3651 results about "Data type" patented technology

The invention discloses an intelligent customer service system based on an AI large model, and belongs to the technical field of intelligent customer service systems. Through the multi-modal interaction technology, the flexibility of the system and the user experience are improved, the data type is intelligently recognized and correspondingly processed, the language style and emotional intonation of the response content are automatically adjusted according to the mood and intention of the user, the knowledge base integrates static data and dynamic interaction data, and the user experience is improved. Real-time updating and comprehensiveness of knowledge content are ensured, rapid retrieval is performed based on a knowledge base, accurate reply is performed in combination with a large language model, high-quality answers and solutions are generated through rule reasoning and model generation, state information of each round of interaction of a user is recorded and managed by constructing a dialogue state tracker, and the user experience is improved. And precise tracking and updating of the context are realized. For a complex task, the system splits a user request into a plurality of sub-tasks, and the sub-tasks are executed in sequence according to the priority and the dependency relationship, so that the high efficiency and continuity of task processing are ensured.

The invention relates to the technical field of archive information query, and discloses an AI-based archive information resource intelligent sharing method and system, and the method comprises the steps: firstly constructing a multi-dimensional data module according to a business data type, and the multi-dimensional data module comprises an archive system database, a high-dimensional document vector database and an archive business knowledge graph; and performing deep analysis on user query through a natural language processing tool to form a structured keyword list. Then, performing semantic analysis on the keyword list by utilizing a pre-trained archive information resource understanding large model, and generating an enhanced query text; the query text is recalled from the data modules of the multiple dimensions, candidate answers are obtained, and the candidate answers are fused and optimized through the general large language model. And finally, returning the optimal answer to the user. By introducing the AI technology, the accuracy and comprehensiveness of file information resource sharing are improved, the retrieval result is optimized, and the retrieval efficiency is improved.

A system and methods for processing diverse financial data types using a jointly trained Vector-quantized Variational Autoencoder (VQ-VAE) and neural upsampler. This system efficiently handles time-series, textual, sentiment, and structured tabular data through specialized encoding modules. A novel fusion module integrates these encodings, capturing cross-modal relationships via attention mechanisms and gated fusion units. The fused representation is compressed into a discrete latent space by the VQ-VAE encoder, then reconstructed and enhanced by the VQ-VAE decoder and neural upsampler, respectively. Joint training optimizes all components simultaneously, using a comprehensive loss function that balances reconstruction quality across modalities with upsampling performance. This approach enables superior data compression, reconstruction, and analysis, leveraging inter-modal correlations to improve financial forecasting, risk assessment, and decision-making. The system's ability to explore the latent space facilitates generation of new, synthetic financial scenarios for robust model testing and strategy development.

The invention provides a science and technology finance platform data sharing method, which comprises the following steps: extracting key field information and a metadata label from a data stream of a science and technology finance platform, the key field information comprising data content and a source identifier; analyzing a data type characteristic value according to the metadata label, and determining a data type and a corresponding sensitivity level by adopting a logistic regression model; querying a source credibility rule base based on the source identifier, and determining a risk level of the data source; according to the sensitivity level, the risk level and the receiver permission identifier, calculating to obtain a dynamic sensitivity score and generating a dynamic sensitivity tag; according to the risk level mapping table, in combination with the sensitivity level, determining a final desensitization mode of the field; and performing format verification on the processed data stream, transmitting the processed data stream, and recording execution information of the dynamic sensitivity label and the desensitization strategy.

The invention discloses a fusion method for automatic coprocessing of multi-source heterogeneous data, which comprises the following steps of: analyzing the format and semanteme of the multi-source heterogeneous data to obtain metadata features of each data source, the metadata features comprising data types, coding modes and semantic tags, and generating a standardized metadata description set; according to the metadata description set, a pre-established protocol template library is adopted, an applicable template is matched, an initial unified exchange protocol is generated, and the initial unified exchange protocol comprises a format conversion rule and a semantic mapping relation; and if the format conversion rule of the initial unified exchange protocol cannot adapt to the metadata features of the newly added data source, dynamically updating the protocol template through a machine learning algorithm to obtain an exchange protocol adaptive to the new data source.

