25 results about "Dynamic Extension" patented technology

The invention discloses a low-code platform design system based on modularization and a dynamic extension component, and relates to the technical field of low-code development, and the system comprises a form module which is used for standardly defining a data model and dynamically rendering a user interface, realizing unification of form configuration through a structured metadata specification, and establishing a form module; bidirectional binding of the view and the data model is realized; the process module is used for realizing modeling, execution and state management of a business process, supporting automatic task scheduling and multi-instance management based on standardized process elements, and keeping consistency of a process state through an increment synchronization mechanism; and the data processing module is used for cooperatively managing platform data resources, supporting dynamic expansion and decoupling dependence between the form module and the process module through an event-driven mechanism. According to the method, the strong binding relationship between the decoupling form and the process is established, the reusability of the process module is improved, and the process module is independent of the form and can independently provide API service to the outside, so that the method is adaptive to more business scenes.

The invention discloses an integrated design management system with an automatic registration communication agent engine and a method thereof, and relates to the technical field of software integrated design simulation. The method has a heterogeneous integration breakthrough, the five communication modes are matched with the software without the API / command line / file, different heterogeneous software can support one communication mode, and the integration cost is remarkably reduced; dynamic expansibility is achieved, a newly added software module only needs to deploy a proxy engine and configure registration, and zero-shutdown upgrading of the platform is achieved; high concurrency reliability is achieved, resource conflicts are solved for an instance mechanism, and parallel task efficiency is improved; and full-process automation is achieved, and the cross-software design process execution time is obviously shortened.

This invention relates to an enterprise-level full-stack low-code development platform based on a model-driven and plug-in architecture, belonging to the field of computer technology. It addresses the problems of poor scalability, performance bottlenecks, high security risks, and inflexible deployment inherent in existing platforms. The invention includes: a visual development tool providing a form designer and workflow engine, supporting drag-and-drop development and real-time preview; a plug-in extension module, implemented based on the PF4J framework, supporting dynamic loading, unloading, and isolation of plug-ins; a data integration and management module providing an API gateway, multi-source data integration, and data model management; and an application deployment and maintenance module for one-click automated deployment and log auditing. This invention achieves dynamic functional expansion through a plug-in architecture, optimizes performance in high-concurrency scenarios through a workflow engine, enhances security, and provides more flexible deployment, improving the development and maintenance efficiency of enterprise-level applications. It is suitable for application scenarios such as government approvals.

The invention relates to the field of program development, and provides an iOS system dynamic interaction scene construction method and system. The method comprises the steps that a dynamic class construction tool is developed, page classes are dynamically registered and class members are added when an application runs, and a page class structure capable of being dynamically expanded is obtained; defining a communication protocol between the application and the cloud, obtaining rendering parameters from the cloud based on the communication protocol, assigning the rendering parameters to view attributes of page class instances corresponding to the page class structures, and obtaining a dynamic view module comprising page objects; pre-defining a user operation event, binding the user operation event to a corresponding view of the page object according to event binding configuration of the cloud, and obtaining an event response module; setting a dynamic code execution entry in a system life cycle event of the application so as to execute a JavaScript code issued by a cloud end and obtain a logic execution module; and assembling a rendering system, generating a native UI interface and responding to user interaction operation. The dynamic interface updating of the application is realized, and the user experience is improved.

This invention relates to the field of information technology, specifically to a method and system for dynamically extending remote invocation based on Java dynamic bytecode. The method includes: establishing a C / S architecture, setting up a communication channel between the client and server, with the server providing an interface for maintaining remote invocation tasks of Java source code type; defining remote tasks of Java source code type through the interface, dynamically pre-compiling the Java source code into task bytecode using Javassist, and storing it on the server; the server initiating a remote invocation to a specified client using remote invocation task ID, version number, and business data as parameters; the client creating a custom ClassLoader to remotely load the task bytecode from the server and generate a Java class; and the client executing the method of the remote task object and returning the invocation result. This invention performs a single compilation, allowing multiple clients to run the code, eliminating the need for repetitive client-side compilation.

The application provides a dynamic expansion method of a model and a device thereof, and relates to the technical field of terminal equipment, financial technology and information technology industry. The application determines a target component from a graphical component library, generates a business model based on the target component, structures the business model in a meta model, obtains metadata of the business model, and reflects the metadata to generate a Java object of the business model runtime. The application provides a method for writing a business model through graphing, which can quickly generate a backend Java object while generating a front-end webpage. Compared with a traditional programming code writing method, when an application developer designs an application through a visual method, the threshold is low, the iteration speed is improved, the model function can be dynamically enhanced at runtime, the model method does not need to be defined in advance, the program maintenance cost is reduced, and the expansion capability of the program is improved.

