AI data docking method and device and computer equipment

Abstract

The invention relates to an AI data docking method and device and computer equipment. The method comprises the following steps: selecting various to-be-configured components from a component library based on a received specific business demand; the various to-be-configured components comprise a universal component and a customized component; the specific business demand is a business demand related to a department where the user is located; determining a component set template based on the selected to-be-configured components and the pointing relationship among the to-be-configured components; the component set template is a component set architecture which combines and configures a plurality of functional components according to a preset framework so as to meet specific business requirements; and utilizing the component set template to acquire AI data of the data source platform, and performing modularization processing on the acquired AI data and then outputting the processed AI data. By adopting the method, the problems that an existing AI data docking method is tedious in docking process and poor in expansibility, and when the AI data throughput is large, the data docking efficiency is low or the accuracy rate is low can be solved.

Description

Technical Field

[0001] The present application relates to the field of information management technology, and in particular to an AI data docking method, device and computer equipment. Background Art

[0002] With the development of information technology, AI (Artificial Intelligence) data connection plays an important role in promoting business innovation and digital transformation.

[0003] The existing AI data docking method needs to develop corresponding codes for each type of AI data to realize the collection, analysis, storage, etc. of customized AI data. When new AI data needs to be docked, or the structure of the source data to be docked changes, it is necessary to redevelop the corresponding code and update the already online program. Therefore, the existing AI data docking method has a cumbersome docking process and poor scalability. And when the AI ​​data throughput is large, the data docking efficiency or accuracy will be low due to untimely code development or program updates.

[0004] However, for the existing AI data docking methods, the docking process is cumbersome and has poor scalability. In addition, when the AI ​​data throughput is large, the data docking efficiency or accuracy is low. There is still no effective solution. Summary of the invention

[0005] Based on this, it is necessary to provide an AI data docking method, device and computer equipment to address the above technical problems.

[0006] In a first aspect, the present application provides an AI data docking method. The method comprises: Based on the received specific business requirements, multiple components to be configured are selected from the component library; the multiple components to be configured include general components and customized components; the general components are components with general functions that are not related to specific businesses; the customized components are components with specific functions related to specific businesses; the specific business requirements are business requirements related to the department where the user is located; The component set template is determined based on the selected components to be configured and the pointing relationship between the components to be configured; the pointing relationship between the components to be configured represents the flow of data between the components to be configured; the component set template is a component set architecture that combines and configures multiple functional components according to a preset framework to meet specific business needs; the functional component is a component that can realize a general function or a specific function; The component set template is used to collect AI data from the data source platform, and the collected AI data is componentized and then output.

[0007] In one embodiment, determining the component set template based on the selected components to be configured and the pointing relationship between the components to be configured includes: Based on the selected components to be configured, in response to the user's selection operation, determine the key components in the components to be configured; the key components include at least a data acquisition component, a data filtering component and a data conversion component; the data conversion component includes at least a key feature recognition component, an integration component and a data output component; In response to a configuration operation by a user, determining the flow direction of data in each of the key components; Based on the flow direction of data in each of the key components, determining the directional relationship between each of the key components; The component set template is determined based on each of the key components and the pointing relationship between each of the key components.

[0008] In one embodiment, the use of the component set template to collect AI data from a data source platform, and componentizing and outputting the collected AI data includes: Using the data collection component to collect AI data from the data source platform; Using the data filtering component, filtering the collected AI data to obtain filtered data; The data conversion component is used to perform data conversion on the filtered data to obtain the converted data and output it; the data conversion includes at least one of data format normalization conversion and data splitting.

[0009] In one embodiment, the using the data conversion component to perform data conversion on the filtered data to obtain the converted data and output it includes: Utilizing the key feature identification component, extracting key feature values ​​from the filtered data, and identifying the hierarchical relationship between the key feature values; Using the integration component, based on the hierarchical relationship between the key feature values, the extracted key feature values ​​are integrated together to obtain a new message; The data output component is used to convert the new message into a standard output format and then output it.

[0010] In one embodiment, the collecting of AI data from the data source platform using the data collection component includes: Based on the specific business needs, determine the type, scope and initial data collection method of the data to be collected; Determine the data source platform to be collected based on the type and scope of the data to be collected; Using the initial data collection method, collect the data source identifier corresponding to the data source platform to be collected; Based on the collected data source identifier, AI data of the data source platform to be collected is collected.

[0011] In one embodiment, the collecting of the AI ​​data of the data source platform to be collected based on the collected data source identifier includes: Based on the collected data source identifier, determine the collection method corresponding to the data source platform to be collected; Based on the collection method corresponding to the data source platform to be collected, the AI ​​data of the data source platform to be collected is collected.

[0012] In one embodiment, the key components further include: a data filling component, a scene storage component, a data collection component, and a collection storage component; the use of the component set template to collect AI data from the data source platform, and componentizing and outputting the collected AI data, further includes: Utilizing the data filling component, according to the attribute field requirements of the target data area, extracting the data to be filled corresponding to the target data area from the converted data to perform data filling, thereby obtaining filled data; Using the scene storage component, the populated data is stored in a scene-based relational database to obtain scene-based data; the scene-based relational database is a database related to the scene of the data; Using the data collection component, the scenario data is collected and processed to obtain collected data; the collection process includes collecting and processing based on at least one dimension of a point, a collection terminal, and a collection environment; The collection and storage component is used to store the collected data in a preset storage location or distribute it to a preset distribution location; the preset storage location is a business database; the preset distribution location is at least one of a preset file, relational or non-relational database, network connection, kafka message queue, big data platform or data warehouse.

[0013] In one embodiment, the using the data collection component to collect the scenario data to obtain the collected data includes: Acquire setting results; the setting results are obtained by the user configuring the collection conditions of scenario data of different scenarios through the configurable policy engine; the collection conditions include at least one of event frequency, content similarity, and time window; the event frequency includes at least one of alarm frequency and feature frequency; the configurable settings include at least one of rule logic configuration and hierarchical rule configuration; Based on the setting result, the scenario data of different scenarios are aggregated to obtain the aggregated data.

[0014] In a second aspect, the present application also provides an AI data docking device. The device comprises: The component selection module is used to select multiple components to be configured from the component library based on the received specific business requirements; the multiple components to be configured include general components and customized components; the general components are components with general functions that are not related to specific businesses; the customized components are components with specific functions related to specific businesses; the specific business requirements are business requirements related to the department where the user is located; The template determination module is used to determine the component set template based on the selected components to be configured and the pointing relationship between the components to be configured; the pointing relationship between the components to be configured represents the flow of data between the components to be configured; the component set template is a component set architecture that combines and configures multiple functional components according to a preset framework to meet specific business needs; the functional component is a component that can realize a general function or a specific function; And a componentization module is used to collect AI data from the data source platform using the component set template, and to componentize the collected AI data and then output it.