The invention relates to the technical field of natural language processing, in particular to a semantic synthesis and text vectorization-based data lake metadata governance method, which comprises the following steps of: performing vector coding of field content keywords and adjacent context words on the basis of field names, data types and description information of data sources in a data lake; and extracting semantic co-occurrence groups of the fields in the differentiated contexts, identifying distances between semantic vectors, and judging whether the distances are in a similar field grouping range or not to obtain a field semantic collection quantity. According to the method, a semantic co-occurrence relation is extracted through field keywords and context vector coding, a semantic hierarchical structure is constructed, the field division precision and abstract ability are improved, the consistency is judged in combination with a context semantic structure, a high-frequency alternative path is adjusted to optimize a semantic structure, and word vector similarity and a structure retention rate are fused to realize field merging; the continuity and the accuracy of a treatment structure are improved, and the intelligence and the consistency of metadata treatment in the data lake are enhanced.

The invention discloses an electric power field knowledge question-answering system construction method based on a large language model, and relates to the field of electric power field knowledge question-answering, and the method comprises the steps: judging the data type of electric power field knowledge, and carrying out the processing of the electric power field knowledge according to the judgment result through matching with a processing technology, and generating an entity relation triple; constructing a power field knowledge graph; optimizing the power field knowledge graph based on the attention network, outputting an answer causal path of the fault problem by using the optimized power field knowledge graph, and marking a confidence score of the answer causal path; and inputting the solution causal path and the confidence score into a language model to obtain a fault question answering result, and optimizing the question answering result according to the consistency of the fault question answering result and the power field knowledge graph. According to the method, on the premise that the fault diagnosis logic is rigorous and the result is traceable, knowledge in large-scale unstructured literatures in the power industry is activated, so that accurate question and answer services can be provided for operation and maintenance personnel in real time.

A system and method for processing and generating multimodal data using a combination of Vector Quantized Variational Autoencoder (VQ-VAE) and Latent Transformer architectures. The system efficiently handles diverse data types including time-series, textual, sentiment, and structured tabular data through specialized encoding modules. A novel fusion module integrates these encodings, capturing cross-modal relationships. The fused representation is compressed into a discrete latent space, processed by a latent transformer, and then reconstructed and enhanced. This approach enables superior data compression, reconstruction, analysis, and generation of synthetic scenarios.

The invention discloses an intelligent data analysis method and system based on an industry large model, and relates to the technical field of data analysis, and the method comprises the steps: collecting business event data, carrying out the feature vector extraction according to the data type, calculating a zoom dot product attention score matrix through employing a self-attention mechanism, and carrying out the weighted output of a multi-modal fusion vector, constructing an industry knowledge database to determine an embedded vector; and calculating a relevancy value, screening a combination of the multi-modal fusion vector and the knowledge fragment, and forming a gating input vector to carry out fusion vector calculation. According to the method, unified expression and correlation mining of heterogeneous information such as structured data, text information and images can be realized through a multi-modal feature fusion and knowledge fragment retrieval scheme, the attention degree of each modal feature can be adaptively adjusted according to a service scene through a self-attention mechanism and a model, information complementation is realized, and the accuracy of information retrieval is improved. And automatically highlighting the multi-modal dimension most related to the current task.

The invention discloses a data quality intelligent auditing system and method based on a dynamic rule base, and belongs to the technical field of data auditing, and the system comprises a rule base construction module which is used for analyzing business scene parameters through a scene analysis unit according to business scene demands and data type features to generate a rule configuration instruction; the multi-source monitoring engine module is connected to the rule base construction module and is used for collecting multi-source data in real time and loading corresponding checking rules; the automatic verification execution module is used for executing normalized quality verification on the multi-source data based on the verification rule base; and the feedback optimization module analyzes a rule hit rate and a false alarm rate in a verification result through a reinforcement learning algorithm, and dynamically iteratively updates a rule threshold value and a logic combination in the rule base. By constructing a full-automatic process of rule generation, execution, feedback and updating, the problems that a traditional system depends on manual intervention, response is slow, the industry average rule updating period is 3-7 days, and real-time updating is achieved through the scheme are solved.

