338 results about "Information extraction" patented technology

The present application is suitable for the field of tower foot welding, and discloses a tower foot weld information extraction method, device, equipment and storage medium, the method comprising: processing the obtained tower foot overhead point cloud data, obtaining the main shoe plate thickness, the secondary shoe plate thickness, the bottom plate size and the vertical weld starting point coordinates of the first quadrant; processing the obtained tower foot side view point cloud data, and combining the main shoe plate thickness, the secondary shoe plate thickness and the bottom plate size to obtain the vertical weld end point coordinates, the horizontal weld end point coordinates, the first horizontal weld starting point coordinates, the second horizontal weld starting point coordinates and the inter-plate angle of the first quadrant; determining the first horizontal weld starting point coordinates, the second horizontal weld starting point coordinates, the vertical weld starting point coordinates and the inter-plate angle of the second quadrant, the third quadrant and the fourth quadrant according to the foregoing information, the method can accurately extract the size information and the weld information of the tower foot through machine vision and point cloud data processing technology.

This invention belongs to the field of natural language processing and information extraction technology, and provides a method and system for extracting parameters from boiler design documents based on multi-round self-feedback RAG. The method includes: preprocessing the boiler design document and the original query, converting them into several text segments and expected inputs respectively; encoding the text segments and expected inputs into high-dimensional vectors and query vectors respectively, storing the high-dimensional vectors and constructing a vector index; generating a response based on retrieved external knowledge to assist a large language model, and iteratively optimizing the generated results through multiple rounds. This invention effectively solves the problems of incorrect and missing parameter extraction in complex or fuzzy query scenarios, significantly improving the reliability of the final results; it overcomes the limitations of traditional methods and standard large language models in handling multimodal data, enhancing the processing capability for complex multimodal documents; and it improves the accuracy of information retrieval, laying a solid foundation for subsequent accurate parameter extraction.

The invention relates to a system (100) for generating real-time causal reasoning insights in real time. The system (100) comprises a global causal graph creation module (101), a natural language query processing module (102), an information extraction module (103), a query-centric subgraph generation module (104), and an adaptive causal reasoning engine (105). The queries from users are processed to identify relevant causal relationships in the global causal graph. The adaptive causal reasoning engine (105) refines the insights iteratively based on user feedback, improving accuracy. The system (100) converts technical insights into structured narratives. A visualization module presents causal relationships through graphs and charts. The feedback mechanism allows users to correct errors and adjust outputs, facilitating user-driven data analysis and causal reasoning, delivering actionable insights for a variety of applications.

The application discloses an electric power information extraction method, device and computer equipment. The method comprises the following steps: recognizing key characters in an industry expansion document by using a multi-modal large model, and performing annotation processing on the key characters; incrementally training an expert model based on the annotated key characters; in response to an information extraction request of a target document uploaded by a user, inputting the information extraction request into the multi-modal large model to obtain an extraction task; calling an expert model corresponding to the extraction task to process the target document to obtain a text extraction result of the target document. The method can improve the intention understanding ability of complex and fuzzy requirements, help improve the accuracy, robustness and response speed of document information extraction for complex and variable business requirements, and greatly improve the automation degree and efficiency of business processing while ensuring high accuracy.

The application provides a multi-time-phase laser point cloud and three-dimensional real scene model registration method, and belongs to the technical field of point cloud registration, and comprises the following steps: sequentially sorting and automatically registering multi-period laser radar point cloud data in a long time period, and calculating point cloud density values. The initial vertex point cloud of the three-dimensional real scene model is extracted, three-dimensional grid creation is performed according to the calculated density values, the curvature adaptive sampling method and the grid plane interpolation method are adopted to expand the point cloud density, and density-enhanced point cloud data is obtained. Feature point information extraction and matching are performed on the point cloud data obtained by automatically registering the multi-time-phase point cloud and the density-enhanced point cloud data, and the automatic registration of the two types of point clouds is realized. By adjusting the calculation model point cloud to meet the target density and be uniformly distributed, the problem of registration difficulty caused by the large density difference between the three-dimensional real scene model point cloud and the multi-time-phase point cloud is effectively solved.

