185 results about "Logical analysis" patented technology

The present application relates to a question answering interaction method and apparatus based on a large language model data vectorization query, and an electronic device and a computer-readable medium. The method comprises: performing semantic recognition on input data of a user, and determining a question category corresponding to the input data; determining a target vector database on the basis of the question category; converting the input data into an input vector; using the input vector to perform a vector query in the target vector database, and performing splicing to generate a query result; generating a prompt question by means of the input data and the query result; inputting the prompt question into a large language model, so as to acquire a logical analysis result; and when the logical analysis result is a positive result, performing question answering interaction on the basis of the logical analysis result. The present application can provide an intelligent question answering method for a user, thereby improving the answering quality and accuracy of a man-machine intelligent dialogue and question query.

The invention relates to the technical field of intelligent decision making, can be applied to business system platforms of financial science and technology, medical health and the like, and discloses a knowledge graph question and answer result generation method, device, equipment and medium, which comprises the steps of performing semantic analysis on a user question to obtain entity information and a query intention, performing logic analysis on the entity information and the query intention to obtain a question and answer result; obtaining an initial execution plan; calling a sub-graph query tool according to the initial execution plan, and querying sub-graph data by using the sub-graph query tool; performing linearization processing on the sub-graph data to obtain a sub-graph data text, extracting a candidate entity set in the sub-graph data text, and screening out a matched target entity from the candidate entity set; performing logic analysis on the target entity and the query intention to obtain a target execution plan, and screening out a matched atomic query tool from the candidate tool set; and querying a knowledge graph question-answer result by using an atomic query tool. The knowledge graph question and answer result generation accuracy can be improved.

The invention discloses a voice interactive console based on artificial intelligence, and relates to the technical field of intelligent consoles, a large language model is combined with context tracking and logic analysis technologies, a composite instruction is disassembled into independent tasks, a task chain is generated, a semantic weight is added to each task through a weight regression model, and the intelligent consoles are obtained. The task execution priority and the logic dependency relationship are identified, the limitation of a traditional system on the complex semantic analysis capability is overcome, a composite instruction can be accurately understood, a logic tree analysis method based on rules is adopted, a task chain is converted into a directed graph structure, the priority is dynamically adjusted, and a task execution sequence table is generated. In the scheduling process, multi-task parallel processing is carried out on tasks without logic dependence, and the resource utilization rate is increased; tasks with a dependency relationship are gradually scheduled by adopting topological sorting, so that the logic consistency is ensured.

The invention relates to the technical field of knowledge processing, and particularly discloses a knowledge proposition error correction method and system based on knowledge graph optimization upgrading, and the method comprises the steps: extracting effective information from multi-source knowledge data through employing entity linking, relation extraction and attribute alignment methods, and constructing an initial knowledge graph. Through semantic analysis based on deep learning, grammar, semantic and logic analysis is carried out on knowledge propositions, and a proposition element triple is generated. A graph neural network is utilized to encode triples, proposition vectors and neighborhood feature vectors are obtained, structural similarity is calculated accordingly, and proposition errors are detected in combination with rules and a statistical method. An integrated learning algorithm is adopted to classify error traceability, and an error correction decision is made in combination with multi-source evidence, so that an accurate error correction result is obtained, and the accuracy and reliability of knowledge propositions are improved; the error correction process is more logical and reliable, blind error correction is avoided, and the quality and efficiency of knowledge proposition error correction are improved.