[0015] In a third aspect, the present application further provides a computer device. The computer device includes a memory and a processor, the memory stores a computer program, and the processor implements the AI ​​data docking method described in the first aspect when executing the computer program.

[0016] The above-mentioned AI data docking method, device and computer equipment, by responding to the specific business needs received, combine and configure multiple functional components according to the preset framework, determine the component set template, and use the component set template to collect AI data from the data source platform, and output the collected AI data after component processing. It processes data through componentization. When docking new AI data, it only needs to configure the components corresponding to the new AI data to obtain the component set template that can process the new AI data. The new AI data is processed using the component set template that can process the new AI data. The expansion is simple and the docking process is simple. When the AI ​​data throughput is large, it only needs to be equipped with the corresponding component set template for each AI data, and there is no need to develop or update the code. It solves the problems of the existing AI data docking method, which has a cumbersome docking process, poor scalability, and low data docking efficiency or low accuracy when the AI ​​data throughput is large.

[0017] Details of one or more embodiments of the present application are set forth in the following drawings and description to make other features, objects, and advantages of the present application more readily apparent. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings: Figure 1 A hardware structure block diagram of a terminal of an AI data docking method provided in an embodiment of the present application; Figure 2 A flowchart of an AI data docking method provided in one embodiment of the present application; Figure 3 A first flow chart of a data processing process provided by an embodiment of the present application; Figure 4 A second flow chart of a data processing process provided by an embodiment of the present application; Figure 5 A flowchart of an AI data docking method provided in a preferred embodiment of the present application; Figure 6 This is a structural block diagram of an AI data docking device provided in one embodiment of the present application. DETAILED DESCRIPTION

[0019] In order to more clearly understand the purpose, technical solutions and advantages of the present application, the present application is described and illustrated below in conjunction with the accompanying drawings and embodiments.

[0020] Unless otherwise defined, the technical terms or scientific terms involved in this application shall have the general meaning understood by people with general skills in the technical field to which this application belongs. The words "one", "a", "the", "these" and the like in this application do not indicate a quantitative limitation, and they may be singular or plural. The terms "include", "comprise", "have" and any variants thereof involved in this application are intended to cover non-exclusive inclusions; for example, a process, method and system, product or device comprising a series of steps or modules (units) is not limited to the listed steps or modules (units), but may include unlisted steps or modules (units), or may include other steps or modules (units) inherent to these processes, methods, products or devices. The words "connect", "connected", "coupled" and the like involved in this application are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. The "multiple" involved in this application refers to two or more. "And / or" describes the association relationship of associated objects, indicating that there may be three relationships, for example, "A and / or B" may mean: A exists alone, A and B exist at the same time, and B exists alone. Generally, the character " / " indicates that the objects associated with each other are in an "or" relationship. The terms "first", "second", "third", etc. in this application are only used to distinguish similar objects and do not represent a specific ordering of the objects.

[0021] The method embodiment provided in this embodiment can be executed in a terminal, a computer or a similar computing device. For example, running on a terminal, Figure 1 FIG. 1 is a hardware structure diagram of a terminal of the AI ​​data docking method of this embodiment. Figure 1 As shown, the terminal may include one or more ( Figure 1 Only one is shown in the figure) a processor 102 and a memory 104 for storing data, wherein the processor 102 may include but is not limited to a processing device such as a microprocessor MCU or a programmable logic device FPGA. The above terminal may also include a transmission device 106 and an input and output device 108 for communication functions. It can be understood by those skilled in the art that Figure 1 The structure shown is for illustration only and does not limit the structure of the above terminal. Figure 1 More or fewer components as shown, or with Figure 1 Different configurations shown.

[0022] The memory 104 can be used to store computer programs, for example, software programs and modules of application software, such as the computer program corresponding to the AI ​​data docking method in this embodiment. The processor 102 executes various functional applications and data processing by running the computer program stored in the memory 104, that is, to implement the above method. The memory 104 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 104 may further include a memory remotely arranged relative to the processor 102, and these remote memories may be connected to the terminal via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0023] The transmission device 106 is used to receive or send data via a network. The above network includes a wireless network provided by the communication provider of the terminal. In one example, the transmission device 106 includes a network adapter (Network Interface Controller, referred to as NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission device 106 can be a radio frequency (Radio Frequency, referred to as RF) module, which is used to communicate with the Internet wirelessly.

[0024] The existing AI data docking method requires redeveloping the corresponding code and updating the already online program when docking new AI data. The work is cumbersome and the docking process is long. Moreover, when there are many types of AI docking, the code is difficult to control. When the AI ​​data throughput is large, data reception and processing bottlenecks are prone to occur, which in turn affects the stability of the entire application. Therefore, it is necessary to provide a new AI data docking method to solve the problem of poor stability of the entire application when the AI ​​data throughput is large.

[0025] In this embodiment, an AI data docking method is provided. Figure 2 is a flow chart of the AI ​​data docking method of this embodiment. Figure 2 As shown, the process includes the following steps: Step S210, based on the received specific business requirements, select multiple components to be configured from the component library; the multiple components to be configured include general components and customized components; general components are components with general functions that are not related to specific businesses; customized components are components with specific functions related to specific businesses; specific business requirements are business requirements related to the department where the user is located.

[0026] The departments where the above users are located can be security departments, data management departments, data science and analysis departments, etc., which can realize certain business needs through AI data. The above business needs related to the departments where the users are located can be one or more of video surveillance and intelligent analysis, environmental monitoring and early warning, pedestrian and vehicle flow monitoring, multi-mode identity authentication, digital identity authentication, etc. when the users are in the security department; they can be one or more of data collection and integration, data storage and management, data governance and quality control, data security and compliance, data sharing and collaboration when the users are in the data management department; they can be one or more of data collection and integration, data analysis and model development, cross-departmental collaboration, etc. when the users are in the data science and analysis department.

[0027] Among them, the above-mentioned components to be configured are components that require configuration of data flows between components. The above-mentioned general components may be components with general functions that are not related to specific businesses, such as data filtering components, data cleaning components, data conversion components, data storage components, etc. The above-mentioned customized components may be customized components written by developers according to video business requirements, such as video frame extraction components, target detection components, key feature recognition components, or components related to other specific businesses. The above-mentioned selection of multiple components to be configured from the component library based on the specific business requirements received may be based on the specific business requirements received, and the user visually selects operations on the web page to select multiple components to be configured from the component library. Among them, the above-mentioned component library may include a general component library and a customized component library. The above-mentioned general component library is a component library of various pre-stored general components.