The invention relates to the technical field of data fusion, and discloses a multi-source heterogeneous data fusion method and system based on edge computing, and the method comprises the steps: obtaining a heterogeneous data stream, carrying out the data type recognition and data feature extraction, and obtaining an original feature set; according to the original feature set, unifying feature dimensions and adjusting a time reference to obtain time sequence vector data; according to the time sequence vector data, filling the feature value of the missing time point to obtain a multi-modal feature; performing block storage on the multi-modal features, verifying the synchronism of adjacent modals, distributing modal synchronization weight coefficients, and finally generating a fusion feature matrix; according to the fused feature matrix, identifying and filtering redundant feature dimensions, establishing a feature association map, and executing feature merging to obtain a simplified feature matrix; according to the simplified feature matrix, feature importance scores are calculated, sorting weight coefficients are arranged and distributed in a descending order according to the scores, and a multi-modal fusion semantic vector is generated. The method improves the accuracy of data analysis and decision.

The invention relates to the technical field of artificial intelligence, can be applied to business scenes of financial science and technology, medical health and the like, and discloses a machine learning task execution and feedback method, device, equipment and medium. And performing semantic reasoning by using a pre-training language model, generating a decision result in combination with an entity relationship associated with a task type and a data type in the knowledge graph, executing a corresponding machine learning task, converting a task execution process and an execution result into natural language feedback information, and outputting the natural language feedback information to a user side. According to the method, natural language input is structured into machine learning task description, large model semantic reasoning and knowledge graph knowledge association are combined, an executable decision result is generated, automatic closed loop from task recognition, model selection to result feedback is achieved, the user operation threshold is lowered, and the model development efficiency and the interaction intelligence level are improved.

A non-transitory, processor-readable medium storing instructions that when executed by a processor, cause the processor to receive data artifacts, encode the artifacts to a standard data type, and compute, for each artifact, a hash function. The hash functions and encoded documents are stored in a first database. The processor is caused to tokenize the encoded artifacts, to produce tokens associated with natural-language identifiers extracted from the encoded artifacts. The processor is caused to transform, using an embedding model, the tokens to produce vectors that are stored in a second database and classified based on categories. The second database is configured to be queried to perform a semantic search in response to receiving a request from a user operating a user compute device. The processor is caused to retrieve, from the semantic search, a subset of vectors from the second database to be displayed on the user compute device.

A system and methods for upsampling compressed data using a jointly trained Vector Quantized Variational Autoencoder (VQ-VAE) and neural upsampler. The system compresses input data into a discrete latent space using a VQ-VAE encoder, reconstructs the data using a VQ-VAE decoder, and enhances the reconstructed data using a neural upsampler. The VQ-VAE and neural upsampler are jointly trained using a combined loss function, enabling end-to-end optimization. The system allows for efficient compression and high-quality reconstruction of various data types, including financial time-series, images, audio, video, sensor data, and text. The learned discrete latent space can be explored and manipulated using techniques such as interpolation, extrapolation, and vector arithmetic to generate new or modified data samples. The system finds applications in data storage, transmission, analysis, and generation across multiple domains.

Various embodiments of the technology described herein cause an LLM to intelligently process data based on a user query and a schema determined for a data set. Certain embodiments programmatically leverage an LLM and utilize its output based on a user query. In this manner, data is processed without the LLM having to access an entire data set, and instead only utilizes information associated with the user query and the schema. The schema comprises a textual description, such as a string of alphanumeric characters, that describes the data, data types, and / or data structure of the data set. Embodiments of the technology described herein are performed by an LLM interface layer separate from a user device layer and an LLM layer. The LLM interface layer is positioned between an LLM abstraction layer and an application layer by which a user can interface with the LLM interface layer.

The invention provides a deep and large foundation pit monitoring method and system, a program product and a storage medium, and relates to the technical field of foundation pit detection. The GPS positioning device collects foundation pit slope crest displacement data, the monitoring sensor network collects deep horizontal displacement data, the two data types complement each other, comprehensive information of surface and internal deformation of the foundation pit is provided, and the generated deep horizontal displacement curve chart visually displays the displacement conditions of the supporting structure at all depths. A potential weak area can be identified, a foundation pit excavation response data acquisition device is arranged at the maximum stress part, response data of a key position is obtained, and the quantitative index of overall deformation is provided by measuring and monitoring the position change of a central point of a sensor network and calculating the deformation amount of the foundation pit. According to the multi-dimensional and multi-type data acquisition and integration method, the description comprehensiveness of the foundation pit environment is improved, potential risks can be judged more accurately, and limitation caused by a single data type is overcome.