The invention discloses an electric power information operation violation risk supervision system based on a knowledge graph, which relates to the field of violation risk supervision and comprises a dynamic graph construction module, a causal analysis module, a strategy analysis module, a strategy modeling module and a supervision decision module. The method comprises the following steps: obtaining multi-type electric power operation field sensing data and information system data, and carrying out data processing and knowledge graph dynamic construction to obtain a dynamic knowledge graph; performing risk situation quantification and risk decision point inference based on the dynamic knowledge graph to obtain a key causal decision point set; based on the key causal decision point set, through strategy logic analysis, obtaining a logic rule set which can be directly deployed and executed; according to the method, a logic rule set which can be directly deployed and executed is subjected to dynamic strategy evolution of a stochastic differential equation to obtain a dynamic strategy model, and the dynamic strategy model is subjected to decision optimal screening of measurement transformation to obtain an optimal supervision decision set, so that a risk value can be accurately calculated, and the supervision response speed can be increased.

The invention relates to the technical field of human resource management, and provides an intelligent paper marking system and method for talent selection and recruitment. The method comprises the following steps: cutting test paper according to a preset test paper template by using an image processing technology to obtain a plurality of plates corresponding to the test paper; converting the image content of each section of the cut test paper into a text format which can be edited and processed by adopting a character recognition technology; setting a preliminary scoring standard according to question type information and knowledge point information corresponding to the test paper, and optimizing the preliminary scoring standard by using a large language model; performing semantic understanding and logic analysis on the converted test paper text answers from a plurality of scoring dimensions including semantic accuracy, logic integrity and knowledge point coverage by using two preset scoring large models to give corresponding scoring information, and marking advantage information and defect information in the answers; and obtaining a comprehensive score corresponding to the test paper according to score information printed by each score big model for each question of each test paper.

The invention relates to the technical field of artificial intelligence in the education industry, in particular to an intelligent composition quality evaluation method and system based on a large language model, and the method comprises the steps: carrying out the text normalization and semantic unit segmentation of a composition, extracting a semantic vector through a first large language model in combination with a context enhancement strategy, and positioning a semantic fracture risk position; recognizing composition core elements through a second large language model, and mapping the composition core elements back to the semantic unit sequence; constructing a demonstration logic diagram, extracting a core demonstration path and abstracting the core demonstration path into a logic role topological graph; in combination with a pre-constructed writing specification knowledge graph, comparing structural compliance, connection strength and an expected support relationship, identifying and demonstrating logic defects, and generating a global deduction item list; semantic clustering is carried out on illegal items to form an error label set, comprehensive weight is calculated in combination with historical data of students, and core weak items are positioned; according to the application, the logic analysis depth of intelligent evaluation of the argument is remarkably improved, and the pertinence and practicability of teaching feedback are improved.

The invention relates to a spacecraft FMEA application method based on a SysML model and a knowledge graph. The spacecraft FMEA application method comprises the steps that logic analysis is conducted on functional faults of a spacecraft, and a functional fault logic model and a fault mode basic element semantic model are obtained; according to the function fault logic model and the fault mode basic element semantic model, utilizing an SysML model tool to construct an FMEA digital model of the spacecraft; an XML metadata conversion method is adopted, and the FMEA digital model is converted into FMEA structured data; mapping concepts in the fault mode basic element semantic model and subclasses and relationships thereof into concepts, subclasses and relationships of an FMEA knowledge graph ontology model to obtain the FMEA knowledge graph ontology model; fMEA structural data and historical fault information are obtained through FMEA digital model conversion, an FMEA knowledge graph ontology model is instantiated, and an FMEA knowledge graph is established; based on the FMEA knowledge graph, spacecraft FMEA analysis is carried out, automatic statistical calculation of fault mode occurrence frequency and fault mode reduction measure recommendation are achieved, and the aim of reducing spacecraft design risks and on-orbit operation and maintenance risks is achieved.

The invention discloses an intelligent medical record quality control method based on a rule and large language model hybrid engine, and relates to the technical field of natural languages, and the method comprises the steps: obtaining original electronic medical record data from a medical information system; performing post-structured processing on the original electronic medical record data, and converting an unstructured or semi-structured text in the original electronic medical record data into a unified structured data object readable by a machine; traversing a plurality of quality control points in a preset quality control rule base, and for each quality control point, adaptively selecting and executing a corresponding quality control path based on the converted structured data object; and based on the quality control results of all the quality control points, performing score calculation and generating a quality control report. According to the method, the automatic and deep quality control of the full-dose medical record can be realized, form inspection is covered, connotation logic analysis can be executed, the system is endowed with excellent flexibility and regional expansion capability, and reliable technical support is provided for improving the overall level of medical quality management.

