142 results about "Logic structure" patented technology

The invention relates to the field of computer vision, and particularly provides a table recognition reconstruction method and system, a terminal and a medium, and the method comprises the steps: firstly decomposing a large-size table image into a plurality of overlapped sub-images, and carrying out the table structure detection and OCR character recognition of each sub-image through parallel recognition; then, sub-graph recognition results are integrated through a coordinate mapping and confidence coefficient weighted fusion algorithm, and boundary errors are eliminated; then, automatically distinguishing common cells based on an area clustering algorithm, merging the cells and a header region, and reconstructing a complete table logic structure; further understanding header semantics through a natural language model and repairing identification errors; and finally, realizing intelligent splicing and standardized output of the cross-page table. According to the method, the memory limitation of the traditional OCR technology is broken through, an oversized table can be processed, the recognition accuracy of a complex structure is improved, and the digitization efficiency of professional documents such as financial statements and engineering drawings is improved.

The invention discloses a multi-modal index knowledge base and a construction method thereof. The construction method comprises the following steps: processing a heterogeneous document to obtain a semi-structured document; identifying a title hierarchical relationship of the document to construct a document logic structure; carrying out minimum chapter blocking on the text of the semi-structured document to obtain logic blocks; performing semantic segmentation on each logic block to obtain text blocks; the method comprises the following steps: constructing text block nodes by meta-information of text blocks, constructing non-text block nodes by meta-information of non-text elements, extracting document nodes, chapter nodes and chapter-chapter inclusion relationships according to a document logic structure, respectively extracting semantic information from the text blocks and the non-text elements, and storing the semantic information in a database; recording the corresponding relationship between the text block nodes and the semantic information and between the non-text block nodes and the semantic information; constructing a knowledge graph based on each node and relationship and storing the knowledge graph into a graph database; and constructing semantic knowledge based on the semantic information and storing the semantic knowledge into a vector database. According to the scheme, lossless retention of multi-modal information and structured organization of document logic are realized, and efficient indexing and accurate recall are facilitated.

The invention discloses a power distribution network layered optimization method oriented to source-network-load interaction, and relates to the technical field of power system automation. The method comprises the following steps: constructing a dynamic adaptive collaborative architecture, and dynamically generating a three-layer logic structure of global coordination, regional autonomy and local execution according to the real-time state of the power distribution network; operating a three-layer mixed game mechanism, and determining an optimal strategy of each main body through master-slave, cooperation and bilateral interactive games; executing an evaluation-planning-operation three-level closed-loop optimization process, and taking a key performance index fed back by an operation layer as a basis for scheme correction of a planning layer; and a mixed intelligent strategy combining multi-agent deep reinforcement learning and an improved particle swarm optimization algorithm is adopted for solving. According to the method, the problems of hierarchical solidification, benefit coordination imbalance and planning operation splitting in the prior art are effectively solved, and the adaptive capacity, the economical efficiency and the flexibility of the power distribution network in the high-permeability source-network-load interaction scene are improved.

The invention discloses a hydraulic machinery feasibility research report grading review method based on a large language model, comprising the following steps: collecting and analyzing an original hydraulic machinery feasibility research report to generate a logic structure tree and a structured text; based on the logic structure tree, hierarchical gating and strategy scheduling are executed, and gating scheduling data are generated; according to review strategy parameters in the gating scheduling data, executing three-constraint parallel partitioning on the structured text to obtain a sub-document set; utilizing the gating scheduling data to execute auditing adversarial compliance review on each sub-document of the sub-document set to form a preliminary review conclusion set; and aggregating the preliminary review conclusion set, executing cross-subdocument consistency judgment and result summarization, and generating a final review conclusion. According to the method, the compliance problem of the hydraulic machinery research-length document report can be completely and accurately identified.