The application discloses a parking space detection method and device, electronic equipment and storage medium, and relates to the field of intelligent parking. The method comprises the following steps: performing feature extraction on a target feature image of a to-be-detected parking space image; performing parking space entrance information extraction based on the target feature image to obtain target parking space entrance information; the target parking space entrance information comprises a parking space entrance key point heat map, a parking space entrance center point heat map, parking space entrance direction information and position offset information; determining a target parking space entrance key point corresponding to each parking space entrance according to the parking space entrance key point heat map, the parking space entrance center point heat map and the position offset information; and determining a target parking space based on the target parking space entrance key point and the corresponding direction information. The application determines and corrects the parking space entrance key point by using the parking space entrance center point and the position offset information, and adds the parking space entrance direction information, so that the parking space detection result is more accurate.

The present application relates to the field of dam inspection, especially to a kind of image processing method and system for dam inspection, the present application is by controlling unmanned aerial vehicle to fly over detection area according to predetermined trajectory, when unmanned aerial vehicle is pre-entered detection area, continuously obtain image information and carry out multidimensional analysis, based on the result of multidimensional analysis and mirror surface reflection index, calculate to obtain spectral characteristic interference representation value, set interference label for detection area, subsequent interference label based on the control of unmanned aerial vehicle in detection area adaptively obtains spectral information in monitoring area, subsequent generation perspective adjustment instruction, by sampling determination selection confidence sampling position obtains spectral information, the present application is by image information in advance sensing interference situation of spectral information extraction in detection area, adaptively select spectral information capture mode, under the premise of guaranteeing inspection efficiency, improve the quality and reliability of the obtained spectral information, provide effective data support for subsequent analysis dam potential anomaly.

Provided in the present application are a multi-agent reasoning system, a data reasoning method, apparatus and device, and a product. The method is applied to a multi-agent reasoning system comprising a plurality of agents, and comprises: a first agent performing key information extraction on reasoning reference data on the basis of a reasoning problem, so as to obtain reasoning key information (S101); a second agent generating a reasoning path on the basis of the reasoning problem and the reasoning key information (S102), wherein the reasoning path comprises prompt information for performing reasoning on the basis of the reasoning key information, so as to solve the reasoning problem; and a third agent performing integrated reasoning on the basis of the reasoning problem and the reasoning path, so as to generate a reasoning result corresponding to the reasoning problem (S103), wherein the reasoning result comprises a reasoning process and / or a reasoning conclusion. The method can improve the flexibility and reasoning efficiency of data reasoning.

The application relates to the technical field of artificial intelligence recommendation, and specifically discloses a medical cooperation partner intelligent matching method and system based on multi-source data fusion, which collects multi-source business data such as telephone recording, instant messaging text and documents in business communication, constructs a multi-dimensional feature vector of cooperation partners through voice recognition, text analysis and information extraction, carries out vector similarity calculation and basic matching degree comparison on the features of cooperation partners and business project features, forms a candidate set, adopts multi-dimensional weighted scoring and normalization processing to generate a precise recommendation list, and dynamically adjusts the model weight according to the operation feedback of business personnel to realize online optimization. The application can effectively integrate unstructured communication data, improve the mining depth and matching accuracy of cooperation partners, reduce the artificial screening cost and compliance risk, form a self-learning intelligent recommendation closed loop, and significantly improve the efficiency of medical academic popularization and business promotion and the rationality of resource allocation.

The application provides a data processing method and system based on a multi-modal large model, relates to the technical field of data processing, and comprises the following steps: analyzing and identifying to-be-inspected data, and determining corresponding engineering types and inspection strategies; sending a query request to a pre-trained knowledge graph and a pre-trained vector knowledge base based on the engineering types and the inspection strategies, obtaining path retrieval results associated with structured knowledge and semantic retrieval results associated with unstructured knowledge, and fusing the path retrieval results and the semantic retrieval results to obtain fused retrieval results; the pre-trained knowledge graph comprises entities, attributes and the association relationships between the entities; the pre-trained vector knowledge base comprises a plurality of professional material text coding blocks; the to-be-inspected data is input into a pre-set inspection model for preprocessing and fine-grained element recognition, global macroscopic description, key path calculation and key information extraction are completed, reasoning is performed in combination with the fused retrieval results, and a structured inspection result is generated.