The invention relates to the technical field of task execution, and provides a workflow parallel computing execution method and system, terminal equipment and a storage medium. The workflow parallel computing execution method comprises the steps that a to-be-executed task is obtained, and after the to-be-executed task is subjected to logic analysis, the to-be-executed task is obtained; dividing a to-be-executed task into a plurality of sub-tasks according to the task type, constructing a task dependency graph, and making an execution model according to the characteristics of the sub-tasks; according to the resource states and the task priorities of the sub-tasks, resource allocation is carried out on the sub-tasks, and an execution sequence is set for the sub-tasks based on the dependency relationship in the task dependency graph; according to the execution mode and the execution sequence, the subtasks enter a parallel execution state, and an execution result of each subtask is obtained; and according to the original logic and data dependence of the to-be-executed task, summarizing, sorting and combining the execution results to obtain a summarized result, and verifying the summarized result to obtain a final execution result.

The invention relates to the technical field of intelligent wheelchair scheduling, in particular to an intelligent wheelchair scheduling method and system. The method comprises the following steps: extracting an obstacle avoidance failure decision in a historical scheduling task, and generating a clustering logic of decision failure through clustering analysis; analyzing the motion trail deduction deviation of the sudden obstacle based on clustering logic, and further performing sudden risk pre-judgment lag analysis to obtain lag regression data; and lagging regression data is utilized to carry out front deviation correction memory learning of wheelchair scheduling braking, decision is optimized through reinforcement learning iteration, and finally optimized decision data is output. According to the method, the intelligent wheelchair scheduling technology is optimized, so that the intelligent wheelchair scheduling technology is more perfect.

The invention discloses a power grid operation safety patrol risk prediction method fused with an artificial intelligence technology, and relates to the technical field of power grid operation safety patrol, and the method specifically comprises the following steps: carrying out the operation logic analysis of all recognized operation tasks which are executed in parallel, screening out all operation tasks with operation conflicts, and carrying out the operation logic analysis; an operation mutual interference task cluster is formed; parallel job interference feature information in the operation mutual interference task cluster is obtained and analyzed, and the task interference degree between job tasks in the operation mutual interference task cluster is evaluated; and according to an evaluation result, executing dynamic regulation and control on the job task input data set for risk prediction. According to the method, the problem of task conflict caused by operation requirements and security constraint differences of different operation tasks during parallel execution of multiple tasks in a power grid operation patrol process is solved, and the task interference degree is accurately evaluated by dynamically regulating and controlling the operation task input data set, so that the accuracy of risk prediction is improved.

The invention belongs to the technical field of criminal investigation three-dimensional scene reconstruction and judicial deduction, and discloses a criminal investigation case three-dimensional scene reconstruction and judicial deduction virtual reality platform, which comprises the steps of fusing field cross-modal evidences, anchoring space-time coordinates through Beidou-block chain double references, and detecting material evidence conflicts. Constructing an adaptive arbitration factor based on a forensic parameter library, and generating a material evidence fusion packet; a resource slicing strategy is constructed to divide four-dimensional resource levels, text description is converted into a skeleton action topological chain, a space-time trajectory simplified set is constructed, case deduction is synchronously triggered, and a resource arbitration log is generated; constructing a physical deduction rule set, executing rainstorm scouring correction, and generating a physical deduction instruction set; performing case deduction resource scheduling, generating a three-dimensional deduction animation stream, constructing a judicial causal chain, and generating a crime logic analysis report; thermally loading the desensitization rule, and identifying and replacing the sensitive field; quantizing the deduction effectiveness, and optimizing and updating the forensic parameter library and the resource slicing strategy.

