providing the question, the answer, and the content; determining a preprompt for a large language model explaining the set of rules; determining a prompt for the large language model instructing the large language model to add at least one rule to the set of rules such that the answer to the question about the content is derivable by automated reasoning on the question, the content, and the set of rules including the at least one rule.

An AI keystone system for medical decision support generates governed, auditable clinical insights. A medical data governance (MDG) engine ingests heterogeneous patient-related data from clinical systems, medical devices, and clinician inputs, validates and normalizes the data under governance rules, and emits structured, context-constrained representations for identified patients and episodes. One or more intermediary components collect and relay real-time device data so downstream processing operates only on governed data. A reasoning layer with artificial-intelligence models, including at least one large language model, cooperates with modular AI components to process clinician queries and generate candidate clinical outputs. A cryptographic governance subsystem evaluates candidate outputs against patient-safety rules, regulatory mandates, institution-specific policies, and stored consent artefacts using machine-verifiable constraints, enforces dual-factor clinician authentication, and maintains immutable, hash-chained audit trails with trusted timestamps that gate workflow progression. The system enables defensible use of AI-assisted clinical insights across hospital, clinic, and virtual-care environments.

This invention discloses an automatic reasoning platform and power distribution system for power big data, relating to the field of power system technology. The platform includes: a knowledge acquisition unit that collects and preprocesses internal and external power data, which is then transformed into structured data by a knowledge processing unit; a dynamic knowledge graph construction unit that organizes triplet data to form an initial graph; an update unit that monitors new dynamics in real time and corrects or adds graph content; and a reasoning unit that, based on the dynamic graph, utilizes fault diagnosis and demand prediction algorithms to achieve accurate reasoning about power faults and predict future demand trends. This invention achieves efficient integration and dynamic evolution of power knowledge through dynamic knowledge graph construction and real-time updates, ensuring the timeliness and accuracy of the graph and providing reliable data for fault diagnosis. Through a closed-loop architecture of data acquisition, reasoning analysis, scheme formulation, and execution control, combined with intelligent sensors and optimization algorithms, it generates distribution schemes adapted to different scenarios, improving the adaptability and economy of power distribution.

Techniques are described for providing a SAT-based solver for a quantifier-free theory of strings and bit vectors. The solver can be used by an automated reasoning service of a cloud provider network to analyze policies and the consequences of policies. The solver reduces an input formula to a Boolean satisfiability problem by encoding the input formula into an equisatisfiable propositional formula, where the satisfiability of the equisatisfiable propositional formula is determined by a SAT solver. Rather than using a traditional DPLL(T) style algorithm, the solver described herein bounds the length of variables in an input formula and reduces the problem to a single formula, which can then be solved using incremental SAT solving. The solver can be used independently or as part of a portfolio of solvers used to determine the satisfiability or unsatisfiability of certain formula corresponding, e.g., to questions about users' policies within a cloud provider network.