CN111158651A - Method, device and system for constructing artificial intelligence application - Google Patents

Method, device and system for constructing artificial intelligence application Download PDF

Info

Publication number
CN111158651A
CN111158651A CN201911364703.1A CN201911364703A CN111158651A CN 111158651 A CN111158651 A CN 111158651A CN 201911364703 A CN201911364703 A CN 201911364703A CN 111158651 A CN111158651 A CN 111158651A
Authority
CN
China
Prior art keywords
information
application
component
configuration
file
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN201911364703.1A
Other languages
Chinese (zh)
Inventor
李瀚�
周凯
郝玥
孙越
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
4Paradigm Beijing Technology Co Ltd
Original Assignee
4Paradigm Beijing Technology Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 4Paradigm Beijing Technology Co Ltd filed Critical 4Paradigm Beijing Technology Co Ltd
Priority to CN201911364703.1A priority Critical patent/CN111158651A/en
Publication of CN111158651A publication Critical patent/CN111158651A/en
Priority to PCT/CN2020/139466 priority patent/WO2021129812A1/en
Pending legal-status Critical Current

Links

Images

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F8/00Arrangements for software engineering
    • G06F8/20Software design
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F8/00Arrangements for software engineering
    • G06F8/30Creation or generation of source code
    • G06F8/33Intelligent editors
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F8/00Arrangements for software engineering
    • G06F8/30Creation or generation of source code
    • G06F8/34Graphical or visual programming

Landscapes

  • Engineering & Computer Science (AREA)
  • Software Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Stored Programmes (AREA)

Abstract

The invention provides a method, a device and a system for constructing artificial intelligence application, wherein the method comprises the following steps: acquiring application configuration information configured for the application, wherein the application configuration information comprises scheme information for realizing an artificial intelligence scheme of the application; and generating an executable file of the application according to the application configuration information and a preset template file.

Description

Method, device and system for constructing artificial intelligence application
Technical Field
The present invention relates to the technical field of artificial intelligence, and more particularly, to a method of constructing an artificial intelligence application, an apparatus for constructing an artificial intelligence application, a system comprising at least one computing device and at least one storage device, a computer-readable storage medium, and a system for running an artificial intelligence application.
Background
The current Artificial Intelligence (AI) application construction is that experts customize according to scenario problems by means of their own experiences to complete the construction of an application service, which makes the construction of an AI application at least include: the method comprises the steps of designing data flow, constructing an offline model, constructing an online service and the like, wherein the parts are generally formed by combining a plurality of systems, and a large amount of information such as rules, process definitions, environments, data and the like related to AI application construction is exposed outside and needs to be processed by a user when the application is constructed, so that the existing application construction is difficult to process and method standardization, and precipitation and migration cannot be carried out, so that any user needs to construct an AI application, a large amount of manpower and time are needed to solve similar problems, efficient copying is difficult, and the efficiency is low.
Disclosure of Invention
The embodiment of the invention provides a new technical scheme for constructing artificial intelligence application.
According to a first aspect of the invention, there is provided a method of building an artificial intelligence application, comprising:
acquiring application configuration information configured for the application, wherein the application configuration information comprises scheme information for realizing an artificial intelligence scheme of the application;
and generating an executable file of the application according to the application configuration information and a preset template file.
Optionally, the scenario information includes task content information of the scenario.
Optionally, the task content information includes content information of each task item and association information between the task items.
Optionally, the task content information includes functional components used by the solution and operation information of each functional component;
the application configuration information further includes component configuration information configured for each of the functional components, respectively.
Optionally, the operation information includes at least one of operation stage information of the corresponding functional component in the scheme, initialization information of the corresponding functional component, and operation condition information of the corresponding functional component.
Optionally, generating the executable file of the application according to the application configuration information and a preset template file includes:
generating a first file defining the scheme according to the scheme information and the scheme template file;
aiming at each functional component, generating a second file defining the corresponding functional component according to the component configuration information of the corresponding functional component and the component template file of the corresponding functional component;
and packaging the first file and each second file to generate the executable file.