[0028] In this step, by selecting common components and customized components related to specific business needs from the component library, it is convenient to use the selected common components and customized components to determine a component set template that can meet the specific business needs.

[0029] Step S220, determining a component set template based on the selected components to be configured and the pointing relationship between the components to be configured; the pointing relationship between the components to be configured represents the flow of data between the components to be configured; the component set template is a component set architecture that combines and configures multiple functional components according to a preset framework to meet specific business needs; the functional component is a component that can implement a general function or a specific function.

[0030] In this step, the component set template is determined based on the selected components to be configured and the pointing relationships between the components to be configured. The pointing relationships between the components to be configured may be configured in response to the user's configuration operation, and the component set template is determined based on the pointing relationships between the components to be configured and the components to be configured.

[0031] In this step, the above-mentioned component set architecture can be an architectural design method that divides the software system into multiple independent and reusable functional components. Among them, each functional component can encapsulate the corresponding function and can realize a general function or a specific function. In addition, each functional component interacts with other components through a clearly defined interface, and multiple components jointly realize specific business needs. The above-mentioned general functions can be functions with high abstraction and versatility, which are independent of specific business logic and can be reused in multiple systems or scenarios. For example, general functions such as user login, registration, and permission verification. The above-mentioned specific functions can be functions designed for specific business needs, which are closely coupled with specific business logic and rely on specific data structures or business processes. For example, functions such as order placement, payment, delivery, and refund of orders on e-commerce platforms. Among them, a component can be a set of highly correlated codes that can implement a module with specific functions.

[0032] Step S230, using the component set template, collects AI data from the data source platform, and componentizes the collected AI data for output.

[0033] Among them, the above-mentioned data source platform can be a system or architecture for centralized management, processing and distribution of data, providing basic support for data collection, integration, analysis and application, and can be used to manage and maintain various data sources, and provide these data to applications, analysis tools or AI models in a standardized and efficient manner. Data can be collected from various data sources through the data source platform. The above-mentioned data sources may include databases, file systems, APIs (Application Programming Interfaces), IoT devices, http or websocket network connections, kafka or flink and other data sources. Among them, the above-mentioned databases may include relational databases and non-relational databases. The above-mentioned AI data (Artificial Intelligence Data, data related to artificial intelligence) can be one or more of structured data, unstructured data, semi-structured data, time series data, multimodal data, etc. The above-mentioned component processing can be a method of decomposing a complex system or task into multiple independent components, and combining these components to achieve specific functions by defining clear interfaces and interaction logic. Component processing can improve development efficiency. Specifically, repetitive development can be reduced by reusing standardized components. Component combinations can be quickly adjusted as needed to adapt to different business scenarios and increase flexibility. Complex tasks can be broken down into multiple small tasks to reduce complexity and facilitate management and maintenance. On-demand expansion of components can be supported to adapt to the growth of data volume and business, thereby improving the scalability of the system.

[0034] In the above steps S210 to S230, in response to the specific business needs received, multiple functional components are combined and configured according to the preset framework, a component set template is determined, and the component set template is used to collect AI data from the data source platform, and the collected AI data is processed and output after componentization. It processes data through componentization. When docking new AI data, it only needs to configure the components corresponding to the new AI data to obtain the component set template that can process the new AI data. The new AI data is processed using the component set template that can process the new AI data. The expansion is simple and the docking process is simple. When the AI ​​data throughput is large, it only needs to be equipped with the corresponding component set template for each AI data, and there is no need to develop or update the code. This solves the problems of the existing AI data docking method, which has a cumbersome docking process, poor scalability, and low data docking efficiency or accuracy when the AI ​​data throughput is large.

[0035] Specifically, in one embodiment, step S220, based on the selected components to be configured and the pointing relationship between the components to be configured, determines the component set template, including: Step S222, based on the selected components to be configured, in response to the user's selection operation, determine the key components in the components to be configured; the key components include at least a data acquisition component, a data filtering component and a data conversion component; the data conversion component includes at least a key feature recognition component, an integration component and a data output component.

[0036] In this step, the user selection operation can be that the user selects in the visual interface of the web page to determine the key components in the components to be configured. The data acquisition component is used to collect AI data from the data source platform. The data filtering component is used to filter the collected AI data to obtain filtered data. The data conversion component is used to perform data conversion on the filtered data to obtain data converted data. The key feature recognition component can be used to extract the key values ​​of the data and the hierarchical relationship between each key feature value.

[0037] The above integration component is used to integrate the extracted key feature values ​​together, and can change the original message format or form a new message and distribute it to the next component. For example, the data under MotorVehicleObject is extracted to form a new format message to send a motor vehicle parking violation message to the next level component, the data under ImageInfo is extracted to form a scene graph message to the next level component, and the data under ThingObject is extracted to form an item message to the next level component. The next component of the above integration component can be a data output component.

[0038] Among them, the above-mentioned data output component can format the different received data streams into standard output streams and distribute them to one or more locations such as target files, relational or non-relational databases, http or websocket network connections, kafka message queues, big data platforms or data warehouses.

[0039] Step S224, in response to the user's configuration operation, determining the flow of data in each key component.

[0040] The above-mentioned user configuration operation can be that the user configures the visual interface of the web page to determine the flow direction of data in each key component. By determining the flow direction of data in each key component, each component can be connected end to end according to the inflow or outflow of data. Among them, a component can accept the inflow of data from multiple components, and data can flow out to multiple components at the same time.

[0041] Step S226, based on the flow of data in each key component, determine the directional relationship between each key component.

[0042] The pointing relationship between the above-mentioned key components is consistent with the flow direction of data in each key component. When data flows from component A to component B, component A points to component B.

[0043] Step S228, determining a component set template based on each key component and the pointing relationship between each key component.

[0044] This step connects each component into a complete process by determining the pointing relationship between each key component, and clarifies the processing process from data collection to output. It can realize the specific business needs by using the determined component set template.

[0045] Preferably, multiple components are connected into a complete process by defining the interface and interaction logic between components. After the component set template is determined, the entire component process can be encapsulated into a template (component set template), and then the component set template is saved. When necessary, the component set template can be adjusted and deployed through a configuration file or interface. In a specific application scenario, the component set template can be deployed to receive and process data using the component set template, and the processing results, throughput, response time and other performance indicators of each component in the component set template can be monitored, and these indicators can be displayed to the user.