The invention discloses an intelligent chart generation method and system suitable for multi-modal data and a storage medium, and the method comprises the steps: receiving and preprocessing the multi-modal data, and completing user intention type recognition and task routing; selecting an adaptive model according to a routing result, and constructing and generating a cue word; standardizing an initial graph code output by the model into an intermediate representation format containing a metadata field, converting the intermediate representation format into a graph code compatible with a target platform, and reserving node definition, a connection relation, layout constraint and metadata information; the method comprises the following steps: respectively completing chart type identification and layout strategy determination, node format size adjustment, overall layout alignment correction and partition adjustment, and connection line identifier optimization through four optimization components, and rendering and outputting a chart. The problem that due to the fact that an existing model lacks a special optimizing and processing mechanism for the structured chart, the effect is remarkably poor when a professional schematic chart needing an accurate logic structure and geometric constraints is generated is solved.

The invention discloses an electrical drawing security collaborative management method, system and device based on a cloud computing technology and a medium. The method comprises the steps that an initial key is generated through a quantum key distribution device, the key is divided into a plurality of sub-blocks, and a dynamic session key is generated; encrypting data blocks divided by the electrical drawing file according to the logic structure through the dynamic session key, and transmitting the encrypted data blocks to a cloud computing platform for storage; generating a temporary key to encrypt the modified content, adding a digital signature and position information, and combining the modified content after the cloud verifies the signature to generate a new electrical drawing; synchronizing the new electrical drawing full-text security operation credential update record to the block chain, and verifying the consistency of the data on the chain and the cloud drawing version; and splitting the dynamic session key into a plurality of fragments and storing the fragments in different cloud nodes. According to the invention, safe transmission, efficient collaborative management and accurate data tracing of the electrical drawings are realized, and the safety, reliability and efficiency of full-life-cycle management of the electrical drawings are remarkably improved.

Disclosed herein are system, method, and computer program product aspects for textual data extraction, generation, and evaluation. Text is input into a first fine-tuned large language model (LLM) to generate an atom (e.g., a textual phrase in a particular category). The atom is input into a second LLM that has been fine-tuned for structured output corresponding to the particular category of information. A logical structure is generated based on a structured output of the second LLM, wherein the logical structure represents the textual phrase of the atom and contains a contextual attribute associated with the textual phrase. The embodiment then stores the logical structure into a knowledge graph as a modifier node having a time-variant attribute (e.g., a timestamp associated with the textual phrase and / or the contextual attribute).

The invention discloses an engineering document generation method and system based on a logic structure tree and LLM and a medium. The method comprises the steps that original data are analyzed to form an initial text segment set X; calling LLM to generate an initial technical draft based on X, extracting a logic structure, and constructing an initial logic structure tree T; maintaining a search queue, for leaf nodes in the queue, acquiring context information related to semantics from X through a retrieval enhancement mechanism, after same-level brother node deduplication and global deduplication judgment, calling LLM to expand the nodes to generate child nodes, optimizing text description, and recursively enriching a tree structure; and finally, based on the X and the optimized tree, calling LLM to generate a title, an abstract and a background, generating content, details and key point chapters through traversal, and combining into a complete engineering document. The system comprises a document analysis module, a knowledge base module, a retrieval enhancement module, a logic structure tree maintenance module and a document generation module. The method is high in automation degree, and the generated document is clear in structure, complete in content and high in specialty.

The invention discloses a report generation method based on a large model, electronic equipment and a computer program product, and relates to the technical field of artificial intelligence, the report generation method comprises the following steps: identifying a logic structure, a placeholder and a task dependency relationship in a preset report template, and initializing a context state carrier based on the logic structure, the placeholder and the task dependency relationship; obtaining a report generation requirement, converting the report generation requirement into SQL query data, and generating a cue word based on the SQL query data; and inputting the cue word into a preset generation model to obtain filling content, updating the filling content to the context state carrier, and filling placeholders of the report template based on the context state carrier to obtain a final report. According to the method, the defect of low efficiency of manual data arrangement is overcome, and the problem that a simple template filling tool cannot process a complex logical relationship is solved, so that the report generation efficiency is improved.