Optionally, the component template file includes a basic information configuration item corresponding to the component, a service information configuration item corresponding to the component, and an operation and maintenance information configuration item corresponding to the component, where the service information includes operation rule information corresponding to the component.
Optionally, the basic information includes at least one of a version of the corresponding functional component, a creator of the corresponding functional component, a unique identifier of the corresponding functional component, a name of the corresponding functional component, and a data set in which the corresponding functional component is located.
Optionally, the service information further includes at least one of context declaration information and message declaration information.
Optionally, the component template file further includes configuration interface information.
Optionally, the method further comprises:
analyzing a component template file corresponding to the functional component in response to the operation of configuring the functional component;
and under the condition that the component template file defines configuration interface information, rendering a configuration interface corresponding to the functional component according to the configuration interface information.
Optionally, the method further comprises:
and responding to the operation of configuring the functional component, and loading the file content of the component template file of the corresponding functional component for configuration.
Optionally, the method further comprises:
and providing a configuration entry interface in response to the operation of configuring the functional component, wherein the configuration entry interface comprises a first configuration entry and a second configuration entry, the first configuration entry is an entry for selecting to load the configuration interface defined by the component template file of the corresponding functional component for configuration, and the second configuration entry is an entry for selecting to load the file content of the component template file of the corresponding functional component for configuration.
Optionally, the method further comprises:
in response to an operation of defining the task content information, loading list items of the functional components in the functional component library;
and detecting the editing operation of the list items to obtain the functional components used by the scheme and the running information of each functional component.
Optionally, the editing operation comprises a selection operation and a shift operation.
Optionally, the method further comprises:
and providing an operation interface for triggering the operation of configuring the corresponding functional component on the list item.
Optionally, the functional components used by the scheme include any one or more of the following: the system comprises a data preprocessing component, a data splitting component, a data merging component, a feature extraction component, a model training component, a model evaluation component, a self-learning component, a real-time estimation component and a batch estimation component.
Optionally, the method further comprises the step of generating the functional component, including:
providing a functional component development template;
acquiring functional component information defined according to the functional component development template and information related to configuration functional components, wherein the functional component information comprises function implementation information of corresponding functional components;
packing the file containing the functional component information to generate a corresponding functional component;
and generating a component configuration template corresponding to the functional component according to the information related to the configuration functional component.
Optionally, the method further comprises:
and responding to the operation of issuing the corresponding functional component, uploading the corresponding functional component and the component configuration template of the corresponding functional component to a component library, and registering the corresponding functional component in the component library.
Optionally, the template file includes a scheme template file, where the scheme template file includes a basic information configuration item of a scheme, a service information configuration item of the scheme, and an operation and maintenance information configuration item of the scheme, where the service information includes the task content information.
Optionally, the task content information includes functional components used by the solution and operation information of each functional component; the service information also includes common configuration information of each of the functional components.
Optionally, the common configuration information includes at least one of a common data dictionary, public system parameters, and common initialization information.
Optionally, the service information further includes at least one of entry information of the scenario and interface information of the scenario.
Optionally, the operation and maintenance information includes operation mode information and operation resource information.
Optionally, the solution information further includes basic information of the solution, where the basic information includes at least one of a solution name, a keyword of a solution applicable field, and solution description information.
Optionally, the obtaining application configuration information configured for the application includes:
and acquiring the application configuration information input through the provided configuration interface.
Optionally, the configuration interface includes at least one of a configuration item providing a check option and a configuration item providing an input box.
Optionally, the method further comprises:
searching whether the template file has configuration information corresponding to the configuration items of the configuration interface;
in the case of having the configuration information, providing the corresponding configuration information in the corresponding configuration item of the configuration interface.
Optionally, the template file has defined corresponding configuration information for at least some of the contained configuration items.
Optionally, the generating an executable file of the application according to the application configuration information and a preset template file includes:
and updating the template file according to the application configuration information to generate an executable file of the application.
Optionally, after generating the executable file of the application, the method further includes:
loading an engine matched with the template file to run the executable file;
and in the case of a run through, publishing the executable file for downloading.