The invention discloses a new-generation emergency command system scene analysis system integrated with a convolutional neural network, particularly relates to the technical field of computer vision and intelligent emergency command, and is used for solving the problem that the dynamic association and situation evolution logic analysis capability among emergency scene elements is insufficient in the prior art. Visual element features are extracted from a video stream through a feature extraction module by using a convolutional neural network, an association graph construction module constructs a space-time association graph based on the features, a movement evaluation module calculates element group movement feasibility indexes, and a situation analysis module analyzes divergence features and antagonism degrees in combination with movement track convergence. And the risk identification module identifies a derived risk aggregation mode according to the situation prediction, and the semantic generation module finally outputs high-level semantic description information containing situation prediction to realize closed-loop analysis from scene perception to cognitive decision.

The invention provides an intelligent diagnosis system of a machine room air conditioner refrigeration system. The intelligent diagnosis system comprises a data acquisition module, a data preprocessing module, a logic diagnosis module, an LSTM neural network module, a fault prediction module and a fault diagnosis module. The data acquisition module is used for collecting operation data of the air conditioning equipment. The data preprocessing module preprocesses original data. And the logic diagnosis module performs preliminary fault diagnosis. The LSTM neural network module learns and memorizes the long-term dependency relationship of the operation data of the air conditioning equipment by training an LSTM neural network. And the fault prediction module is used for processing and analyzing the real-time operation data of the air conditioning equipment. And the fault diagnosis module performs analysis according to the long-term dependency relationship and the fault data. According to the equipment, main operation data of an air conditioner compressor and other core parts are collected, on the basis of data collection, classification and feature extraction, comprehensive logic analysis is firstly carried out after big data screening, and then possible alarms are diagnosed and pre-judged in advance according to features.

The invention belongs to the technical field of artificial intelligence, and discloses a spherical tank weld joint intelligent detection system fusing dynamic convolution and three-dimensional reconstruction, which comprises a structural logic analysis module for analyzing a spherical tank CAD design drawing and a process welding process document, extracting each weld joint node position, a connection sequence, a welding path and a corresponding spherical topological relation, and determining a weld joint node position; constructing a structural logic graph; the welding seam point set acquisition module is used for extracting a welding seam candidate region in each frame of point cloud on the basis of a welding seam theoretical track provided in the structural logic map, and identifying an actual welding seam point set through combined clustering of reflection intensity and geometric features; the method is suitable for automatic detection of complex curved surface structure welding seams, the intelligent and automatic degree is improved, and a reliable solution is provided for intelligent detection of spherical tank welding seams.

The invention relates to the technical field of chip design, in particular to a test method and system for a chip simulation model. The method comprises the following steps: acquiring chip data, and performing chip circuit logic analysis to obtain circuit logic data; performing signal transmission simulation based on the circuit logic data to obtain signal transmission data; constructing a chip simulation model according to the circuit logic data and the signal transmission data; performing particle bombardment on the chip simulation model to obtain particle data; performing energy spectrum classification based on the particle data to obtain proton data and heavy ion data; performing single event upset fault analysis on the chip simulation model according to the proton data to obtain single event upset data; and performing multi-particle upset fault analysis on the chip simulation model according to the heavy ion data to obtain multi-particle upset fault data. The reliability of the chip in the radiation environment is evaluated and optimized based on the chip design technology, so that the radiation resistance of the chip is improved.