The invention relates to the technical field of multi-modal document analysis, and discloses a multi-modal document analysis method and device, electronic equipment and a storage medium. The method comprises the steps that picture elements in a document are automatically recognized and extracted through a third-party library module, placeholders are generated in combination with picture positions, and the placeholders are stored in the third-party library module; and then semantic understanding and text generation are carried out by using a multi-modal large model, so that the visual content is converted into a structurable information text. The method has the advantages that by introducing the multi-modal large language model, unified analysis of multi-modal information such as texts, pictures, tables and flow charts is achieved, layout semantics and logic structures of documents are reserved, fusion analysis of image-text content is achieved, the integrity and the intelligent level of document analysis are remarkably improved, and the method is suitable for large-scale popularization and application. Compared with a traditional single-mode method, the method has the advantages that unified understanding and structured expression can be carried out on multi-mode content, and richer and more accurate structured results can be output.

The invention relates to a streaming data real-time intelligent analysis processing method and device, equipment and a medium. The method comprises the following steps: acquiring a query request for streaming data, and analyzing a logic structure of the query request to obtain a corresponding query syntax tree; based on the query syntax tree, metadata of a unified storage cluster is read, and a physical execution path is generated; estimating the cost of each physical execution path based on a preset cost rule, and selecting the physical execution path with the minimum cost as a physical execution plan; executing the physical execution plan, obtaining a target data set from the unified storage cluster, and analyzing the target data set through a preset machine learning analysis model to obtain a deep analysis result set; and generating a final analysis report corresponding to the streaming data based on the deep analysis result set. By adopting the method, the resource utilization rate can be improved, and real-time data and historical data are considered for analysis.

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 distributed database query optimization, in particular to a distributed database query dynamic optimization method and device and electronic equipment.The distributed database query dynamic optimization method includes the steps that a multi-dimensional feature unified coding framework is constructed, an SQL logic structure, a fragment topology state and a historical load rule are fused into a three-dimensional tensor, and the three-dimensional tensor is obtained; and a dynamic routing decision mechanism is formed by combining the multi-scale feature extraction capability of the convolutional neural network. According to the scheme, the problem of dynamic response missing caused by feature splitting of a traditional method is effectively solved, the execution efficiency of complex query is remarkably improved by quantifying a fragment weight matrix to optimize a node selection strategy, the utilization rate of calculation and network resources is improved, and the adaptability to a long-period load rule is enhanced; and full-link optimization from feature coding to intelligent decision making is realized.

The invention relates to the technical field of intelligent chips, in particular to a low-delay interface control module priority arbitration method for a chip, which comprises the following steps of: receiving a plurality of access requests, associating static priorities, QoS (Quality of Service) levels and timestamps, and generating dynamic sorting counts; and the dynamic sorting count is decreased progressively through a QoS differential aging mechanism, an updated sorting count is generated, conflict arbitration is performed by using a secondary arbitration strategy, and the access type of the optimal request is determined. Grouping the access requests based on the access type of the optimal request, calculating a sorting counting difference value, and generating a quantization difference value; and constructing a load decision model, determining a corresponding dynamic switching threshold value, comparing and judging the quantization difference value through the dynamic switching threshold value, and generating a switching enable signal. And calculating an authorization signal of the optimal arbitration request through the flattened logic structure, and outputting bus authorization information. By constructing the flattening arbitration method, the balance among high throughput, low delay and high hardware efficiency is realized.

The invention relates to a code generation method based on a large model dynamic routing mechanism, which comprises the following steps of: constructing a difficulty-perceived task routing module, and performing complexity judgment on an input programming problem; for simple problems, code generation is carried out by directly adopting few-sample prompt, and complex reasoning of excessive thinking is avoided; for complex problems, an intention thinking chain is introduced, reasoning is divided into two stages of specification description and thought generation, and the method is used for explicit modeling of task intention and logic structures. Then, according to a difficulty judgment result of the task, executing a generation process along the determined scheduling path: for a simple task, outputting a code by adopting a direct generation path; for a complex task, an intention thinking chain is generated firstly, and then a structured code consistent with a task intention is generated based on the thinking chain. By using the method provided by the invention, efficient generation of the programming task in the industrial intelligent system can be realized.