Optionally, after generating the executable file of the application, the method further includes:
and issuing the executable file to an application providing device for downloading.
According to a second aspect of the present invention, there is also provided an apparatus for building artificial intelligence applications, comprising:
the application configuration module is used for acquiring application configuration information configured for the application, wherein the application configuration information comprises scheme information for realizing an artificial intelligence scheme of the application; and the number of the first and second groups,
and the file generation module is used for generating the executable file of the application according to the application configuration information and a preset template file.
Optionally, the scenario information includes task content information of the scenario.
Optionally, the task content information includes content information of each task item and association information between the task items.
Optionally, the task content information includes functional components used by the solution and operation information of each functional component;
the application configuration information further includes component configuration information configured for each of the functional components, respectively.
Optionally, the operation information includes at least one of operation stage information of the corresponding functional component in the scheme, initialization information of the corresponding functional component, and operation condition information of the corresponding functional component.
Optionally, when the file generation module generates the executable file of the application according to the application configuration information and a preset template file, the file generation module is configured to:
generating a first file defining the scheme according to the scheme information and the scheme template file;
aiming at each functional component, generating a second file defining the corresponding functional component according to the component configuration information of the corresponding functional component and the component template file of the corresponding functional component; and the number of the first and second groups,
and packaging the first file and each second file to generate the executable file.
Optionally, the component template file includes a basic information configuration item corresponding to the component, a service information configuration item corresponding to the component, and an operation and maintenance information configuration item corresponding to the component, where the service information includes operation rule information corresponding to the component.
Optionally, the basic information includes at least one of a version of the corresponding functional component, a creator of the corresponding functional component, a unique identifier of the corresponding functional component, a name of the corresponding functional component, and a data set in which the corresponding functional component is located.
Optionally, the service information further includes at least one of context declaration information and message declaration information.
Optionally, the component template file further includes configuration interface information.
Optionally, the application configuration module is further configured to:
analyzing a component template file corresponding to the functional component in response to the operation of configuring the functional component; and the number of the first and second groups,
and under the condition that the component template file defines configuration interface information, rendering a configuration interface corresponding to the functional component according to the configuration interface information.
Optionally, the application configuration module is further configured to: and responding to the operation of configuring the functional component, and loading the file content of the component template file of the corresponding functional component for configuration.
Optionally, the application configuration module is further configured to:
and providing a configuration entry interface in response to the operation of configuring the functional component, wherein the configuration entry interface comprises a first configuration entry and a second configuration entry, the first configuration entry is an entry for selecting to load the configuration interface defined by the component template file of the corresponding functional component for configuration, and the second configuration entry is an entry for selecting to load the file content of the component template file of the corresponding functional component for configuration.
Optionally, the application configuration module is further configured to:
in response to an operation of defining the task content information, loading list items of the functional components in the functional component library; and the number of the first and second groups,
and detecting the editing operation of the list item, and acquiring the functional components used by the scheme and the running information of each functional component according to the editing operation.
Optionally, the editing operation comprises a selection operation and a shift operation.
Optionally, the application configuration module is further configured to:
and providing an operation interface for triggering the operation of configuring the corresponding functional component on the list item.
Optionally, the functional components used by the scheme include any one or more of the following: the system comprises a data preprocessing component, a data splitting component, a data merging component, a feature extraction component, a model training component, a model evaluation component, a self-learning component, a real-time estimation component and a batch estimation component.
Optionally, the apparatus further comprises a component generation module, configured to:
providing a functional component development template;
acquiring functional component information defined according to the functional component development template and information related to configuration functional components, wherein the functional component information comprises function implementation information of corresponding functional components;
packing the file containing the functional component information to generate a corresponding functional component; and the number of the first and second groups,
and generating a component configuration template corresponding to the functional component according to the information about the configured functional component.
Optionally, the apparatus further comprises a component publishing module, configured to:
and responding to the operation of issuing the corresponding functional component, uploading the corresponding functional component and the component configuration template of the corresponding functional component to a component library, and registering the corresponding functional component in the component library.