The invention discloses a mineral resource intelligent prediction method and system based on multi-source heterogeneous data fusion and deep learning, and the method comprises the steps: collecting and preprocessing multi-source heterogeneous data, and carrying out the standardization processing to form a structured data set; multi-source heterogeneous data fusion: realizing data layer space registration and feature layer weight dynamic allocation through an attention mechanism multi-modal fusion module, and outputting a high-dimensional metallogenic feature vector; constructing a CNN-LSTM mixed deep learning model and completing initialization training, and outputting an initial mineralization probability graph; and establishing a dynamic updating engine, performing model increment training based on transfer learning, correcting the mineralization probability through positive and negative sample reinforcement learning in combination with a newly added data type, and outputting a time sequence dynamic mineralization probability graph. According to the method, mineralization probability dynamic evaluation and risk quantitative updating are realized, the prediction precision and the model updating efficiency are improved, the method is adaptive to a multi-stage exploration scene, and accurate real-time support is provided for exploration decision making.

The invention discloses a method for simulating and forecasting flood in a cold highland area based on hydrological and hydrodynamic coupling, and belongs to the technical field of disaster forecasting. The method specifically comprises the following steps: S1, multi-source basic data collection and preprocessing: collecting multi-type and multi-scale basic data for a target cold and cold mountainous area drainage basin; and S2, deep learning correction and fusion of the satellite rainfall data: aiming at the local overestimation and underestimation problems of the satellite rainfall data, a deep learning algorithm is adopted to carry out hour scale correction and fusion. Four types of core data of satellite remote sensing, reanalysis, ground observation and geographic space are collected, total factors of'rainfall-runoff-terrain-underlying surface 'required by flood simulation in the cold and cold mountainous area are covered, simulation one-sidedness caused by lack of data types in traditional modeling is avoided, rainfall and runoff abnormal values are eliminated by adopting a 3-sigma criterion, data formats and spatial-temporal scales are unified, and the modeling efficiency is improved. A standardized data set is formed, and interference of abnormal values, format incompatibility and space-time mismatching on subsequent model input is avoided.

The invention provides a multi-modal multi-scale feature fusion-based power load prediction method and system, and relates to the technical field of power load prediction. According to the method, a cross-modal attention mechanism is adopted to fuse multi-scale load time sequence features and satellite image features; decoding the fusion features through a multi-step prediction decoder to obtain a multi-step power load prediction result; the acquisition mode of the satellite image features comprises the following steps: performing feature extraction on a satellite image based on a GRVT model to obtain the satellite image features; the GRVT model includes a number of stacked gated residual encoder network elements. According to the method, satellite image feature extraction is carried out based on the GRVT model, the global information processing advantage of Transform is utilized, the single-layer information retaining capability of a residual connection mechanism is also utilized, the two are combined to efficiently extract satellite image features, the satellite image features are fused during power load prediction, and the situation that the data type is single, and the prediction efficiency is high is avoided. And the precision of power load prediction is improved.

The invention discloses a virtual power plant energy scheduling method and system under a local data abnormal condition, and the method comprises the steps: recognizing a data type and an incidence relation related to abnormal data through an incidence matrix after the abnormal data of a virtual power plant are collected and recognized; and calculating an abnormal data correction value based on the association relationship function and the real-time association data, and performing weighted average on the abnormal data correction value and the original abnormal data to obtain correction data. And then calculating an abnormal data time difference and a correlation deviation degree, and distributing a credibility weight for the corrected data. And finally, calculating the correlation degree between the scheduling parameter of the preset scheduling scheme and the weighted correction value, and selecting the scheme with the highest correlation degree as a virtual power plant resource scheduling scheme. By implementing the technical scheme provided by the invention, the scheduling accuracy in a data exception state is improved.