This application discloses a method, apparatus, device, and storage medium for motion redirection, relating to the field of computer technology. The method includes: acquiring a motion capture sequence to be processed, the motion capture sequence including an upper limb motion sequence and a lower limb motion sequence; obtaining human arm span information based on a preset parametric human model and the upper limb motion sequence; converting the upper limb motion sequence based on preset arm span parameters of a target robot and the human arm span information to obtain an upper limb motion sequence of the target robot; obtaining an upper limb joint angle sequence of the target robot using an inverse kinematics algorithm based on the upper limb motion sequence of the target robot; performing information extraction processing on the upper limb joint angle sequence of the target robot to obtain three-dimensional spatial pose information corresponding to the upper limb joint angle sequence of the target robot; and determining the motion data of the target robot to be redirected based on the upper limb joint angle sequence and the three-dimensional spatial pose information of the target robot.

The application discloses a power marketing text-oriented information extraction method and device, terminal equipment and storage medium, and belongs to the technical field of data processing. The method comprises the following steps: based on a preset power text example library, the similarity of a to-be-extracted power marketing text is reordered to obtain a target example; the obtained information extraction instruction and the to-be-extracted power marketing text are input into a preset information extraction model; and information extraction results are extracted according to the information extraction task and output format constraint of the information extraction instruction. The training of the information extraction model comprises the following steps: an initial training set is generated by marking historical marketing texts according to historical extraction results; the similarity between the historical marketing texts is calculated; samples with a similarity not greater than a preset similarity threshold value in the initial training set are taken as a current training set; the current training set is input into a to-be-trained information extraction model to repeatedly perform an optimization operation until a preset loss function value converges to obtain the information extraction model. The application solves the problem of low extraction accuracy in the prior art.

This invention discloses a method for outputting detailed OpenFOAM boundary surface mesh information. This method directly extracts mesh information from specified physical boundary surfaces based on OpenFOAM numerical simulation results. It automatically obtains geometric information such as coordinates, area, and normal vectors of all user-interested facets from a large number of complex topologically connected mesh cells; topological information such as the index number and mesh number of all facets; and physical field data information on all facets. The results are written to one or more data files in a format easily readable and visualized by common data post-processing software. This invention, employing the aforementioned method for outputting detailed OpenFOAM boundary surface mesh information, can flexibly adapt to various meshes, automates information extraction and output, reduces the complexity of post-processing operations, saves time and costs, supports subsequent boundary surface flux calculations, visualization analysis, and other tasks, and improves the efficiency and accuracy of post-processing for complex numerical simulations.

This invention discloses a method for mining alumni information based on a large-scale model, comprising the following steps: 1) acquiring raw alumni-related data; 2) preprocessing the acquired data; 3) performing natural language processing on the acquired data; 4) semantic parsing and information extraction using a large-scale model; 5) constructing a knowledge graph; storing the extracted alumni information and relationships in a graph database; and constructing a visualized alumni knowledge graph; 6) updating the knowledge graph data; and 7) providing intelligent retrieval and question-answering services for alumni information based on the knowledge graph data. This invention uses a large-scale model to perform deep semantic understanding and information extraction from unstructured text. Compared to traditional keyword-based or small-scale model-based methods, it can more accurately identify alumni identities, professional experiences, and social relationships, while reducing the false matching rate.

A method for auto-correcting information extracted by an open information extraction system includes inputting an input triple extracted from an input sentence to a grammaticality model trained to output whether the input triple is grammatical or ungrammatical. The input triple is corrected based on the grammaticality model outputting that the input triple is ungrammatical. Embodiments of the present invention can be used in a variety of machine learning and artificial intelligence applications, including but not limited to several anticipated use cases in material informatics, data security, data extraction, and medical / healthcare, for example, for supporting decision making or optimizing and calibrating of extracted chemical compounds from libraries of textbooks and publications.

The application discloses a kind of view extraction method, electronic equipment and computer readable storage medium.Therein, method includes obtaining the extraction result after the information extraction model in the view extraction device is handled to positive sample training data;According to the extraction result, generate the training negative sample of the aspect word classification model, the sentiment classification model and the category and sentiment pairing identification model in the view extraction device;According to the positive sample training data and the generated training negative sample, the aspect word classification model, the sentiment classification model and the category and sentiment pairing identification model in the view extraction device are trained;Utilize the view extraction device after training to the text to be extracted and process, obtain the view information in the text to be extracted.The scheme provided in the application can improve the accuracy of view extraction.