Optionally, the template file includes a scheme template file, where the scheme template file includes a basic information configuration item of a scheme, a service information configuration item of the scheme, and an operation and maintenance information configuration item of the scheme, where the service information includes the task content information.
Optionally, the task content information includes functional components used by the solution and operation information of each functional component; the service information also includes common configuration information of each of the functional components.
Optionally, the common configuration information includes at least one of a common data dictionary, public system parameters, and common initialization information.
Optionally, the service information further includes at least one of entry information of the scenario and interface information of the scenario.
Optionally, the operation and maintenance information includes operation mode information and operation resource information.
Optionally, the solution information further includes basic information of the solution, where the basic information includes at least one of a solution name, a keyword of a solution applicable field, and solution description information.
Optionally, the application configuration module is configured to, when acquiring the application configuration information configured for the application, acquire the application configuration information input through a provided configuration interface.
Optionally, the configuration interface includes at least one of a configuration item providing a check option and a configuration item providing an input box.
Optionally, the application configuration module is further configured to:
searching whether the template file has configuration information corresponding to the configuration items of the configuration interface; and, in the event of having the configuration information, providing the corresponding configuration information in the corresponding configuration item of the configuration interface.
Optionally, the template file has defined corresponding configuration information for at least some of the contained configuration items.
Optionally, when the file generation module generates the executable file of the application according to the application configuration information and a preset template file, the file generation module is configured to: and updating the template file according to the application configuration information to generate an executable file of the application.
Optionally, the apparatus further includes an application publishing module, where the application publishing module is configured to load an engine adapted to the template file to run the executable file after the file generating module generates the executable file of the application; and, in the event of a pass through, publishing the executable file for download.
Optionally, the apparatus further includes an application publishing module, where the application publishing module is configured to publish the executable file to an application providing apparatus for downloading after the file generating module generates the executable file of the application.
According to a third aspect of the present invention, there is also provided an apparatus comprising at least one computing device and at least one storage device, wherein the at least one storage device is adapted to store instructions for controlling the at least one computing device to perform the method according to the first aspect of the present invention; alternatively, the apparatus implements the apparatus according to the second aspect of the present invention by the computing means and the storage means.
According to a fourth aspect of the present invention, there is also provided a computer readable storage medium having stored thereon a computer program which, when executed by a processor, implements the method according to the first aspect of the present invention.
According to a fifth aspect of the present invention, there is also provided a system for running an artificial intelligence scheme, comprising an engine means and the means for constructing an artificial intelligence application according to the second aspect of the present invention, wherein the means for constructing an artificial intelligence application is configured to generate an executable file of an application; the engine device is used for running the application according to the executable file.
Optionally, the system further comprises an application providing device, and the device for building the artificial intelligence application is further configured to publish the generated executable file to the application providing device for downloading;
the engine means is for running the application in accordance with the executable file downloaded from the application providing means.
In the embodiment of the present invention, a template file for constructing an application is preset, and configuration items of the template file, which need to be configured individually, are exposed to the outside, and after a user provides corresponding individualized configuration information for the exposed configuration items, the method of this embodiment may generate an executable file that can be analyzed by an engine device according to the acquired configuration information and the template file, so that the engine device can run a corresponding artificial intelligence application through the executable file. In this embodiment, since the configuration items exposed by the template file only need to be configured by one client for constructing the AI application, and the template file defines configuration information that can be shared by different artificial intelligence applications, the embodiment realizes standardization of the process for constructing the AI application, and can realize precipitation and migration of the AI application construction by storing the configured template file, thereby greatly reducing the difficulty in constructing one AI application and effectively improving the construction efficiency.
Other features of the present invention and advantages thereof will become apparent from the following detailed description of exemplary embodiments thereof, which proceeds with reference to the accompanying drawings.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description, serve to explain the principles of the invention.