[0046] The specific implementation process is as follows: { "SubscribeNotificationListObject": { "SubscribeNotificationObject": [{ "SubscribeID": "003301020001032023091420574500184", "NotificationID": "003301020001042024062014414501107", "ImageObjectList": { "ImageObject": [{ "MotorVehicleObjectList": { "MotorVehicleObject": [{ "eventSort": 659, "DeviceID": "35010100861321022886", "SubImageList": { "SubImageInfoObject": [{ "Type": "09", "StoragePath": "https: / / xxxxxx / motorVehicle / 35010100861212733.Jpeg", "DeviceID": "35010100861321022886", }] }, "PlateNo": "Unknown", "PlateDescribe": "{\"Score\":0.947833,\"AnalysisType\":32008,\"AreaId\":1,\"EventId\":\"1410\"}" }] }, "ImageInfo": { "EventSort": 659, "Title": "Scene Diagram", "StoragePath": "https: / / xxxxxx / imageInfo / 350101008613210221367.Jpeg", "SecurityLevel": "5"}, "ThingObjectList": { "ThingObject": [{ "eventSort": 659, "DeviceID": "35010100861321022886", "Characteristic": "{\"Score\":0.830342,\"AnalysisType\":32004,\"AlarmDescription\":\"Out-of-store operation\",\"AreaId\":1,\"EventId\":\"1408\"}"}]}]}, "Title": "Universal Application Live Network Subscription" }]}} In the above steps S222 to S228, the key components in the components to be configured are determined based on the selected components to be configured, and the pointing relationship between the key components is determined, and then the component set template is determined based on the key components and the pointing relationship between the key components. The determination of the component set template facilitates the subsequent use of the component set template to collect AI data of the data source platform corresponding to the specific business needs, and the collected AI data is processed into components and then output.

[0047] Figure 3 The first flow chart of the data processing process provided by an embodiment of the present application is as follows: Figure 3 As shown, the data is output in a standard manner after passing through the filtering component, the key feature recognition component, the integration component, and the output component. Among them, the key feature recognition component includes the key feature recognition component 1, the key feature recognition component 2, and the key feature recognition component 3. The integration component includes the integration component 1 and the integration component 2. Among them, the components can be selected manually from the general component library, and can be selected from the custom component library through customized components.

[0048] In addition, in one embodiment, step S230, using a component set template to collect AI data from a data source platform, and componentizing the collected AI data and outputting the componentized data, includes: Step S231, using the data collection component to collect AI data from the data source platform.

[0049] Among them, the above-mentioned data collection component is used to collect data on video AI detection results of different big data capability platforms. It can be based on specific business needs to determine the type, scope and initial data collection method of the data to be collected, and then, based on the type and scope of the data to be collected, determine the data source platform to be collected, use the initial data collection method to collect the data source identifier corresponding to the data source platform to be collected, and based on the collected data source identifier, collect the AI ​​data of the data source platform to be collected.

[0050] Step S232, using the data filtering component to filter the collected AI data to obtain filtered data.

[0051] The above-mentioned use of the data filtering component to filter the collected AI data to obtain the filtered data can be through the data filtering component, setting the filtering conditions, and calling the corresponding data filtering component to filter the multi-source data. When the filtering component filters the data source, the above-mentioned filtering component can be a combination of multiple filtering components to filter information or multiple filtering components can extract preset key values ​​at the same time. The above-mentioned preset key values ​​can be one or more of SubscribeNotificationListObject, SubscribeID, NotificationID, etc.

[0052] Step S233, using the data conversion component, performing data conversion on the filtered data, obtaining the converted data and outputting it; the data conversion includes at least one of data format normalization conversion or data splitting conversion.

[0053] In this step, the above-mentioned data conversion component converts the filtered data to ensure the normalization, consistency or adaptation of the converted data to specific business needs. The above-mentioned data conversion can be a process of processing data to change its format, structure or content. The above-mentioned data format normalization conversion can be to convert the data into a unified format for subsequent processing. The above-mentioned unified format can be specifically set according to the specific situation, and this embodiment does not make specific limitations here. The above-mentioned data splitting can be to split a complex data structure into a simpler target structure for analysis or storage. The above-mentioned target structure can be specifically set according to the specific situation, and this embodiment does not make specific limitations here.

[0054] In the above steps S231 to S233, the AI ​​data of the data source platform is collected by using the data collection component, and the collected AI data is filtered by using the data filtering component to obtain the filtered data, and then the filtered data is converted by using the data conversion component to obtain the converted data and output it. It is convenient for further processing of the data after data collection, filtering and format normalization.

[0055] In one embodiment, step S233, using a data conversion component to perform data conversion on the filtered data, obtain the converted data and output it, includes: Step S2332, using the key feature recognition component, extracting key feature values ​​from the filtered data, and identifying the hierarchical relationship between the key feature values.

[0056] The extraction process of the key feature recognition component can be extracted by multiple key feature recognition components at the same time. The key value of the data can include one or more of ImageObjectList, EventSort, MotorVehicleObjectList, ImageInfo, ThingObjectList, SubImageInfoObject, etc. The ImageObjectList can be a list containing multiple image objects (ImageObject), which can be used to store and manage metadata or information of multiple images. Specifically, the ImageObjectList may contain information such as the path, format, and shooting time of multiple images. The EventSort can be an event type or sorting identifier, which can be used to mark the type or priority of an event so that the system can classify or sort it according to the value. The MotorVehicleObjectList can be a list containing multiple motor vehicle objects (MotorVehicleObject), which is used to store and manage data related to motor vehicles, for example, for storing and managing data such as license plate numbers, vehicle types, and entry times. The ImageInfo can be a collection of image information, which is used to store detailed information of a single image, such as image path, shooting time, resolution, and tags. The ThingObjectList mentioned above may be a list containing multiple general object objects (ThingObject), which is used to store and manage data related to general objects, such as object type, location, status, etc. The SubImageInfoObject mentioned above may be a sub-image information object, which is used to store detailed information of a sub-image, such as the cropping area, label, and associated parent image of the sub-image.

[0057] The hierarchical relationship between the above-mentioned extracted key feature values ​​can be to identify eventSort under MotorVehicleObject, eventSort under ThingObject, and eventSort under ImageInfo. The above-mentioned MotorVehicleObject, ThingObject, and ImageInfo are names of different objects. The above-mentioned eventSort can be a field used to identify event types or sorting, which can be used to classify, sort, or mark data. In MotorVehicleObject, eventSort is usually used to identify event types or sorting rules related to motor vehicles. For example, it may be used to distinguish different vehicle events, such as vehicle entry, exit, violation, etc. In ThingObject, eventSort can be used to identify event types or sorting rules related to general objects. For example, it may be used to distinguish object detection events, state changes, or other object-related operations. The specific meaning needs to be determined according to the actual application scenario and data definition. In ImageInfo, eventSort can be used to identify event types or sorting rules related to images. For example, it may be used to distinguish when an image was taken, the image type (such as base image, subimage), or other events related to image processing.

[0058] Step S2334, using the integration component, based on the hierarchical relationship between the key feature values, the extracted key feature values ​​are integrated together to obtain a new message.

