22 results about "Dynamic query" patented technology

This application discloses a method and system for identifying 2D engineering drawing elements based on dynamic query-guided consistent projection, proposing a dynamic query-guided consistent projection decoder. This decoder achieves a single forward propagation projection from a random noise vector to an accurate element state vector. The method first initializes a set of vectors following a predetermined noise distribution as initial dynamic queries. These queries are fed into the dynamic query-guided consistent projection decoder (DQ-CPD). At each layer of the decoder, the current dynamic query vector not only interacts with other queries through a self-attention mechanism, but it is also input as a controller into a "contextualized hypernetwork." This hypernetwork generates a set of dedicated weight parameters in real time based on the current inference information carried by the query vector. This method organically combines dynamic queries, dynamic context, and one-step generation theory to construct a novel, efficient, accurate, and adaptive element identification method.

The application discloses a kind of multilayer cascade safety management methods for constructing with house as carrier, comprising: obtaining the physical structure data and spatial coordinate data of house, and fusing with standardized address matching, generate house unique basic identification;Based on the identification, obtain historical property change record, carry out address consistency check and association, generate complete property chain;Obtain enterprise subject registration information, and associate mapping with the latest property record, determine the enterprise subject and its occupation relationship of current occupation house;Obtain house physical state detection data and accident history data, identify risk hidden point in combination with enterprise subject and occupation relationship, and determine corresponding direct and indirect responsibility subject, build responsibility traceability path;Finally, house basic identification, property chain, enterprise occupation relationship and responsibility traceability path are associated with graph data structure Storage, form the safety responsibility knowledge graph that can be dynamically queried and traced back.The application realizes the whole chain through and dynamic tracing from house entity to responsibility subject.

This invention belongs to the field of data analysis and query optimization technology. Specifically, it relates to an interactive, visual data analysis and dynamic query optimization system, comprising: a multi-source data aggregation module for extracting query feature sets from interactive query data; a feature extraction and construction module for building an interactive query prediction model based on the user's analytical intent and a pre-defined query optimization engine and query feature set; a model intent prediction module for outputting a dynamic query plan set based on the interactive query prediction model and establishing a switching mechanism; a plan generation and switching module for generating a dynamic analysis view set based on the query execution result set using a multi-version view adaptation strategy combined with the switching mechanism; and a view rendering and output module for optimizing rendering performance and incrementally updating the dynamic analysis view set, outputting a real-time responsive analysis visualization interface. This invention enhances the level of analytical intelligence and automatically and accurately identifies the user's true analytical intent.

This invention discloses a visualization report analysis method based on task discrimination and query generation. It receives user queries, generates a sequence of subtasks through task planning using a large language model, and evaluates and optimizes the subtask sequence using a task planning discriminator. For each subtask, after vectorization, similarity calculation is performed with a tool description vector library to obtain the top K tool descriptions, and the large language model further refines the selection to N tools. When a subtask involves a database query, similarity matching is performed based on a query example library to obtain query examples, which are then combined with table structure information to generate and execute a query statement. Finally, a visualization report is generated based on the subtask processing results. This invention effectively improves the accuracy of the large model in task decomposition, tool invocation, and query statement generation by introducing a task planning discriminator, vectorized tool recommendation technology, and dynamic query statement suggestion technology, thereby enhancing the execution effect of complex data queries.

The application relates to the technical field of database intelligent query, in particular to a dynamic query method and system based on a running scene and electronic equipment thereof, the method comprises the following steps: generating real-time running data based on business link data and a preset business program after mounting; obtaining a query statement corresponding to a query user, outputting relevant training data based on a query vector combination and a preset vector database; inputting the query statement and the relevant training data into a large language model, and calling a preset parameter mechanism to obtain a structured parameter; positioning an abstract filtering template in a template database based on a positioning feature, and dynamically cutting the abstract filtering template based on a condition feature to obtain an SQL standard statement corresponding to the query statement. The application adopts a template and a dynamic cutting strategy, enhances query accuracy and user intention matching degree, and significantly improves the accuracy and practicability of natural language query in a complex business scene under the premise of guaranteeing correct SQL execution.

Systems and methods for caching queries and query results data to improve the efficiency of responding to query requests are described. A dynamic query data caching system may generate query data structures for newly received query requests, adding the generated query and a weight for individual queries to such query data structures. Query results data may be obtained and stored for use in responding to future query requests. Weight may be reduced over time and compared periodically to a threshold to determine if the query should be removed. If the query gets used before it expires, the weight may be reinitialized. Once a weight meets the threshold, the query and any associated results data may be deleted.

The application discloses a kind of genomic variation atlas visual interactive systems, system includes data access and pretreatment module, multidimensional variation atlas rendering module, interactive control module, dynamic query and filtering module and result output module.Data access and pretreatment module generates standardized variation feature vector set to multiple types of variation data, multidimensional variation atlas rendering module simultaneously renders the position, type, frequency and pathogenicity of variation with position, shape and color semantic coding triple dimensions;Interactive control module realizes continuous seamless zooming from whole genome macroscopic thermal view to single base resolution;Dynamic query and filtering module supports composite condition real-time filtering and atlas synchronous linkage through multidimensional inverted index;The application effectively solves the problems of insufficient multidimensional attribute coding, fragmented interaction levels and missing screening and atlas linkage in existing tools, significantly improves the efficiency of visual analysis of massive genomic variation data.