Fig. 1 is a block diagram showing an example of a hardware configuration of an electronic device that can be used to implement an embodiment of the present invention;
FIG. 2 is a flow diagram of a method of building an artificial intelligence application, according to one embodiment;
FIG. 3 is an interface schematic of a configuration portal interface according to one embodiment;
FIG. 4 is a functional block diagram of an apparatus to build an artificial intelligence application, according to one embodiment;
FIG. 5 is a schematic diagram of a component architecture of a system for building artificial intelligence applications, according to one embodiment;
FIG. 6 is a functional block diagram of an engine arrangement according to one embodiment;
FIG. 7 is a flow diagram illustrating operation of an engine arrangement to run an online predictive application, according to an example.
Detailed Description
Various exemplary embodiments of the present invention will now be described in detail with reference to the accompanying drawings. It should be noted that: the relative arrangement of the components and steps, the numerical expressions and numerical values set forth in these embodiments do not limit the scope of the present invention unless specifically stated otherwise.
The following description of at least one exemplary embodiment is merely illustrative in nature and is in no way intended to limit the invention, its application, or uses.
Techniques, methods, and apparatus known to those of ordinary skill in the relevant art may not be discussed in detail but are intended to be part of the specification where appropriate.
In all examples shown and discussed herein, any particular value should be construed as merely illustrative, and not limiting. Thus, other examples of the exemplary embodiments may have different values.
It should be noted that: like reference numbers and letters refer to like items in the following figures, and thus, once an item is defined in one figure, further discussion thereof is not required in subsequent figures.
Various embodiments and examples according to embodiments of the present invention are described below with reference to the accompanying drawings.
< hardware configuration >
The method of the embodiments of the present invention may be implemented by at least one electronic device. Fig. 1 shows a hardware structure of an arbitrary electronic device. The electronic device shown in fig. 1 may be a laptop computer, a desktop computer, a mobile phone, a tablet computer, or any other device having a computing device such as a processor and a storage device such as a memory, and is not limited herein.
As shown in fig. 1, the electronic device 1000 may include a processor 1100, a memory 1200, an interface device 1300, a communication device 1400, a display device 1500, an input device 1600, a speaker 1700, a microphone 1800, and the like. Wherein the processor 1100 is adapted to execute computer programs. The computer program may be written in an instruction set of an architecture such as x86, Arm, RISC, MIPS, SSE, etc. The memory 1200 includes, for example, a ROM (read only memory), a RAM (random access memory), a nonvolatile memory such as a hard disk, and the like. The interface device 1300 includes, for example, a USB interface, a headphone interface, and the like. The communication device 1400 is capable of wired or wireless communication, for example, and may specifically include Wifi communication, bluetooth communication, 2G/3G/4G/5G communication, and the like. The display device 1500 is, for example, a liquid crystal display panel, a touch panel, or the like. The input device 1600 may include, for example, a touch screen, a keyboard, a somatosensory input, and the like. The electronic device 1000 may output voice information through the speaker 1700, and may collect voice information through the microphone 1800, and the like.
The electronic device shown in fig. 1 is merely illustrative and is in no way meant to limit the invention, its application, or uses. In any embodiment of the present invention, the memory 1200 of the electronic device 1000 is configured to store instructions for controlling the processor 1100 to operate so as to execute the method for constructing an artificial intelligence application according to any embodiment of the present invention. The skilled person can design the instructions according to the disclosed solution. How the instructions control the operation of the processor is well known in the art and will not be described in detail herein.
In further embodiments, the electronic device 1000 may also comprise an apparatus for building an artificial intelligence application according to any of the embodiments of the present invention, wherein the apparatus is arranged to perform the method for building an artificial intelligence application according to any of the embodiments of the present invention.
In one embodiment, an apparatus is provided that includes at least one computing device and at least one storage device to store instructions to control the at least one computing device to perform a method according to any embodiment of the invention.
The apparatus may include at least one electronic device 1000 as shown in fig. 1 to provide at least one computing device, such as a processor, and at least one storage device, such as a memory, without limitation.
< method examples >
FIG. 2 illustrates a flowchart of a method of building an artificial intelligence application, which can be implemented by the electronic device 1000 as shown in FIG. 1, for example, by an apparatus of the electronic device 1000 that builds an artificial intelligence application, which can include an application development apparatus equivalent to an application editor, according to one embodiment.