[0059] Step S2336, using the data output component, convert the new message into a standard output format and then output it.

[0060] In the above steps S2332 to S2336, by using the key feature recognition component, the key feature values ​​are extracted from the filtered data, and the hierarchical relationship between the key feature values ​​is identified, and the extracted key feature values ​​are integrated together based on the hierarchical relationship between the key feature values ​​by using the integration component to obtain a new message, and then the new message is converted into a standard output format and then output. It recognizes the key features and integrates the extracted key features together to obtain a new message, which is convenient for subsequent standardized output. It converts the data into a unified format through standardized output, which is convenient for further processing or analysis of the data.

[0061] In addition, in one embodiment, step S231, using a data collection component to collect AI data from a data source platform, includes: Step S2312, based on specific business needs, determine the type, scope and initial data collection method of the data to be collected.

[0062] The above-mentioned initial data collection method may be a method of collecting a data source identifier corresponding to a data source platform to be collected.

[0063] Step S2314, based on the type and scope of the data to be collected, determine the data source platform to be collected.

[0064] Step S2316: using the initial data collection method, collect the data source identifier corresponding to the data source platform to be collected.

[0065] The data source identifier corresponding to the above data source platform may be a name or code used to uniquely identify a data source in the data source platform. The data source identifier corresponding to the above data source platform is usually defined by the user when creating a data source and remains unique in the entire platform.

[0066] Step S2318: Based on the collected data source identifier, collect the AI ​​data of the data source platform to be collected.

[0067] The above-mentioned AI data of the data source platform to be collected based on the collected data source identifier can be determined based on the collected data source identifier, and then, based on the type and characteristics of the data source, a suitable data collection method is selected, and the AI ​​data of the data source platform to be collected is collected using the suitable data collection method. The above-mentioned suitable data collection method can be one or more of a web crawler, sensor collection, database connection, etc.

[0068] In the above steps S2312 to S2318, the type, scope and initial data collection method of the data to be collected are determined based on specific business needs, and then the data source platform to be collected and the data source identifier corresponding to the data source platform to be collected are determined based on the type, scope and initial data collection method of the data to be collected, and then the AI ​​data of the data source platform to be collected is collected based on the collected data source identifier. The collection of AI data of the data source platform to be collected facilitates further processing such as filtering of the collected data.

[0069] Further, in one embodiment, step S2318, based on the collected data source identifier, collects AI data of the data source platform to be collected, including: Step S1, based on the collected data source identifier, determine the collection method corresponding to the data source platform to be collected.

[0070] Step S2: Collect AI data of the data source platform to be collected based on the collection method corresponding to the data source platform to be collected.

[0071] In the above steps S1 to S2, by determining the collection method corresponding to the data source platform to be collected, it is convenient to collect the AI ​​data of the data source platform to be collected based on the collection method corresponding to the data source platform to be collected. By collecting the AI ​​data of the data source platform to be collected, it is convenient to filter the collected data and perform further processing.

[0072] In one embodiment, the key components further include: a data filling component, a scene storage component, a data collection component, and a collection storage component; step S230, using the component set template to collect AI data from the data source platform, and outputting the collected AI data after component processing, further includes: Step S234, using the data filling component, according to the attribute field requirements of the target data area, extracting the data to be filled corresponding to the target data area from the converted data to perform data filling to obtain the filled data.

[0073] In this step, before using the data filling component to extract the data to be filled corresponding to the target data area from the converted data according to the attribute field requirements of the target data area for data filling, it is necessary to determine the attribute field requirements of the target data area. The attribute field requirements of the above-mentioned target data area may include one or more of the user name, contact number, email address, and other business-related attribute fields. The above-mentioned extraction of the data to be filled corresponding to the target data area from the converted data may be to first map the converted data field with the field of the target data area to ensure the consistency of the data, and then use data processing tools such as Pandas to extract the data to be filled corresponding to the target data area according to the field name or index. The above-mentioned data filling may be performed by using a preset filling method to fill the data. Among them, the above-mentioned preset filling method may be one or more of rule-based filling (such as single value filling or formula filling, etc.), model-based filling (such as machine learning model, etc.), and data filling based on data binding technology.

[0074] The data after filling above is unified and complete data content.

[0075] Step S235, using the scene storage component, the filled data is stored in the scene-based relational database to obtain scene-based data; the scene-based relational database is a database related to the scene of the data.

[0076] Step S236, using the data collection component, collects and processes the scenario data to obtain collected data; the collection processing includes collecting and processing based on at least one dimension of the point, the collection terminal and the collection environment.

[0077] The above-mentioned collection processing may be the process of centrally collecting, organizing and storing data scattered in different sources, formats and storage locations. The above-mentioned point may be the location where the data is collected. The above-mentioned collection terminal may refer to the terminal used to collect data (such as a camera, sensor, etc.). The above-mentioned collection environment may be the current environment or application scenario of the collected data. In this step, data from different sources are centrally managed and processed through the data collection component to meet the needs of specific business logic and AI detection goals.

[0078] Step S237, using the collection storage component, stores the collected data to a preset storage location or distributes it to a preset distribution location; the preset storage location is a business database; the preset distribution location is at least one of a preset file, relational or non-relational database, network connection, kafka message queue, big data platform or data warehouse.

[0079] Among them, the above-mentioned business database can be a database system used to support the daily business operations and management of an enterprise or organization, and can be one or more of a relational database, a non-relational database, a hybrid database, a data warehouse, a columnar database, an in-memory database, a time series database, etc.

[0080] The above steps S234 to S237 realize data integration and standard output of different platforms in the field of video AI analysis by filling, storing, and collecting data, which facilitates further processing such as data retrieval, analysis, and report generation.

[0081] Figure 4 A second flow chart of the data processing process provided in an embodiment of the present application is as follows: Figure 4 As shown in the figure, data is collected by the data collection component, transformed by the data conversion component, output through componentized multi-data streams, and output after passing through the data filling component, scene storage component, data collection component, and collection storage component. By listing the processing of data stream 1 and data stream 2, the process from data collection to output is provided.

[0082] In one embodiment, step S236, using a data collection component to collect and process the scenario data to obtain collected data, includes: Step S2362, obtain the setting result; the setting result is obtained by the user through the configurable policy engine, by configuring the collection conditions of scenario data for different scenarios; the collection conditions include at least one condition in event frequency, content similarity, and time window; the event frequency includes at least one of the alarm frequency and feature frequency; the configured settings include at least one of the rule logic configuration and the hierarchical rule configuration.

[0083] This step can eliminate a large number of useless and repeated AI events through a configurable collection strategy, reducing the amount of alarm processing by managers. Through the method of this embodiment, subsequent step expansion and subsequent process continuation can be conveniently performed.