The application provides an API information processing method and an electronic device, wherein the method comprises: fusing abstract syntax tree analysis, large language model and retrieval enhancement generation technology, obtaining code fragments and a vector library by analyzing target application source code, extracting a route group and constructing a route mounting tree representing hierarchical relationships by using a large language model, determining complete target API information in combination with the route mounting tree and the extracted initial API information, and finally generating a function description through semantic retrieval of the vector library to associate related code fragments. The API information extraction and description generation full process are automated, the API information integrity and the accuracy and richness of the function description are improved, the manual maintenance burden is effectively reduced, and the application scenarios and business change scenarios lacking API document maintenance are adapted.

The application discloses a table information extraction method, system and device based on a compressed pre-training language model. The application models the table information extraction as a reading comprehension problem and inputs the pre-training language model, thereby avoiding error propagation and reducing the participation of rules. Meanwhile, the strong generalization ability and robustness of the pre-training language model can obtain relatively accurate cells in the table that can answer natural language questions, thereby completing the extraction of the associated information in the table.

This application discloses a method, apparatus, electronic device, and readable storage medium for entering questions. The method includes: acquiring a currently captured preview frame image, the preview frame image containing question information; determining whether a historical reference frame image matching the preview frame image is stored, the historical reference frame image being associated with corresponding historical entry information, the historical entry information being obtained based on feature information extraction processing of the historical reference frame image; and, if it is determined that a historical reference frame image matching the preview frame image is stored, entering questions based on the historical entry information. This technical solution can avoid redundant capture and detection operations, greatly reducing computational overhead.

This application discloses a method, method, and apparatus for training an information extraction model. The main technical solution includes: acquiring a third dataset, which includes multiple unlabeled third text samples; using a pre-trained first information extraction model to extract target information of a preset type from the third text samples in the third dataset, as a second target sample; inputting the second target sample into a pre-trained text generation model to obtain text generated by the text generation model using the second target sample as a second text sample; forming a second sample pair by combining the second target sample and its corresponding second text sample; obtaining a second dataset using multiple second sample pairs; and training a second information extraction model using the second dataset. This application can improve the information extraction effect of the information extraction model while reducing manual costs and time consumption.

This disclosure relates to an information extraction method, apparatus, electronic device, storage medium, and program product, belonging to the field of natural language processing technology. The information extraction method includes: determining a commonly used text segment corresponding to a first element; wherein the number of times the extraction result corresponding to the first element appears in the commonly used text segment satisfies a frequency threshold; determining a first text segment from multiple text segments of the text based on the commonly used text segment; the similarity between the first text segment and the commonly used text segment satisfies a similarity condition; determining prompt information based on the first element and the first text segment; and performing information extraction on the text based on the prompt information to obtain the extraction result of the first element. This disclosure improves the accuracy of information extraction by accurately selecting the first text segment containing the element and then obtaining prompt information based on the first text segment containing the element for information extraction.

This invention discloses a joint estimation method and system for state and characteristic parameters based on maneuver intensity perception. It primarily addresses the problems of filter divergence and low parameter estimation accuracy caused by sudden changes in target maneuvering when estimating parameters for hypersonic targets in existing technologies. The implementation scheme is as follows: receiving observation data and calculating innovation; introducing a forgetting factor to adaptively update the process noise covariance, and using the innovation norm to construct a gain factor to adjust the observation weights, completing the posterior state update; calculating the sliding window length based on the adaptive process noise covariance, and constructing the arrival cost using the output of the posterior state update; transforming physical constraints into a penalty function, and combining the arrival cost to establish a rolling time-domain estimation objective function and solve for the optimal estimate. This invention can perceive maneuver intensity in real time, dynamically adjust the estimation window and observation weights, significantly improving the accuracy and robustness of the joint estimation, and can be applied to guidance information extraction and trajectory prediction for near-space hypersonic gliders.

The present application relates to the technical field of bid evaluation system, in particular to a bid document structured information extraction and intelligent bid evaluation integrated system, which comprises a rule analysis module, a constraint calculation module, a field construction module, an element integration module and a consistency verification module.In the present application, semantic boundary recognition and sequence annotation are performed on the scoring clause text to make the scoring condition expression structured and analyzable, a constraint parameter strength analysis is introduced to form a weight sequence, different conditions are presented with a distinguishing degree in the review, the bid document field comparison is focused on high correlation content, and the element correlation relationship is constructed in combination with the paragraph and chapter position information, the consistency and conflict review is carried out at the bid evaluation element level, the redundant field influence is compressed, the scoring basis integrity and logical coherence are enhanced, the review process expression is clearer, and the result formation path is more checkable and stable.