The invention relates to the technical field of computer science, and discloses a cryptographic logic vulnerability detection method and system based on a large language model. The method comprises the following steps: firstly, constructing a diversified cryptographic vector knowledge base, and organizing and managing through text vectorization; based on cryptographic algorithm standards such as FIPS, performing logic compliance check on the target code; in combination with a small number of examples and thinking chain prompts, guiding a large language model to identify potential logic vulnerabilities; summarizing a compliance check and reasoning process result, and retrieving related knowledge in a knowledge base based on cosine similarity to enhance model judgment; and finally, integrating knowledge enhancement and logic analysis results, and outputting a structured vulnerability detection report. According to the method, a universal cryptographic vulnerability detection framework which does not need to run codes and is suitable for multiple programming languages is constructed, automatic recognition and detection of complex logic defects in cryptographic codes are achieved, and the accuracy and reasoning ability of a large language model in a cryptographic vulnerability detection task are improved.

The invention discloses an atmospheric pollution case system, a case analysis method and a medium, and the system comprises a basic information layer which is used for storing the whole-process feature data and treatment measure data of a pollution event; the logic analysis layer is used for carrying out quality control on the data, carrying out source analysis and cause analysis on the whole-process feature data of the pollution event after the quality control, and carrying out effect evaluation on the treatment measures after the quality control; and the comprehensive decision-making layer is used for comparing historical cases according to the retrieval conditions and matching similar historical cases, and finding out treatment measures corresponding to the similar historical cases according to the similar historical cases, or formulating targeted new treatment measures according to the corresponding treatment measures. The invention provides a'basic information-logical analysis-comprehensive decision 'three-layer progressive architecture case library, so that other modules can be conveniently expanded on the architecture of the case library subsequently; and multi-source data are systematically integrated, so that the problem of fragmentation of existing case data is solved.

The invention discloses a general data exchange system based on a configurable label structure, which relates to the technical field of data processing and information integration, and comprises a parameter module used for establishing an identity authentication mechanism of two exchange parties; the business document module is used for establishing a data structure definition based on a preset exchange demand; the receipt logic module is used for carrying out logic analysis and code decoupling on the data structure definition, generating a marked language script code and an execution instruction, constructing an execution matrix and realizing mapping between a database physical field and a memory logic variable; and the standard adaptation module is used for comparing and checking the content of the generated data and feeding back a checking result to the parameter module to correct the exchange rule and parameter configuration. A more intelligent logic separation strategy is implemented through structural insight, data reading, writing and transmission are cooperatively controlled, complete decoupling of service logic and program codes is achieved through a nested tag structure, and flexibility, universality and maintenance efficiency of data exchange are optimized.

The invention discloses a road accident prediction system based on big data, and relates to the technical field of big data analysis and prediction, the road accident prediction system comprises an acquisition module, a fusion module and a prediction module, the acquisition module obtains multi-source data of road accident prediction, transmits the multi-source data to the fusion module, carries out space-time alignment processing on the multi-source data through a processing unit, and carries out prediction on the multi-source data; generating feature vectors, transmitting the feature vectors to a prediction module, generating an accident risk prediction result through a prediction unit according to the real-time input feature vectors, constructing a three-dimensional data set by integrating multi-source data, establishing a unified coordinate system by means of space-time grid division, and realizing data differential fusion, static parameter positioning according to road section ID, and dynamic data space-time two-dimensional association. Historical accident space-time backtracking matching is carried out, a feature set with a space-time label is generated, the problem that traditional data benchmarks are not uniform is solved, statistical features are extracted through multiple windows in the prediction stage, multi-factor coupling is analyzed in combination with risk assessment logic, prediction space-time resolution and robustness are improved, and accurate safety management is supported.