[0084] The above rule logic configuration can be to configure multi-conditional logic in the rule, including AND, OR, and NOT operations, to meet complex aggregation requirements. For example, a rule may require that an event meets the frequency and similarity conditions within a specified time window to trigger alarm merging. The above hierarchical rule configuration can be a hierarchical rule configuration, specifically the configuration of upper-level rules and lower-level rules. For example, the upper-level rule can determine the scope of the aggregation condition (such as a large time window), and the lower-level rule can define specific details (such as thresholds for specific event types).

[0085] Step S2364: Based on the setting result, the scenario data of different scenarios are aggregated to obtain aggregated data.

[0086] In the above steps S2362 to S2364, by obtaining the setting results, the scenario data of different scenarios are aggregated to obtain the aggregated data, which facilitates the subsequent storage of the data.

[0087] In one embodiment, in certain application scenarios, a video application system involves multiple AI detection and analysis platforms or view libraries.

[0088] For example, in the main road detection scenario, the installed cameras are used to detect illegal parking of motor vehicles, and the AI ​​analysis results are pushed to the view library through 1400. However, due to regulatory needs, users need to introduce new cameras installed by other manufacturers to perform road occupation business detection, garbage detection, and illegal parking of motor vehicles. Since the AI ​​analysis results of the new manufacturers are not in accordance with the 1400 standard, the view library cannot receive them.

[0089] Therefore, it is necessary to introduce components to realize componentized processing of data. The following is a detailed description of a low-code componentized processing method for video AI monitoring results provided by an embodiment of the present application in a video surveillance scenario, and the specific steps include: Step S1, data collection, this step mainly completes the data collection work of the video AI detection results of different big data capability platforms, and extracts the data source identifiers of different capability platforms from them; the different big data capability platforms here can be various application systems or databases, etc. Through the data collection function of the system, the data collection and integration of video AI detection results of different big data capability platforms can be quickly realized; at the same time, the scope and method of data collection can be flexibly adjusted according to needs. Through graphical configuration to support relational databases, message middleware, files and other data, the data source can be easily expanded.

[0090] Step S2, data filtering, this step mainly completes the filtering of multi-source data to achieve data cleaning and deduplication operations. By setting filtering conditions and using filtering functions, you can easily filter out the required data and remove redundant and invalid information; at the same time, you can also sort or group the filtered data as needed. You can graphically configure the cleaning logic, or introduce script cleaning when cleaning complex data, reducing the steps of program customization and development and reducing the difficulty of subsequent data operation and maintenance.

[0091] Step S3, data conversion, this step mainly completes the data conversion of the filtered target data set, including data format normalization and data splitting. By setting corresponding conversion rules and templates in the system, data in different formats can be converted into a unified standard format; at the same time, data can be split or merged as needed to meet the business needs of different management departments of users. This step can be used to graphically configure the conversion logic with low code, customize the program development according to the needs, and optimize the program version management.

[0092] Step S4, data filling, this step mainly completes the business data filling of the converted multi-source data. According to the attribute field of the target data in the area to be filled, the data to be filled is obtained from the business data, and the data to be filled is filled into the target data area according to the field. At the same time, by setting corresponding filling rules and templates in the system, missing or erroneous data can be replaced with correct values; at the same time, the data can be formatted and standardized as needed to meet the business needs of different management departments of users. Through this step, the convenience of subsequent data use and statistics can be greatly reduced, and the source http or database of external business data can be graphically configured.

[0093] Step S5, scene storage, this step mainly completes the storage of the filled data into the scene-based relational database. It is an important link between general and customized. By configuring the corresponding storage strategy and database connection information and other parameters in the system, the processed data can be stored in the specified database or other storage medium; at the same time, the storage method and location can be adjusted as needed to meet the business needs of different management departments of users. The scene storage step is a key step in the big data monitoring system to perform scene-based, structured and multi-dimensional storage of event data in different business scenarios. Through scene-based storage, the system can realize personalized data organization for different business needs, making subsequent data analysis and query more targeted and accurate. This step uses a flexible scene storage component, combined with business tags and scene-based rule libraries, to store data from different scenarios in a structured relational database for subsequent data retrieval, analysis and report generation. The following are detailed storage technology points and logical designs for various scenarios.

[0094] In the personnel entry and exit scenario, the system records the data of each person who enters and leaves the monitoring area. The scene storage step uses the basic information of the personnel (such as ID number, image features, etc.), entry and exit time and location identification to achieve efficient storage of personnel entry and exit. The entry and exit records are stored in chronological order, and a time series index is established to support efficient query by time range retrieval.

[0095] The vehicle entry and exit scenario involves the storage and management of vehicles entering and exiting the monitoring area. The system implements scenario-based storage of vehicle data through license plate recognition, vehicle feature data (such as vehicle model, color), and time and location records. The vehicle data is stored in order of entry and exit identification and time, and is associated with the location table to provide support for vehicle flow analysis and entry and exit anomaly detection.

[0096] Key vehicle scenarios: For key vehicles, the system needs to provide key monitoring and scenario storage for quick retrieval and correlation analysis. The key vehicle data is stored separately in the key table and associated with the vehicle entry and exit record table to provide effective support for future comprehensive analysis.

[0097] In the key personnel scenario, the storage of key personnel data requires special permission control and confidentiality measures. The system needs to record the activity trajectory of key personnel and provide an alarm mechanism when necessary. On the basis of the ordinary personnel storage table, the key personnel data is stored in an encrypted database table, and access permission management is added to meet data security requirements.

[0098] The scenario of personnel settlement point analysis is to determine the possible settlement points by clustering the individual's movement trajectory. The system aggregates the data of different settlement points through database aggregation operations, and forms frequent patterns, which is convenient for extracting the analysis of frequent settlement locations.

[0099] The multi-unit frequent occurrence scenario is used to detect the situation where the same person frequently appears in multiple monitoring units. This scenario records the activity frequency of the same person in multiple locations, determines whether there is abnormal behavior, stores the frequent occurrence records in the abnormal behavior table, and associates them with the location table to implement activity statistics and alarm mechanisms between units.

[0100] In the case of a sudden increase in the number of people, the system needs to monitor the changes in the number of people in the monitoring area to detect whether a large number of people are gathering. The system records the sudden increase event as an alarm and stores the relevant data. It is stored in the sudden increase alarm table and associated with the location table and personnel data table for event tracing.

[0101] The checkpoint traffic scenario is mainly used to count and analyze the traffic at the checkpoint. The system provides support for traffic analysis and abnormal warning by counting the traffic of people and vehicles at the checkpoint in different time periods. The checkpoint traffic is stored in the traffic statistics table and indexed by time and location to facilitate subsequent trend analysis.