This information processing device comprises: a storage means that stores, in a storage unit as document information in association with an item to be extracted, an item name and item content corresponding to the item to be extracted, the item name and item content having been extracted from a document processed in the past on the basis of the item to be extracted, and position information indicating each position of the item name and the item content; a retrieval means that retrieves, from among the document information stored in the storage unit on the basis of a character string included in a document to be processed and position information indicating a position of the character string, the document information of a document processed in the past and corresponding to the document to be processed; and an information extraction means that inputs, to a large language model, a prompt including the retrieved document information to perform inference by the large language model, and extracts, from the document to be processed, an item name and item content corresponding to the item to be extracted.

This invention discloses an image semantic segmentation processing method with adaptive context information extraction, belonging to the field of image processing technology. The specific steps of the method are as follows: First, basic features in three domains—spatial, frequency, and semantic—are collected and regularized into an original set. Then, cross-domain mapping is used to generate globally interconnected and fused features. Next, based on regional attributes, the features are decomposed into low-frequency structures and high-frequency detail components, and their proportions are adjusted to extract complete contextual features. Finally, after consistency verification and optimization of the three-domain information, pixel category determination is completed, and a high-precision semantic segmentation result is output. This invention solves the core technical problems of insufficient information coordination, difficulty in balancing global structure and local details, and inability to correct information conflicts between dimensions in traditional methods through a full-chain adaptive information processing mechanism. This improves the accuracy and regional clarity of semantic segmentation, making the segmentation results more consistent with actual scene requirements.

The application provides an industrial robot unpacking and feeding method supporting natural language instructions, and the method comprises the following steps: S1, receiving the unstructured natural language demand information about the unpacking and feeding task input by a user through a natural language interaction interface, wherein the demand information comprises at least one of the material type, the packaging specification, the feeding rate and the production line environment layout; S2, performing semantic analysis on the natural language demand information by using a large language model fine-tuned by professional knowledge in the field of robot unpacking and feeding, and performing intention recognition and key information extraction; if it is identified that the demand information is missing or ambiguous, generating natural language prompt information for follow-up questioning and returning to the user through the interaction interface, and repeating the step until a complete and clear task parameter set is obtained.

The application discloses a vehicle key information extraction method, device and system and electronic equipment, and relates to the technical field of vehicles. The method comprises the following steps: acquiring video data and CAN data of a vehicle; for each video frame in the video data, if a time difference between the video frame and a starting time of the video data is greater than a preset time, a first reference frame is selected from the video data according to a first time corresponding to the video frame, and a pixel difference value between the first reference frame and the video frame is obtained; a key frame is selected from the video data according to the pixel difference value and the CAN data; and a key information extraction result is generated according to the key frame. Thus, the vehicle video data, including a vehicle video without audio text information, can be subjected to key information extraction.

This invention relates to the field of intelligent document processing technology, specifically to a method and system for document processing in power transmission and transformation scenarios based on a brain-like memory mechanism. The method includes the following steps: generating a query plan based on query intent; recalling relevant documents from a hierarchically modeled power transmission and transformation document knowledge base; and extracting and reasoning information from the recalled documents using a dual-path parallel processing mechanism and a hierarchical abstraction processing mechanism; filling in a structured form according to a predefined template to generate preliminary structured output; simultaneously transforming the extracted and reasoned information into interpretable reasoning chain records; generating memory cases based on the reasoning chain records and confidence assessment results; and storing and optimizing the memory cases using a brain-like memory mechanism. This method possesses the ability to handle complex multi-step reasoning tasks, significantly improves reasoning accuracy, automates document information extraction, and greatly reduces manual processing time.

The application discloses a kind of generation methods, devices and electronic equipment of component production process file, wherein the method comprises: obtaining the design drawing file of the component to be produced;Using machine vision model, visual information extraction is carried out on design drawing file, and the first component characteristic information in design drawing file is obtained;Based on the first component characteristic information, the second component characteristic information in the initial production process file is proofread, and the proofreading result is obtained;Based on the proofreading result, the initial production process file is adjusted, and the target production process file is obtained, wherein the target production process file is used to indicate the file of the production process of the component to be produced.The present application solves the technical problem that the accuracy of generating component production process file in related art is low.