A compare circuitry that can be implemented in a memory circuit includes: selection logic coupled to a sense amplifier of the memory circuit, the selection logic structured to receive a sense amplifier enable signal, a column output from the sense amplifier, and a column output bar from the sense amplifier, wherein the selection logic is structured to pass through the column output and the column output bar when the sense amplifier enable signal is active and hold the column output and the column output bar at a same value when the sense amplifier enable signal is inactive; and an XNOR gate structured to receive the column output and the column output bar passed through the selection logic, a data input, and a data input bar and output a bit compare output.

This invention relates to the field of teaching aids technology, and more particularly to a method for adaptive generation of speech content based on a multimodal large model and a real-time question-answering system. The method includes the following steps: acquiring multimodal raw materials provided by the speaker and parsing them using a multimodal large model to extract the speech's logical structure, key entities, and content attribute tags; organizing the multimodal large model parsing results into a content tree and constructing a knowledge graph representing arguments, sub-arguments, cases, data, and conclusions, wherein the knowledge graph defines support, exemplification, and comparison relationships between nodes; learning from the speaker's historical audio / video and text data and outputting a speaker style vector; generating an initial speech suite and configuring adjustment strategy templates, priority definitions, and safety boundaries. In this invention, by organizing content into an operable knowledge graph / content tree and constructing a real-time closed loop of "perception—understanding—decision—generation," dynamic path selection and content generation based on real-time multimodal perception are achieved.

The application discloses a complex specification multi-modal understanding method and system based on YOLO two-stage segmentation, comprising the following steps: performing pre-processing on an input specification image; inputting the pre-processed image into a coarse-grained detection stage of an improved YOLOv8 network to generate a coarse-grained candidate region set; inputting the coarse-grained candidate region set into a fine-grained segmentation network based on an improved YOLOv8 feature extraction structure; constructing a layout hierarchical relationship structure based on an accurate layout segmentation result; according to the layout hierarchical relationship structure, performing multi-modal feature extraction on the accurate layout segmentation result, and calculating a matching weight; inputting a graphic-text semantic alignment result and the layout hierarchical relationship structure into a logical structure modeling module together to construct a layout element graph; and generating a knowledge graph semantic output result according to a structured logical reasoning result. The application realizes automatic structured understanding of a complex specification by adopting YOLO two-stage segmentation and multi-modal semantic alignment.

The application discloses a multi-agent collaborative reasoning method and system based on memory sharing and conflict resolution, comprising: constructing a hybrid structured memory pool containing a vector database for semantic recall and a graph database for storing logical structure; deploying functionally heterogeneous multi-agent to normalize original multi-source data into observation nodes and write into the graph database; using the graph database trigger mechanism to listen to the writing event, and automatically triggering the reasoning to generate the hypothesis node and establish the deduction edge when matching the preset rule; detecting whether there is a logical conflict for the hypothesis node of the same subject, and triggering dynamic arbitration if there is; the arbitration mechanism calculates the weighted score based on the confidence, data source credibility and context consistency calculated based on KL divergence, and finally outputs the reasoning conclusion and evidence chain path. The application breaks the information island through the hybrid memory pool, realizes the automatic reasoning by using the trigger, and significantly improves the accuracy, logical consistency and explainability of reasoning in complex scenarios.

The invention relates to a trade rule reasoning method and system based on bidirectional mapping and a storage medium, and relates to the technical field of trade management.The trade rule reasoning method comprises the steps that a trade rule text is collected and split, and a plurality of rule terms are obtained; analyzing the rule terms according to preset key entity elements to obtain rule condition segments, and combining the rule feature weight values to obtain trade rule feature vectors; identifying and converting trade rule feature vectors according to the rule attributes and the dependency relationship between the rules, and constructing a rule dependency graph; converting the trade rule text into formalized rule elements according to the semantic framework and the logic structure in combination with the rule dependency graph; generating an optimal rule reasoning path according to the rule query request in combination with the formalized rule elements; aggregating all the optimal rule reasoning paths to form a trade rule corresponding to the rule query request; and establishing a mapping relationship between the semantic structure and the formalized representation of the rule text through a bidirectional mapping mechanism.