[0102] The data collection step mainly completes the collection of data used for scenarios by configuring corresponding business scripts in the system, such as points, collection terminals, and collection environments, and stores them. It meets the alarm intensity requirements of different AI event alarms in different management departments of users. The data collection step is a key process for merging high-frequency, repeated or redundant data in the AI ​​analysis system. It can optimize data traffic, reduce system resource usage, and improve alarm processing efficiency. The purpose of data collection is to dynamically filter and merge duplicate data through the policy engine in large-scale data monitoring systems, reduce the burden of manual management, and ensure high-quality event push. To achieve this goal, the data collection step includes a multi-level policy system, a complex similarity algorithm, a dynamic configuration mechanism, and an optimized batch processing architecture.

[0103] The configurable policy engine is the core of data collection. The system uses the policy engine to flexibly configure the alarm strategy, allowing operators to configure multi-dimensional conditions such as event frequency, content similarity, time window, etc. for different scenarios. The features of configuration include: multi-dimensional parameterization, rule logic configuration, and hierarchical rule configuration.

[0104] Among them, the above-mentioned multi-dimensional parameterization mainly provides parameterized configuration in terms of time window, event similarity, alarm frequency, feature frequency, etc., and allows flexible adjustment of thresholds according to different business needs. The time window is the basic concept of data collection, which is used to determine whether similar events occurring within a specific period of time should be merged for processing. The configuration of the time window can be a fixed time window or a dynamically adjusted sliding time window. Fixed time window: The system detects and merges events within a predefined time range, such as merging the same event type within 60 minutes into one alarm. Sliding time window: Dynamically adjust the time window and slide it automatically according to the frequency of the event. For example, if the system detects a sharp increase in the frequency of events within 10 minutes, the time window is shortened to improve the real-time performance of alarm processing.

[0105] The complex formula for the time window is: Among them, T start and T end are the start and end time of the time window, and ΔT is the frequency increment condition. When the event or feature frequency increases, the sliding window will automatically decrease to improve the collection efficiency, such as when the number of people identified by the crowd gathering detection results returned for analysis increases in a short period of time. The alarm frequency is a key indicator for determining whether the same type of events are too frequent within a certain time window. The frequency threshold is dynamic and will be adjusted in real time according to the system load and event changes to cope with the impact of high concurrent alarms.

[0106] The event frequency formula is as follows: Where F is the alarm frequency, N is the number of events that occurred in the time window, and T is the length of the time window. When F exceeds the set frequency threshold, the merging operation will be triggered, such as when the facial features, device features, and personnel location cross features returned by the knife and weapon detection results exceed the frequency threshold. In data aggregation, content similarity is used to identify events with similar features. Similarity calculations are based on key fields of events, such as time, capture location, event description, etc. Complex similarity algorithms combine technologies such as feature weighting, vectorized representation, and capture location distance calculation.

[0107] The calculation formula is: Where S is the event similarity, ranging from 0 to 1. A and B represent the feature sets of two events, respectively, and |A∩B| represents the number of common features. α, β, and γ are the weighting coefficients of features, capture location, and time, respectively. geo Indicates the difference in capture position, d time represents the time difference, d maxis the difference normalization parameter. S exceeds the predetermined similarity threshold S threshold , the system will consider the events to be of the same type and merge them.

[0108] Combine into an instruction diagram, visualize the combined operation on the web page, select the filtering component, key feature recognition component, integration component, and output component. Each component is connected end to end according to the data inflow / outflow. One component can accept data inflow from multiple components, and data can flow out to multiple components at the same time.

[0109] The present embodiment is described and illustrated below through preferred embodiments.

[0110] Figure 5 : is a flow chart of an AI data docking method provided by a preferred embodiment of the present application. Figure 5 As shown, the AI ​​data docking method includes the following steps: Step S501, based on the received specific business requirements, select multiple components to be configured from the component library; Step S502, based on the selected components to be configured and the pointing relationship between the components to be configured, determine a component set template; the component set template includes a data acquisition component, a data filtering component, a data conversion component, a data filling component, a scene storage component, a data collection component and a collection storage component; Step S503, using the data collection component to collect AI data from the data source platform; Step S504, using a data filtering component to filter the collected AI data to obtain filtered data; Step S505, using the data conversion component to perform data conversion on the filtered data, obtain the converted data and output it; Step S506, using the data filling component, according to the attribute field requirements of the target data area, extracting the data to be filled corresponding to the target data area from the converted data to perform data filling, and obtain the filled data; Step S507, using the scene storage component, storing the filled data in the scene-based relational database to obtain scene-based data; Step S508, using the data collection component to collect and process the scenario data to obtain collected data; the collection processing includes collecting and processing based on at least one dimension of the point, the collection terminal and the collection environment; Step S509: using the collection storage component, the collected data is stored in a preset storage location or distributed to a preset distribution location.

[0111] In the above steps S501 to S509, in response to the specific business needs received, multiple functional components are combined and configured according to the preset framework, a component set template is determined, and the component set template is used to collect AI data from the data source platform, and the collected AI data is processed and output after componentization. It processes data through componentization. When docking new AI data, it only needs to configure the components corresponding to the new AI data to obtain the component set template that can process the new AI data. The new AI data is processed using the component set template that can process the new AI data. The expansion is simple and the docking process is simple. When the AI ​​data throughput is large, it only needs to equip each AI data with the corresponding component set template, and there is no need to develop or update the code. This solves the problems of the existing AI data docking method, which has a cumbersome docking process, poor scalability, and low data docking efficiency or accuracy when the AI ​​data throughput is large.

[0112] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0113] Based on the same inventive concept, an AI data docking device is also provided in this embodiment, which is used to implement the above-mentioned embodiments and preferred implementation modes, and will not be repeated hereafter. The terms "module", "unit", "subunit", etc. used below can implement a combination of software and / or hardware of predetermined functions. Although the devices described in the following embodiments are preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceived.

[0114] In one embodiment, Figure 6 is a structural block diagram of an AI data docking device provided in an embodiment of the present application, such as Figure 6 As shown, the AI ​​data docking device includes: The component selection module 62 is used to select a plurality of components to be configured from the component library based on the received specific business requirements; the plurality of components to be configured include general components and customized components; the general components are components with general functions that are not related to specific businesses; the customized components are components with specific functions related to specific businesses; the specific business requirements are business requirements related to the department where the user is located; The template determination module 64 is used to determine the component set template based on the selected components to be configured and the pointing relationship between the components to be configured; the pointing relationship between the components to be configured represents the flow of data between the components to be configured; the component set template is a component set architecture that combines and configures multiple functional components according to a preset framework to meet specific business needs; the functional component is a component that can realize a general function or a specific function; And the componentization module 66 is used to collect AI data from the data source platform using the component set template, and to componentize the collected AI data and then output it.