As shown in fig. 2, the method may include steps S2100 and S2200 as follows.
Step S2100, acquiring application configuration information configured for an application, where the application configuration information includes scheme information of an artificial intelligence scheme implementing the application.
The scenario information may include task content information of the scenario, i.e., information defining the task content performed by the scenario.
The solution may include one task item or may include a plurality of task items. The task content information may include content information of each task item and association information between the task items. The content information of each task item refers to the task content of the corresponding task item, and the association information reflects the dependency relationship between the task items, for example, the dependency relationship reflects the execution sequence between the task items.
Step S2200 is to generate an executable file of the application according to the application configuration information acquired in step S2100 and a preset template file.
In this embodiment, each configuration item is set in the preset template file according to a data structure that can be analyzed by the engine device, so as to generate an executable file that can be analyzed by the engine device for the application.
The data structure is, for example, a data structure conforming to the JSON specification, and is not limited herein.
In one embodiment, the template file has corresponding default configuration information defined for at least some of the configuration items therein. The at least part of the configuration items may include configuration items that may share configuration information for any application. The at least partial configuration items may also include configuration items that are individually configurable by the application developer.
In this embodiment, for each configuration item set in the template file, the application editor may directly load and display file content of the template file, so as to add or modify corresponding configuration information for the configuration item that needs to be configured by the application developer, for example, the application developer adds or modifies corresponding configuration information for the configuration item that needs to be configured individually, and after the application developer directly inputs these pieces of application configuration information through the template file, the application editor acquires application configuration information configured by the application developer for the developed application through step S2100.
In this embodiment, for each configuration item set in the template file, the application editor may also provide a configuration interface, and an application developer may input configuration information configured for the developed application through the configuration interface, so as to improve usability of the application editor and reduce a threshold for the application developer to construct an AI application. Correspondingly, the application editor may acquire application configuration information input through the provided configuration interface at step S2100.
In one embodiment, the application editor may provide the configuration interface only for the configuration items that are available for personalized configuration by the application developer, that is, the configuration interface exposed to the outside by the application editor may include only a part of the configuration items in the template file, for example, the configuration items that are available for personalized configuration by the application developer, and the like.
The configuration interface may include at least one of a configuration item providing check-out options (including drop-down lists), a configuration item providing input boxes.
In one embodiment, the method may further comprise the steps of: searching whether the template file has configuration information of a configuration item corresponding to a configuration interface; and providing corresponding configuration information in a corresponding configuration item of the configuration interface under the condition of having the configuration information. In this embodiment, for the configuration item in the configuration interface, if the selected template file has defined the configuration information of the corresponding configuration item, the configuration information may be provided in the corresponding configuration item of the configuration interface, so that an application developer may select whether to modify the configuration information of the configuration item in the configuration interface according to the application occasion of the constructed AI application, thereby further reducing the configuration amount of the constructed AI application.
In one embodiment, the generating an executable file of the application according to the application configuration information and the preset template file in step S2200 may include: and updating the template file according to the application configuration information to generate an executable file of the application. In this embodiment, for a configuration item in the template file where default configuration information is not defined, corresponding information in the application configuration information may be filled in the configuration item; for the configuration items with default configuration information defined in the template file, the original configuration information can be replaced by the corresponding information in the application configuration information, and then the updating of the template file is realized.
In one embodiment, after generating the executable file of the application, the method may further include: the executable file is published to the application providing device for downloading. In this embodiment, the application providing apparatus is provided with respect to an application store in which executable files of all the constructed AI applications are published, for example, an application developer constructs a single-table general application by the method of the embodiment, and the single-form general application is released to the application store, another application developer constructs an anti-money laundering application by the method of the embodiment, and issues the anti-money laundering application to the application store, and an application developer constructs an anti-fraud application by the method of the present embodiment, and also distributes the anti-fraud application to the application store and the like, so that the application store aggregates various constructed AI applications, and an application user who needs to use the AI application can download an executable file of a desired application to the application providing apparatus to run the desired application.