The invention discloses an electromagnetic spectrum identification and tracking algorithm based on transform. Comprising the following steps: step 1, electromagnetic spectrum signal acquisition, step 2, electromagnetic spectrum data preprocessing, step 3, Transform model construction, step 4, target identification, step 5, frequency hopping logic analysis, step 6, target signal parameter acquisition, step 7, target signal type customization, step 8, alarm and countering processing, step 9, historical information playback, and step 10, electromagnetic spectrum signal acquisition, step 5, frequency hopping logic analysis, step 6, target signal parameter acquisition, step 7, target signal type customization, step 8, alarm and countering processing, step 9, historical information playback, and step 10. And optimizing and updating the algorithm. According to the method, interlaced signals in a complex electromagnetic environment are accurately identified, long-distance weak signals are detected with high sensitivity, frequency hopping signals are accurately tracked, the false alarm rate and the missing report rate are reduced, preprocessing is optimized through a band-pass filter and wavelet transform, the feature extraction capacity is improved through a Transform model, the type of a user-defined target signal is supported, continuous optimization is carried out according to feedback, and the detection accuracy is improved. And high performance and adaptability are maintained.

The invention discloses an English teaching writing error diagnosis method based on artificial intelligence, and relates to the technical field of artificial intelligence, and the method comprises the following steps: obtaining an English composition text, and obtaining initial writing data; performing grammatical structure analysis according to the initial writing data to obtain grammatical features; performing vocabulary error analysis according to the grammatical features to obtain vocabulary errors; performing syntactic and semantic consistency analysis according to the vocabulary errors to obtain syntactic and semantic errors; performing article logic analysis according to the syntactic and semantic errors to obtain article logic errors; and performing comprehensive diagnosis according to the article logic error to obtain a diagnosis result. According to the method, the conjunction matching degree and the semantic compatibility are combined, cross-sentence logic analysis is achieved, misjudgment of isolated sentences is avoided, a complete error set is formed, the comprehensiveness and accuracy of diagnosis are improved, and practicability and guidance are enhanced.

The invention discloses an AI homework correction method and system based on a large model, and the method comprises the steps: obtaining text data and question requirements from answers submitted by students, carrying out the word segmentation and syntactic analysis of the answer content through a natural language processing technology, and obtaining the semantic representation and logic structure of the answer content; according to the semantic representation and the logic structure of the answer content, performing context understanding by adopting a pre-training language model, extracting key arguments and arguments in the answer content, and determining the integrity and continuity of the answer content; according to the integrity and coherence of the answer content, analyzing the association strength between the key argument and the argument by adopting a logical reasoning algorithm, judging whether the logical structure of the answer content is reasonable or not, and obtaining a logical analysis result; and obtaining an unreasonable part from a logic analysis result, matching a question requirement and a standard answer by adopting a knowledge graph technology, and determining a deviation point and an improved space of answer content. The teaching resource configuration is optimized, and the teaching effect and the learning quality are remarkably improved.

The invention relates to the technical field of artificial intelligence and natural language processing, in particular to a sensitive word and intention recognition method and device based on multiple modes and a medium. According to the method, feature extraction is carried out on different modal data, and specific information of each modal is reserved; characteristic extraction is directly carried out on multi-modal characteristics, and potential association and interaction information of different modals can be captured. Therefore, in the feature extraction stage, original data or feature vectors of different modes are effectively fused, association and complementarity among information of each mode are fully mined, comprehensive and comprehensive screening of multi-mode information is realized, and information omission caused by mode separation is avoided. Meanwhile, the information intention is deeply mined through multi-dimensional means such as sentiment analysis, semantic logic analysis and behavior pattern analysis, so that sensitive words can be found, the real intention of a publisher can be known, more targeted measures can be taken, and the health and safety of the network environment can be maintained.

The invention discloses a resource-oriented academic paper technical scheme automatic reproduction method. The method comprises the steps of obtaining text content of a target academic paper and generating a resource demand list; technical resources are retrieved and screened on an open source platform, and a resource support library is constructed; generating a comprehensive reproduction blueprint in combination with the resource support library and the text content; formulating a resource use guide, and coordinating a plurality of large language models to perform logic analysis on each module according to a dependency sequence to obtain initial technical implementation; iteratively repairing the initial technical implementation to obtain a repaired technical implementation; and performing fine adjustment on the quality evaluation model by adopting a reinforcement learning algorithm to maximize the matching degree between the model output and the key technical key points of the thesis, and performing quality evaluation and optimization on the repaired technical implementation by utilizing the fine-adjusted model to complete reproduction of the target academic thesis. According to the reproduction method provided by the invention, resources can be effectively utilized, and the reproduction success rate of complex papers is improved.