[0115] The above-mentioned AI data docking device, by responding to the specific business needs received, combines and configures multiple functional components according to the preset framework, determines the component set template, and uses the component set template to collect AI data from the data source platform, and outputs the collected AI data after component processing. It processes data through componentization. When docking new AI data, it only needs to configure the components corresponding to the new AI data to obtain the component set template that can process the new AI data. The new AI data is processed using the component set template that can process the new AI data. The expansion is simple and the docking process is simple. When the AI ​​data throughput is large, it only needs to equip each AI data with the corresponding component set template, and there is no need to develop or update the code. It solves the problems of the existing AI data docking method, the docking process is cumbersome, the scalability is poor, and when the AI ​​data throughput is large, the data docking efficiency or accuracy is low.

[0116] It should be noted that the above modules can be functional modules or program modules, and can be implemented by software or hardware. For modules implemented by hardware, the above modules can be located in the same processor; or the above modules can be located in different processors in any combination.

[0117] In one embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, any one of the AI ​​data docking methods in the above embodiments is implemented.

[0118] In one embodiment, a method for docking video AI monitoring data is provided, including: utilizing any one of the AI ​​data docking methods in the above embodiments to componentize and output the collected video AI detection data.

[0119] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but are not limited to this.

[0120] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0121] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. An AI data docking method, characterized in that: The method comprises: Based on the received specific business requirements, multiple components to be configured are selected from the component library; the multiple components to be configured include general components and customized components; the general components are components with general functions that are not related to specific businesses; the customized components are components with specific functions related to specific businesses; the specific business requirements are business requirements related to the department where the user is located; The component set template is determined based on the selected components to be configured and the pointing relationship between the components to be configured; the pointing relationship between the components to be configured represents the flow of data between the components to be configured; the component set template is a component set architecture that combines and configures multiple functional components according to a preset framework to meet specific business needs; the functional component is a component that can realize a general function or a specific function; The component set template is used to collect AI data from the data source platform, and the collected AI data is componentized and then output.

2. The AI ​​data docking method according to claim 1, characterized in that: The determining the component set template based on the selected components to be configured and the pointing relationship between the components to be configured includes: Based on the selected components to be configured, in response to the user's selection operation, determine the key components in the components to be configured; the key components include at least a data acquisition component, a data filtering component and a data conversion component; the data conversion component includes at least a key feature recognition component, an integration component and a data output component; In response to a configuration operation by a user, determining the flow direction of data in each of the key components; Based on the flow direction of data in each of the key components, determining the directional relationship between each of the key components; The component set template is determined based on each of the key components and the pointing relationship between each of the key components.

3. The AI ​​data docking method according to claim 2, characterized in that: The using of the component set template to collect AI data from the data source platform and componentizing the collected AI data for output includes: Using the data collection component to collect AI data from the data source platform; Using the data filtering component, filtering the collected AI data to obtain filtered data; The data conversion component is used to perform data conversion on the filtered data to obtain the converted data and output it; the data conversion includes at least one of data format normalization conversion and data splitting.

4. The AI ​​data docking method according to claim 3, characterized in that: The step of using the data conversion component to perform data conversion on the filtered data, obtaining the converted data and outputting the converted data comprises: Utilizing the key feature identification component, extracting key feature values ​​from the filtered data, and identifying the hierarchical relationship between the key feature values; Using the integration component, based on the hierarchical relationship between the key feature values, the extracted key feature values ​​are integrated together to obtain a new message; The data output component is used to convert the new message into a standard output format and then output it.

5. The AI ​​data docking method according to claim 3, characterized in that: The using the data collection component to collect AI data from the data source platform includes: Based on the specific business needs, determine the type, scope and initial data collection method of the data to be collected; Determine the data source platform to be collected based on the type and scope of the data to be collected; Using the initial data collection method, collect the data source identifier corresponding to the data source platform to be collected; Based on the collected data source identifier, AI data of the data source platform to be collected is collected.

6. The AI ​​data docking method according to claim 5, characterized in that: The collecting of the AI ​​data of the data source platform to be collected based on the collected data source identifier includes: Based on the collected data source identifier, determine the collection method corresponding to the data source platform to be collected; Based on the collection method corresponding to the data source platform to be collected, the AI ​​data of the data source platform to be collected is collected.

7. The AI ​​data docking method according to claim 3, characterized in that: The key components also include: a data filling component, a scene storage component, a data collection component and a collection storage component; the component set template is used to collect AI data from the data source platform, and the collected AI data is processed and output after componentization, and also includes: Utilizing the data filling component, according to the attribute field requirements of the target data area, extracting the data to be filled corresponding to the target data area from the converted data to perform data filling, thereby obtaining filled data; Using the scene storage component, the populated data is stored in a scene-based relational database to obtain scene-based data; the scene-based relational database is a database related to the scene of the data; Using the data collection component, the scenario data is collected and processed to obtain collected data; the collection process includes collecting and processing based on at least one dimension of a point, a collection terminal, and a collection environment; The collection and storage component is used to store the collected data in a preset storage location or distribute it to a preset distribution location; the preset storage location is a business database; the preset distribution location is at least one of a preset file, relational or non-relational database, network connection, kafka message queue, big data platform or data warehouse.

8. The AI ​​data docking method according to claim 7, characterized in that: The utilizing the data collection component to collect the scenario data to obtain the collected data includes: Acquire setting results; the setting results are obtained by the user configuring the collection conditions of scenario data of different scenarios through the configurable policy engine; the collection conditions include at least one of event frequency, content similarity, and time window; the event frequency includes at least one of alarm frequency and feature frequency; the configurable settings include at least one of rule logic configuration and hierarchical rule configuration; Based on the setting result, the scenario data of different scenarios are aggregated to obtain the aggregated data.

9. An AI data docking device, characterized in that: The device comprises: The component selection module is used to select multiple components to be configured from the component library based on the received specific business requirements; the multiple components to be configured include general components and customized components; the general components are components with general functions that are not related to specific businesses; the customized components are components with specific functions related to specific businesses; the specific business requirements are business requirements related to the department where the user is located; The template determination module is used to determine the component set template based on the selected components to be configured and the pointing relationship between the components to be configured; the pointing relationship between the components to be configured represents the flow of data between the components to be configured; the component set template is a component set architecture that combines and configures multiple functional components according to a preset framework to meet specific business needs; the functional component is a component that can realize a general function or a specific function; And a componentization module is used to collect AI data from the data source platform using the component set template, and to componentize the collected AI data and then output it.

10. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, the steps of the AI ​​data docking method according to any one of claims 1 to 8 are implemented.

Images