The application discloses an online state monitoring method and system based on an Internet of Things device, and the method comprises the following steps: periodically measuring and reporting network quality data; adaptively adjusting the reporting period based on the network quality data; receiving the network quality data and updating state record data, and performing online state detection; and updating state confidence scores based on the state record data. Thus, the Internet of Things device self-checks the network quality data and adaptively adjusts the reporting period, thereby reducing the reporting frequency and network occupation. The server side maintains state record data based on the network quality data, sets state confidence scores for each Internet of Things device, dynamically sets timeout judgment thresholds according to the classification, performs dynamic queries on the Internet of Things device, and reduces the occupation of computing resources in the query process. The state record data is maintained in a ring queue and is updated based on the state change increment of the Internet of Things device, so that the state information obtained in the active query or passive push can be ensured to be accurate and reliable.

The application relates to the field of internet search services and discloses a semantic retrieval system for cultural gene indexes, which comprises a basic retrieval subsystem, a dynamic graph generation subsystem and a dynamic query arbitration subsystem. The dynamic graph generation subsystem acquires an internet text stream, determines a variant word item through co-occurrence distribution drift rate and information entropy increment, constructs a dynamic evolution graph to determine a graph mapping path, and generates a bit mask according to the graph mapping path. In the storage addressing stage, the bit mask shields a corresponding storage block in a static inverted index structure, addressing interception of a variant noise document is realized, the application realizes decoupling of a semantic evolution capture mechanism and an underlying inverted index architecture, effectively suppresses storage write amplification and retrieval load avalanche caused by sudden increase of global document frequency of the variant word item, and guarantees response stability of the retrieval system under a high concurrency environment.

The application discloses a kind of intelligent customer service dialogue generation methods based on multimodal interaction and dynamic knowledge enhancement, specific steps are as follows: S1: intention recognition: according to the content input by user, the intention of user is identified;S2, knowledge retrieval: based on user intention and historical interaction record, generates dynamic query vector, and carries out weighted retrieval in a double-source knowledge base combined with knowledge graph and document paragraph, wherein, the weight of knowledge is dynamically updated according to user feedback;S3: reply generation: input user intention and the knowledge segment related to intention into large language model, obtain several candidate replies;S4, compliance verification;S5, response output.The application significantly improves the interaction accuracy, knowledge timeliness and service personalization level of intelligent customer service through multimodal fusion understanding, dynamic knowledge enhancement and controllable content generation.

This invention discloses a method and system for filtering dynamic model data. The method includes: S1: parameter receiving, receiving raw data, filtering parameters, and sorting parameters transmitted from the front end; S2: dynamic condition parsing and filtering; S3: intelligent data type recognition; S4: dynamic multi-field sorting, constructing a comparator chain and performing multi-field priority sorting; S5: result truncation, truncating a predetermined number of rows from the sorted data and outputting the final result. This invention can be widely applied to backend interface data processing, data analysis platforms, dynamic query systems, and other scenarios, solving the core needs of dynamically configuring filtering conditions and intelligently adapting data types in multiple scenarios.

The application discloses a speech large model adaptation method and device for a low-resource language, and relates to the technical field of natural language processing. The method comprises the following steps: first, frame a speech signal, input an encoder with frozen parameters, and extract a time sequence acoustic feature sequence; then, pass through a feature enhancement module composed of linear projection, two-layer Transform and a feedforward network, capture global acoustic context, fuse fine-grained acoustic features through a target modal encoder and linear mapping, construct an adaptation structure based on a dynamic query transformer, utilize an explicit inductive bias connected with time sequence classification to guide the dynamic generation of a query vector, and input the query vector into a large language model trained through a frozen training strategy to generate text. The application realizes efficient feature compression and cross-modal alignment of a variable-length speech signal in a low-resource scene, and realizes the best convergence of model performance under very few samples.

The application discloses a scene graph generation method based on dynamic query distribution and decoupled attention mechanism, comprising the following steps: inputting a training image into a backbone network and a Transformer encoder; generating a relationship query through an adaptive semantic relationship query generation module; inputting the relationship query into a relationship query attention decoupling decoder; in a one-to-many relationship coarse positioning module, calculating a one-to-many stage loss; in a one-to-one relationship fine positioning and deduplication module, calculating a one-to-one stage loss; weighting and combining the one-to-many stage loss and the one-to-one stage loss to obtain a total loss, and training the scene graph generation; if the set batch processing size is reached, returning the image training; if all the training pictures are read, outputting the trained model. The application can select high-quality tokens from the encoder tokens to construct a relationship query through the adaptive semantic relationship query generation module.