The invention provides an unmanned aerial vehicle risk handling question-answering system and method based on knowledge enhancement, and belongs to the technical field of computer data processing. According to the method, two-way analysis is performed on user emergency scene query, dynamic reconstruction is performed on the unmanned aerial vehicle risk knowledge graph based on the analysis result, a highly purified and structured disposal knowledge flow is provided for a large model, a scheme generated by the large model is guided to be a finally generated standardized disposal scheme, and the risk of the unmanned aerial vehicle risk knowledge graph can be quickly and efficiently handled. The accuracy and reliability of a processing scheme are improved, the core defect that a general large model easily generates factual illusion in the professional field is avoided, a macroscopic logic framework of answers is preset through question and answer structure planning, and a processing priority is marked for specific operation in processing knowledge flow; it is ensured that the generated standardized disposal scheme meets actual combat regulations and safety requirements of emergency disposal in the aspects of logic structure, step organization and operation sequence.

The invention provides a legal instrument writing guidance processing method and system based on a language model, and relates to the technical field of legal instrument writing, and the method comprises the steps: obtaining an evidence material, an interaction action sequence and pressure data, and constructing a real-time updated initial space-time diagram according to the evidence material; processing the initial space-time diagram by using a time sequence knowledge graph embedding model, and dynamically adjusting the semantic association degree between events through pressure data to obtain an enhanced space-time diagram; on the basis, event evolution under different vertical fields is simulated in parallel through anti-factual reasoning, so that a plurality of logic diagrams are generated; inputting the logic diagram and a preset strategy template library into a reinforcement learning decision framework, and performing multiple rounds of adversarial evaluation to obtain a target logic structure and a text construction strategy matched with the target site; and finally, generating a first text draft, and backtracking to generate guidance information. According to the method and the device, the disordered evidence can be constructed into the legal document which is strict in logic, adaptive in vertical field and high in persuasion.

Embodiments of the present specification provide a method, device and storage medium for detecting invocation of a trusted execution environment (TEE) by an Android application. The method comprises: obtaining a data packet of a software to be detected; obtaining an identifiable file from the data packet based on a reverse engineering algorithm; identifying whether the software to be detected invokes a TEE function based on a preset TEE function vector set and the identifiable file, the TEE function vector set containing vectors representing general functions / interfaces, indirect invocation functions / interfaces or specific TEE functions / interfaces; obtaining a function vector of the invoked TEE function, text data and logical structure data of a module in the software to be detected that invokes the TEE function from the identifiable file when the software to be detected invokes the TEE function; and determining a functional module to which the invoked TEE function belongs based on the function vector of the invoked TEE function, the text data and the logical structure data. The technical solution provided by the present application solves the problem of insufficient precision in detecting TEE in the prior art.

The invention relates to the technical field of databases, and discloses a data recovery method and device for a PostgreSQL database, which are executed on the premise of not depending on pre-stored backup of the PostgreSQL database, and comprises the following steps: scanning file system metadata of a storage partition where the database is located, positioning a physical storage area which is occupied by database data files, and storing the physical storage area in the storage partition; the range is used as a target range of subsequent scanning data fragments; within the target range, executing deep scanning for the PostgreSQL storage structure to extract an effective data fragment set; based on the effective data fragment set, reconstructing a logic structure definition of a deleted table; and analyzing and assembling the effective data fragment set according to the logic structure definition, recombining the effective data fragment set into a complete logic data row record, and exporting a data file in a preset format to complete data recovery of the database.