The invention discloses a large language model acceleration system and method based on binary quantization, and the system comprises an extraction and quantization unit, a grouping packaging and splicing unit, an operation realization unit, and an inverse quantization and output unit, and comprises the specific steps: extracting parameters needing matrix / vector multiplication in a model; performing binary quantization processing on parameters of the model matrix / vector multiplication operation; splicing the grouped binary quantized parameters into single integer data bit by bit according to an order when matrix / vector multiplication is calculated by adopting a grouping and packaging algorithm; matrix / vector multiplication which is equivalent to original model reasoning and is friendly to hardware is executed; the result is converted back to the floating point data type through an inverse quantization algorithm, a final calculation result of the large language model in the matrix / vector multiplication operation is obtained, the final calculation result is transmitted into a subsequent network layer, and calculation acceleration for the matrix / vector multiplication operation in the large language model is completed. According to the method, the accuracy is ensured, and meanwhile, the speed is increased during large language model reasoning.

The invention discloses a metadata query method based on a knowledge graph, which relates to the technical field of data query, and comprises the following steps: collecting metadata from various data sources; then respectively performing knowledge extraction on the data, wherein the knowledge extraction comprises entity extraction, relationship extraction and attribute extraction; constructing a data asset directory according to the metadata after knowledge fusion and a knowledge extraction result of the metadata; then knowledge modeling is carried out, and a knowledge graph is established; and finally, constructing a retrieval enhancement generation process by utilizing a large language model, and retrieving the metadata. According to the method, the natural language can be understood and generated by utilizing the large language model and the retrieval enhancement generation process, and the RAG process retrieves related knowledge and inputs the related knowledge into the language model as a context, so that the processing capability of the model on complex queries is enhanced. According to the method, the extraction accuracy and efficiency can be improved, the method can better adapt to new fields and data types, and the requirement for manual intervention is reduced.

The invention discloses a privacy protection data sharing and cooperative training method and system based on federated learning. The method comprises the steps of receiving software development log data, adaptively judging the sensitivity degree according to a data type, dynamically adjusting noise disturbance intensity according to the sensitivity degree to perform data desensitization, and generating a sensitivity index; selecting a feature extraction strategy, extracting time sequence correlation features from the desensitization data, constructing a dynamic graph structure with a weight, and obtaining a time sequence feature vector through iterative fusion; calculating the time sequence correlation of the time sequence feature vector to obtain a data quality score, and setting a contribution weight based on the quality score to perform parameter aggregation; combining sensitivity indexes with data quality scores to construct a security sharing domain, decoupling global training parameters into knowledge fragments in the domain, formulating a recombination rule, and selectively acquiring the required knowledge fragments by all parties for local training. According to the method, deep collaboration is realized on the premise of protecting data privacy, and the collaboration training effect is improved.

The invention belongs to the technical field of multi-source heterogeneous data quality and analysis, and relates to a data management and intelligent analysis method oriented to power multi-source heterogeneous. According to the method, ontology model alignment, temporary ontology generation and multi-mode semantic embedding technologies are adopted, semantic unification of cross-data types is achieved, meanwhile, a spatio-temporal joint indexing mechanism is established, geographic coordinates and timestamps are bound and stored, efficient multi-dimensional query is supported, and data integration efficiency and semantic consistency are improved; the problem of poor adaptability of a fixed threshold value is solved by dynamically adjusting the operation data safety threshold value and evaluating the abnormal confidence coefficient, the detection precision and credibility of the power multi-source heterogeneous data anomaly are improved, the anomaly detection accuracy is improved, and the false alarm rate is reduced; by constructing an abnormity confirmation logic and combining equipment parameters, environment data and communication states, reasons such as equipment faults, environment interference and communication abnormity can be accurately positioned, root causes can be quickly positioned, and time can be shortened.

The invention provides a fault diagnosis question-answering system based on a multi-modal knowledge graph and a large language model, and the system comprises a time sequence data processing module which is used for carrying out the noise reduction of time sequence data, carrying out the feature extraction of the time sequence data after noise reduction, and representing a time sequence data body through a plurality of feature data; the multi-modal knowledge graph module is used for constructing a multi-modal knowledge graph, and the multi-modal knowledge graph comprises a text modal knowledge graph and a time sequence data modal knowledge graph; the large model fine tuning module is used for performing fine tuning on the large model by adopting a LoRA model based on different fault problems and corresponding solutions so as to obtain a question and answer large model; and the RAG question and answer module is used for enhancing the user question based on the multi-modal knowledge graph, inputting the enhanced statement into the fine-tuned large model, and outputting a result. According to the method, the fault diagnosis speed is increased, fault questions and answers are carried out on various data types, and the problem that a large model lacks professional domain knowledge is solved.