Provided are a dynamic expansion method and device for attribute data and a storage medium, which can realize dynamic expansion of attribute data of a complex type. The method comprises: obtaining attribute data to be dynamically expanded and target attribute data templates corresponding to the attribute data, which have been subjected to flattening mapping processing. Further, the target attribute data is subjected to flattening mapping processing to obtain a second mapping relationship, which can be a primary key value format. Then, according to configuration information or verification rules of each attribute defined by the target attribute data templates subjected to flattening mapping processing, dynamic attribute verification processing is performed on the value of each attribute of the attribute data subjected to flattening mapping processing to obtain a dynamic attribute verification processing result of the attribute data. Further, the target attribute data is dynamically expanded according to the dynamic attribute verification processing result of the target attribute data.

The application provides a multi-level page table traversal acceleration method, system, device and medium, and belongs to the technical field of computers. The method comprises the following steps: inputting a virtual address into a path predictor and a dynamic weight hierarchical cache, and collecting an instruction access proportion, a TLB hit rate and a prefetch success rate in real time; performing hash calculation on the virtual address according to a preset bit of the virtual address by the path predictor, querying a history record table to obtain a predicted path and a confidence level, so as to determine a query range; when the cache hits, the score of the hit entry is calculated, the priority of the cache entry is sorted and the elimination decision is made, and a hybrid replacement algorithm is used to manage the cache entry; under a preset condition, a prefetch operation is started by an adaptive prefetch engine, the prefetch step is calculated and dynamically expanded, and memory bandwidth contention is prevented through two-level monitoring; when the instruction access proportion, the TLB hit rate and the prefetch success rate meet the conditions, an instruction priority mode is activated, the score of the hit entry is updated by adjusting the related calculation parameters.

The invention provides a public component development management method. The method comprises the following steps: establishing a code writing specification, a dynamic expansion mechanism, a code annotation specification and a version control strategy for developing a public component; establishing a development team training mechanism and a code review mechanism based on the code writing specification; designing dynamic parameters based on a dynamic expansion mechanism; optimizing the annotation document based on the code annotation specification; and managing the version of each public component in the public component library based on the version control strategy. By establishing a series of management specifications and optimization mechanisms, the development efficiency, quality and maintainability of the public component are improved, and the flexibility and expandability of the public component are enhanced. Through the implementation of the steps, it is ensured that developers can rapidly and easily integrate and call the public components, and meanwhile customized configuration and expansion can be conducted according to specific requirements of the project, so that the development efficiency is improved, the maintenance cost is reduced, and successful implementation of the project is promoted.

The application relates to the technical field of computers and provides a document automatic generation method and system based on a language generation model, which comprises the following steps: in response to generation request information, creating a task scheduling management unit, defining a plurality of initial tasks and creating corresponding task description bodies, and constructing a task association topology based on data dependency relationships; the task association topology is cyclically scanned, schedulable tasks are identified and submitted to a concurrent execution unit for parallel processing, and a language generation model is called to generate content segments; after target task execution is completed, a new task description body is generated based on the output result, corresponding dependency relationships are established and registered to the task association topology, and result-driven incremental topology growth is realized; after all the tasks are completed, the content segments are synthesized and content integration processing is performed, and a target document is generated. The application realizes efficient concurrent generation and dynamic expansion of the structure of the document, and guarantees context consistency.

Systems, computer program products, and methods are described herein for developers to initiate development operations from within a development environment of a user application. This can also be extendable dynamically for different user groups (e.g., testers, support, internal associates or employees, or the like), and environments (e.g., lower-level environments, production environments, or the like). As such, the invention offers a dynamic debug mode feature for virtual assistants based on different sections of users and environments.

The class incremental learning method based on multi-granularity distillation and weight alignment comprises the steps that under a given wireless electromagnetic signal data set, an incremental deep learning classification model is initialized, and the model dynamically expands an output layer to adapt to a newly-added class; for a training sample of a current task, inputting the training sample into the model for forward propagation, obtaining a prediction score vector of the model on each category, and calculating cross entropy classification loss for learning discriminant features of a newly introduced category in the task; after training of each task is completed, forward propagation is carried out on all training samples of the current task by using the current model to obtain logs representation of the training samples, and clustering is carried out; in each iteration in the training process, calculating cross entropy loss by using a training sample of a current task, and forming a loss function of a historical task sample; after task training is finished, weight alignment operation is executed, and the difference of new and old category classifiers in weight amplitude is eliminated; and repeating the above steps, performing incremental learning task by task, and continuously performing model parameter updating and memory bank maintenance until all incremental learning tasks are completed.