The specification discloses a file compliance judgment method and related devices, wherein the method introduces a multi-modal compliance large model to uniformly analyze and compare external rules files and internal rules files, integrates text, logical structure and image elements into the same analysis framework, and realizes systematic judgment on file compliance. By constructing a structured compliance rule set and a compliance knowledge graph, and combining a gap analysis large model to identify and quantitatively evaluate the gap, the compliance analysis is changed from traditional manual comparison to automatic, structured and quantifiable processing, which significantly improves the comprehensiveness, accuracy and efficiency of compliance review, and reduces the risk caused by human understanding bias.

The invention provides a task processing method and electronic equipment, and is applied to the technical field of artificial intelligence. The task processing method comprises the steps that a target task is decomposed into a plurality of sub-tasks, and the sub-tasks are different execution nodes of the target task; determining a sub-action sequence corresponding to the sub-task, wherein the sub-action sequence has at least one type of current semantic information, current logic structure information and current context information; according to at least one type of information of the sub-action sequences, at least one candidate action sequence corresponding to the target task is determined from a knowledge base, and the knowledge base comprises a plurality of historical tasks and historical action sequences corresponding to the historical tasks; processing the target task based on the large model, and generating a target action sequence corresponding to the target task by taking the at least one candidate action sequence as prompt information; the target action sequence is used for executing the target task.

An apparatus includes a central processing unit (CPU) coupled to a system-on-a-chip (SOC) interconnect. The apparatus also includes multiple logic structures coupled to a memory. The apparatus further includes a safety mechanism coupled to and inline with the memory, the logic structures, and the CPU via the SOC interconnect. The safety mechanism comprises a meta cache and is configured to detect errors in one or more of the logic structures and the memory.

The invention discloses a control system deployment method and device, equipment and a storage medium, and the method comprises the steps: determining at least one to-be-deployed physical equipment in a target control system according to a subsystem of the target control system, and determining a data communication protocol between different physical equipment according to the equipment type of the physical equipment, generating an equipment topological structure; determining at least one system function block corresponding to a subsystem based on system function requirements of different subsystems in the target control system, and generating a function block topological structure based on a function logic relationship between the system function blocks; and determining a system deployment model of the target control system based on the equipment topological structure and the functional block topological structure. According to the technical scheme, the system logic structure of the target control system is optimized, and the association degree between different subsystems is improved.

The application discloses a kind of intelligent chart generation method, system and storage medium suitable for multi-modal data, by receiving and pre-processing multi-modal data, complete user intent type identification and task routing;Again according to routing result, select adaptive model and build generation prompt word;The initial graph code output by model is standardized to intermediate representation format containing metadata field, and is converted into graph code compatible with target platform and retains node definition, connection relationship, layout constraint and metadata information;Through four optimization components, respectively complete chart type identification and layout strategy determination, node format size adjustment, overall layout alignment correction and partition adjustment, connection line identification optimization, and render output chart.It solves the problem that existing model is significantly poor in generating professional schematic chart requiring accurate logical structure and geometric constraint due to lack of special optimization and processing mechanism for structured chart.

The invention belongs to the technical field of visual large model training. The invention provides a rapid autonomous evolution method for a large visual model of a ubiquitous embedded terminal. According to the embodiment of the invention, the most important layer is preferentially selected for updating on the premise that the cost requirement is met, and the optimal balance between the precision and the time delay can be realized. In order to evaluate layer time delay, in the reasoning process of the deep learning model, the calculation process relates to a plurality of layers in a large model, and input and output tensors with similar calculation sequences and logic structures are processed when each layer is executed. Through inductive analysis of forward propagation, reverse propagation and weight update delays of different layers, a model is established, different delay types are abstracted, and an accurate delay evaluation basis is provided for scheduler design. The intermediate activation result only depends on the input data and the output of the previous layer, and the same activation result is generated by the same input data, so that repeated calculation on a non-updated layer is avoided by adopting a memory-aware calculation multiplexing method, and redundant calculation is remarkably reduced.