The invention discloses a subjective question intelligent marking system and method based on a large language model, and relates to the technical field of online education and examination intelligent marking, and the system comprises a scoring rule configuration module, an AI scoring engine module, a teacher review interaction module and a result presentation and archiving module. The scoring rule configuration module is used for receiving AI scoring related parameters configured by a user; the AI scoring engine module integrates a large language model and is used for receiving test question related information, examinee answers and AI scoring related parameters and carrying out natural language understanding, knowledge reasoning, logic analysis and standard comparison on the basis of the large language model to obtain an AI scoring result; the teacher re-checking interaction module provides an interaction interface for the user to check the AI scoring result to obtain a final scoring result; and the result presenting and archiving module is used for archiving, storing and visually displaying the final scoring result. According to the invention, the manual reviewing cost can be reduced, and the accuracy and efficiency of subjective question reviewing are improved.

The application provides an engineering education system and method fusing explainable multi-objective intelligent optimization, and relates to the technical fields of intelligent education and artificial intelligence optimization. The method comprises the following steps: obtaining an initial engineering design scheme of a user, and obtaining a standardized multi-dimensional index vector through physical simulation analysis and logical analysis; configuring weights with structured reasons for each index, identifying critical weight values leading to decision reversal through sensitivity analysis, and then determining optimization objectives and constraint conditions; performing multi-objective optimization search using an explainable multi-objective optimizer, generating a Pareto optimal solution set and outputting explainable data; simultaneously generating a behavior interaction sequence of the user; generating a capability evaluation report of the user according to the behavior interaction sequence; and adjusting a preset teaching logic sequence according to the capability evaluation report to form a closed-loop teaching process. The application is suitable for cultivating and evaluating the decision-making ability of students in balancing technical performance and environmental sustainability in engineering design.

The invention relates to the technical field of voice real-time translation, in particular to a voice real-time translation and proofreading system oriented to cross-border e-commerce, which comprises the following modules: a voice semantic processing module used for receiving voice input, converting the voice input into text data, performing semantic analysis on the text data, extracting commodity names, attribute information and logic relationships, and sending the extracted commodity names, attribute information and logic relationships to a server; semantic structure data is generated and stored in a context cache; and the anaphora analysis module is used for reading the semantic structure data from the context cache, detecting anaphora words in the semantic structure data, determining anaphora objects and updating the semantic data subjected to anaphora resolution to the context cache. According to the method, word segmentation, part-of-speech tagging, syntactic analysis and dependency analysis are carried out on a text, commodity names, the number, colors, sizes and attribute relations are extracted to generate structured semantic data, the structured semantic data are stored in a context cache, and when anaphora words are detected, anaphora objects are determined through semantic similarity and context logic analysis, and the semantic data are updated.

The invention provides an evidence-driven large language model MISRA C rule review method and system for aviation safety. According to the method, formalized feature extraction is carried out from four aspects of grammar structure features, semantic structure features, logic analysis features and preprocessing behavior features for target codes according to various rules of MISRA C forcing class rules, and diagnosis information is diagnosed in combination with an industrial-grade compiler; generating a structured evidence set comprising abstract syntax tree node statistical features, code context structure features, control flow graph features, function internal data flow analysis features, symbol and type table features and macro definition analysis features; and based on the structured evidence set and the large language model, executing evidence tracing, rule matching and logical reasoning according to a preset thinking chain process, and outputting a structured judgment result containing illegal rule numbers, evidence description and code positions. According to the method, the false alarm rate can be remarkably reduced while the high rule coverage rate and the detection accuracy are kept.