The present application belongs to the technical field of computer vision, and particularly relates to an open-vocabulary object detection method based on multi-modal fusion and dynamic expansion, aiming to solve the problems of insufficient robustness of visual features, static multi-modal fusion, and poor flexibility of class expansion in the prior art. The method first acquires an image to be detected and text information containing class and scene description, extracts features through a visual encoder and enhances them through a robust noise reduction layer, and then generates a candidate detection frame through a full convolutional network, while extracting text features through a CLIP text encoder. Subsequently, the candidate region features are extracted, dynamically fused through an adaptive multi-modal fusion module to generate cross-modal joint features, and finally the detection frame is screened and the confidence is calculated according to the feature similarity. The model training fuses the InfoNCE loss and the cross-entropy loss to optimize the parameters, and the new class dynamic expansion can also be realized through the analysis of natural language by a generative language model. The present application significantly improves the robustness and accuracy of detection in complex scenes, optimizes the multi-modal fusion efficiency, greatly reduces the cost of new class expansion, and the model meets the real-time detection requirements after acceleration, has good engineering deployment advantages, and can be widely applied in the fields of intelligent monitoring, automatic driving and the like.

The embodiment of the invention relates to the field of data conversion, and provides a multi-source domain-specific language metadata conversion method and computing device.The method comprises the steps that in response to received domain-specific language DSL data, the language type of the DSL data is obtained; based on a first grammar rule corresponding to the language type in a preset grammar rule base, analyzing a semantic structure corresponding to the DSL data; converting the semantic structure into first metadata according to the semantic structure and a preset mapping rule base; and processing the first metadata to generate second metadata conforming to a preset format. Based on the scheme, unified analysis and conversion can be carried out on the multi-source DSL data, the adaptation cost is effectively reduced, and the expansibility and compatibility of the system are improved. And through the modular design of the grammar rule base and the mapping rule base, the analysis capability of dynamically expanding the new type DSL is supported, and the adaptation cost of an external development platform is reduced.

The invention provides a self-evolution model generation and real-time optimization method for an intelligent dialogue system and a related device, and relates to the technical field of artificial intelligence. By establishing a collaborative optimization mechanism, dynamically expanding a model architecture and realizing real-time iterative update, the problems of model update lag, learning mechanism isolation, architecture stiffness and iteration delay are solved, and the method has the advantages of improving the dynamic adaptability, collaborative optimization capability and real-time response efficiency of the system.

The present application provides a multi-version dynamically extensible data source method, device, terminal and storage medium, which relates to the technical field of database data source loading. The method includes: creating a JAVA SDK package, the JAVA SDK package includes core dependencies, core entity classes, core interface classes, database connection pool management core classes, thread isolation, routing data source classes and basic properties of routing data source classes and target data source mapping, the target data source mapping is used to store the loaded database connection pool; based on the JAVA SDK package, initialize the system database connection pool in the application system; after initializing the system database connection pool in the application system, based on the JAVA SDK package, dynamically expand the new database connection pool in the application system. The present application has powerful dynamic expansion capabilities and loosely coupled design features, which can greatly improve the flexibility and extensibility of the system and reduce development and maintenance costs.

According to the low-code platform dynamic expansion system and method based on the standardized plug-in architecture, metadata, interfaces, dependency and authority specifications of the plug-ins are defined through the standardized meta-model, compatibility between the plug-ins and the platform and compatibility between the plug-ins are ensured, and the fragmentation problem of plug-in development is solved. The AI capability engine supports plug-in full-life-cycle intelligent decision making, the development threshold is greatly reduced, and the development and use efficiency is improved. During running, the container adopts a modular framework, dynamic loading / unloading of the plug-in is supported, the platform does not need to be restarted, the platform is guaranteed to be continuously available, and the problem that a traditional plug-in needs to be shut down during loading is solved. Through resource isolation and role-based access control (RBAC) authority definition, the influence on the overall performance of the platform due to the abnormity of a single plug-in is avoided, the unauthorized operation is prevented, and the operation safety of the platform is ensured.

The invention relates to a function library maintenance and formula editor function, aims to provide a method for efficiently and conveniently managing and using a function library, and overcomes the defects of an existing formula editor in the aspects of function management and automatic complementation. The method comprises the following steps: constructing a function library management module for storing, classifying and maintaining mathematical functions and related information thereof; a formula editor module is provided, a user is supported to input and edit a formula through a graphical interface, and a function in a function library can be quickly inserted; a function automatic association function is realized, and functions are intelligently recommended according to user input contents; the method supports dynamic expansion and maintenance of a function library, allows a user to customize addition, modification and deletion of functions, and supports team cooperation and sharing. The method can significantly improve the efficiency of formula editing, simplify the management of the function library, support personalization and team cooperation, is especially suitable for the fields of scientific research, engineering, industry and the like, and has a wide application prospect.