The application discloses a rule engine execution method, device and equipment and a computer readable storage medium. The application assembles a rule instance object structure value according to preset rule parameter classes, obtains a target condition object instance, constructs a hash key value according to the target condition object instance and a fixed query condition array, constructs a combined key value according to the target condition object instance, a dynamic query condition array and a classification combination condition array, and obtains a rule result from the rule engine according to the hash key value and the combined key value. The application processes rule conditions based on the rule engine, separates business logic from code, and thus simplifies development of business code.

The application discloses an infrared video small target detection method based on global dynamic query. The application constructs a dynamic query aligner (DQAligner) framework, uses global random large displacement data enhancement to expand long tail sample distribution, and improves the adaptability of the model to the rare large motion scene. A pyramid deformable convolution with a dynamic receptive field is used to decompose complex multi-scale motion, and fine alignment of the target is realized. Further, the generalized residual form of deformable convolution is used to iteratively learn the target dynamic query storage, and the target position is searched in the frame to maintain the semantic consistency of the target in multiple frames. The modules of the DQAligner work cooperatively, realize the paradigm shift from rigid alignment to flexible matching, and significantly improve the detection performance in the large displacement and complex dynamic scene.

The application provides a three-dimensional Gaussian code generation system based on multi-agent cooperation of multi-dimensional feedback, integrates visual quality verification, performance analysis and static code analysis into a dynamic optimization cycle, and uses multi-agent cooperation for 3D Gaussian splat code optimization; meanwhile, cross-task optimization migration and automatic error detection and repair are carried out by using a knowledge base; compared with existing manual optimization or single-agent optimization solutions, the application uses multi-dimensional feedback integration (visual quality + performance analysis + static analysis) to enter the optimization process, can be dynamically queried to the optimization experience knowledge base, corresponds to the reconstructed scene, and thus provides higher performance portability and adaptability, and can efficiently and stably generate and optimize Gaussian splat codes for complex 3D scenes, such as large 3D reconstruction, real-time rendering application and dynamic scene.

The application relates to the technical field of gait recognition, and provides a lower limb gait action recognition method. The method comprises the following steps: collecting multi-source sensor data of a lower limb, and constructing a multi-modal feature sequence; inputting a bidirectional recurrent neural network for feature coding to obtain a hidden state sequence; generating a dynamic query vector for a preset action category through an action-specific attention mechanism, calculating attention weights in a local sliding window with the current time as the center, weighting and fusing the hidden state sequence based on the attention weights, and mapping the hidden state sequence to a preliminary action probability sequence through a classifier; taking the preliminary action probability sequence as an observation probability, combining a preset state transition probability matrix to construct a hidden Markov model, globally decoding the hidden Markov model to obtain a corrected action sequence that satisfies a biomechanical state transition constraint. While improving the recognition accuracy of sparse key actions, the method ensures that the output action sequence conforms to the biomechanical logic of a human body.

The application provides a health and hygiene data classification and grading method based on multi-agent cooperation, wherein a general intelligent agent analyzes health and hygiene data in parallel with a basic form track and a regular format track and generates a health and hygiene data set, a specialist intelligent agent sequentially performs de-identification processing and classification and grading processing according to the health and hygiene data set, and a cross-class arbitration intelligent agent generates calibration category level data; a privacy protection intelligent agent acquires all calibration category level data and de-identified health and hygiene data; the privacy protection intelligent agent performs static data anonymization processing on an input static data set for public or internal use, and performs privacy protection processing on a query action through a dynamic query interface. By using the above method, high recall rate and high precision are realized in a full-automatic process, and the workload of manual review is greatly reduced. Sensitive data is strictly protected, and low-risk data can be reasonably utilized, thereby realizing the optimal balance between safety and utility.

The application discloses a data processing method, device, equipment and medium, and belongs to the technical field of data processing. The method comprises the following steps: in response to the system time reaching a preset dynamic processing time point in a dynamic data processing time point sequence, triggering a batch data processing event automatically, performing a dynamic query on each registration and registration institution according to the data file delivery state of each registration and registration institution, obtaining all registration and registration institution identifiers, performing an aggregation operation on all registration and registration institution identifiers, and generating a current batch of to-be-processed registration and registration institution task list; performing a batch data processing task on the current batch of to-be-processed registration and registration institution task list; and in response to all data processing time points in the dynamic data processing time point sequence being completed, completing the dynamic batch data processing based on a time event driving. The application can realize the on-demand starting of a data processing task, thereby maximizing the data processing efficiency and meeting the diversified business timeliness requirements.

This invention belongs to the field of computer-aided design secondary development technology, and specifically relates to a method for dynamically querying CAA component information. The method includes: utilizing publicly available information of the target software, obtaining the address of the QueryMetaObject metadata query function through debugging; entering the function body, and within the function body, finding the nearest relative addressing function instruction; entering the corresponding function body, locating the global dictionary through the relative addressing instruction of global variables, parsing the global dictionary according to the hash linked list data structure, and constructing a query tool; inputting query conditions in the query tool interface, traversing the linked list nodes, and completing the CAA component information query. This method has high retrieval performance, richer query fields, and more complete query results, making it suitable for CAA component secondary development scenarios. This invention also provides a non-transitory readable recording medium storing the program of this method and a system containing the medium, which can call the program through processing circuitry to execute the above method.