The invention discloses a fault diagnosis method and device, a medium and a product, and relates to the technical field of data processing. Multi-dimensional data such as text logs, time series data, topological graph structures, operation records and environmental parameters are obtained, various factors possibly involved when the fault occurs are included, different types of data can complement and verify each other, the fault features are accurately described, the limitation of single data type analysis is avoided, and the analysis efficiency is improved. And the accuracy of fault diagnosis is effectively improved. Then converting different types of multi-dimensional data into target feature representations, matching the target feature representations with data in a historical fault case library, and when a historical fault of which the matching degree is greater than a preset matching threshold value is found, determining a diagnosis conclusion and a repair scheme of the historical fault as a diagnosis conclusion and a repair scheme of the current fault. And each new fault does not need to be analyzed and reasoned, so that the fault diagnosis time is shortened, and the fault diagnosis efficiency is improved.

Predicting the salience of one or more data entities to a particular (target) data entity from among a plurality of data entities may comprise generating a graph of the plurality of data entities and a machine-learned model architecture that predicts the salience of the one or more data entities output by the machine-learned model architecture using the graph. For example, the machine-learned model architecture may comprise a first machine-learned model for generating an embedding using the content of the target data entity, a second machine-learned model for generating a vector using the data type indicated by the target data entity, and a third machine-learned model (e.g., a graph neural network or other feed-forward neural network) for generating a contextual representation of the target data entity to which other contextual representations associated with the plurality of data entities may be compared (e.g., using Euclidean distance, cosine similarity, dot product).

The invention discloses an industrial internet data processing method based on big data, which comprises the steps of designing a visual data processing flow arrangement tool, allowing a user to flexibly define logic and steps of data processing, and realizing common data processing operation through a predefined data processing operator library in the flow arrangement process. A user is allowed to select a required data processing component through dragging operation through a visual configuration interface, and corresponding data types and processing conditions are set to form a complete data processing flow chart; acquiring data type information configured by a user, performing primary condition filtering judgment on the data according to the information, and screening out original data meeting conditions; judging whether data needs to be dynamically calculated along with time or not according to whether conditions are changed or not; a real-time data processing engine is introduced, and data are collected, cleaned, converted and analyzed in real time; through a visual interface, NLP interaction, real-time processing and intelligent scheduling, the flexibility of data processing is improved.

The invention provides a heterogeneous data migration method and device, equipment and a medium, and the method comprises the steps: determining a reading path according to a source cluster, and obtaining to-be-migrated source data from the source cluster based on the reading path; determining a required target data mapping rule according to the current data migration scene; constructing an intermediate mapping model based on the data type of the source cluster and the data type of the target cluster, and mapping a field of the source data to a mapping field of the intermediate mapping model by using the target data mapping rule to obtain converted source data; and loading the converted source data into the target cluster by adopting a batch loading mode so as to complete migration of the source data. By means of the method and device, data migration in various heterogeneous database environments is supported, and the efficiency of data migration synchronization in the heterogeneous cluster environment is effectively improved.

The invention discloses an AI model service platform data set labeling method and system, and relates to the technical field of data set labeling, a to-be-labeled data set is split according to a data type, and whether an existing labeling model capable of respectively processing the currently split to-be-labeled data set exists in an AI model service platform is identified; evaluating the annotation performance of the annotation model based on the evaluation index of the annotation model, and then updating the annotation model based on data enhancement; labeling the split to-be-labeled data set by using the updated labeling model to generate a first labeling result; performing secondary annotation on the first annotation result based on multi-model joint annotation to generate a second annotation result; and screening and identifying an unqualified annotation data set in the second annotation result through the confidence interval so as to trigger manual annotation. An efficient, low-cost, high-quality, flexible and extensible data set labeling process is realized, so that the labeling capability and the service quality of an AI model service platform are remarkably improved.