The invention relates to the field of distributed operating systems, and discloses a dynamic expansion-supporting hybrid architecture-based swan-gap system adaptation method, which is technically characterized by comprising a registration verification step, verifying the configuration file of the application module through a preset multi-dimensional verification mechanism to obtain a registration verification result; an operation monitoring step: during the operation period of the application module, acquiring the resource use condition of the application module in real time by using the improved monitoring index set, and generating multi-dimensional monitoring data with a time stamp; a resource demand prediction step: obtaining the multi-dimensional monitoring data, generating a prediction result based on the time sequence prediction model, the prediction result reflecting the resource demand of the application module in the future time period; and a dynamic resource adjustment step: according to the prediction result and the registration verification result, dynamically adjusting the resource allocation of the application module by using a preset dynamic resource adjustment strategy.

The invention discloses a large model tool learning method based on sample enhancement and dynamic course learning. The method comprises the following steps of: 1) performing sample enhancement through methods of randomizing functions, parameter names, dynamic extension function description and the like, and forcing a model to deeply understand function description of a tool instead of depending on a surface name so as to cope with challenges of non-uniform tool names, incomplete documents and the like; and 2) in model training, a dynamic sample selection mechanism is introduced, and the mechanism firstly represents the complexity of a dynamic evaluation task in real time based on the model in the training process, and then adaptively selects a training sample with the most suitable difficulty according to the complexity, so that the learning efficiency and effect are optimized. Compared with the prior art, through sample enhancement and a dynamic training strategy, the accuracy and generalization ability of a large language model during calling of complex and special tools are remarkably improved.

The invention relates to the technical field of document processing and text analysis, and discloses a text comparison algorithm based on a sliding window and a dynamic extension KMP algorithm, which comprises the following steps of: setting a minimum matching threshold value, extracting substrings character by character in a document by using the sliding window with the step length of 1 as a to-be-matched mode, and then performing scanning matching by using the improved KMP algorithm to obtain a matched mode. The actual position information of the characters in the original text is completely reserved, position deviation and context loss caused by preprocessing in a traditional method are effectively avoided, after preliminary matching succeeds, a system automatically enters a dynamic expansion stage, the matching range is prolonged character by character until continuous matching cannot be achieved, and therefore the longest repeated text segment is determined, and after matching is finished, the matching efficiency is improved. The adjacent or overlapped matching fragments are intelligently merged and structured through a special result merging module, and finally a clear result containing original starting and ending positions, repeated content and length information is output, so that the efficiency and accuracy of large-scale document comparison are remarkably improved through the operation.

The application provides a trusted data space dynamic expansion method and system of a micro-service architecture, relates to the technical field of data processing, and establishes a hybrid expansion strategy of a computing resource layer, performs task prediction according to a task database, establishes a task prediction result, performs adaptive evaluation of the hybrid expansion strategy in the computing resource layer by using the task prediction result, preconfigures the hybrid expansion strategy, performs hot spot calling prediction of data, synchronously establishes a distributed data storage layer by using a hot spot data prediction result, and after mapping the distributed data storage layer and the hybrid expansion strategy, the dynamic expansion of the trusted data space is completed. The technical problem that the trusted data space of the micro-service architecture cannot flexibly cope with high concurrency and dynamic changes in tasks in the prior art is solved. The technical effects that the trusted data space is dynamically expanded, the flexibility of resource scheduling is improved, and data security and high availability are ensured are achieved.

The invention is applicable to the technical field of data processing, and provides an interface expansion and data processing method based on an applet platform, which comprises the following steps: if detecting that an applet is in a dynamic expansion mode and a first interface instance exists, sending an interface release request to a platform server; an interface deletion instruction from the platform server is received, the interface deletion instruction comprises a data migration rule, and the data migration rule is used for dynamically selecting a migration mode according to the to-be-processed data size, the data type matching rate and the aging level to achieve fault-adaptive data flow recombination and lossless migration; according to a data migration rule, transferring the to-be-processed data associated with the first interface instance to a second interface instance; and deleting the first interface instance, and sending an operation confirmation message to the platform server. According to the method, the data loss rate is effectively reduced by setting the data migration rule, and the interface switching time delay can be reduced while the resource utilization rate can be improved.