In one embodiment, after generating the executable file of the application, the method may further include: loading an engine device matched with the template file to run the executable file; and in the case of a run through, publishing the executable file for download. In this embodiment, by loading the engine apparatus to run the executable file for verification before the release, the validity of the executable file of the application released to the application providing apparatus can be ensured.
In one embodiment, for the task content information of a solution in the solution information, the template file may provide a configuration item, such as a jobs configuration item, which defines the task item and the associated information between the task items, and the application developer may edit the defined configuration information for the configuration item, so as to implement the expression of the task content information.
In one embodiment, in order to facilitate the construction of a more complex AI application, the device for constructing an AI application may provide a component development device that meets the requirements of an engine device framework, and a component developer may develop some functional components that meet the requirements of the engine device framework in advance through the component development device, so that the component developer may select and use the functional components for constructing an AI application by the component developer to generate an executable file of the application, and the engine device may call the functional components related to the executable file to run the executable file. Here, each functional component is equivalent to a packaged module, each functional component can complete at least one task item, and when an application developer constructs an AI application, the application developer can express task content information of the constructed AI application by selecting a required functional component and configuring operation information of each functional component, without customizing the task items.
Therefore, in this embodiment, the task content information of the recipe may include the functional components used by the recipe and the operation information of each functional component. The operation information may include, for example, at least one of operation phase information to which the corresponding functional component belongs in the recipe, initialization information of the corresponding functional component, and operation condition information of the corresponding functional component. Therefore, when the engine device runs the executable file of the AI application, the corresponding functional components can be called according to the task content information and the running information so as to provide the application service. With this embodiment, it is possible to simplify the configuration for constructing the AI application, and to construct a complicated AI application in a simpler manner.
The operation phase represents a task phase to which the corresponding functional component belongs in the constructed AI application, and includes a data template phase, a scheme generation phase and the like.
The initialization information indicates the initialization order, whether to initialize, etc. of the corresponding functional components.
The operation condition information indicates a condition for operating the corresponding functional component, that is, when the condition is satisfied during the process of operating the executable file by the engine device, the corresponding functional component is called again to complete the corresponding task item.
In one embodiment, these functional components may include, for example, any one or more of the following: the system comprises a data preprocessing component, a data splitting component, a data merging component, a feature extraction component, a model training component, a model evaluation component, a self-learning component, a real-time estimation component, a batch estimation component and the like.
The data pre-processing component may, for example, perform a task event of selecting a data table in the input data table and selecting data in the selected data table. The data merge component can, for example, complete a task transaction that merges selected ones of the retrieved data sheets into one data sheet. The data splitting component may, for example, complete a task transaction that splits the data into training data and validation data. The feature extraction component may, for example, perform a task of extracting features required for model training or model prediction from the data. The model training component can, for example, complete a task item for model training based on a specified algorithm, such as a binary algorithm. The model evaluation component may, for example, complete the evaluation of the trained model by the validation data and give the task items of the set model indices. The real-time estimation component may, for example, complete a task event for real-time prediction of the input data by the model. The batch prediction component may, for example, complete a task for batch prediction of input data by the model. The self-learning component may, for example, complete a task event that continues to train a prior model through newly generated data to effect model updates.
In one embodiment, the method may further include a step of generating a functional component, the step of generating a functional component being implemented by, for example, a component development apparatus of an apparatus for building an artificial intelligence application, the step including, for example, the following steps S3100 to S3400:
step S3100, providing a functional component development template.
In this embodiment, the system for constructing the AI application may provide a component development apparatus that provides a functional component development template that can be analyzed by the engine apparatus based on the architecture of the engine apparatus, and a component developer may implement a functional component that can complete a certain task or tasks using the functional component development template.
In this embodiment, the component developer may define the functional component information through the functional component development template, where the functional component information includes function implementation information corresponding to the functional component, that is, implementation codes corresponding to the functional component.
The functional component information may also include makefile files written for the corresponding functional components, and the like.
In this embodiment, the component developer may further define information about configuring the functional component through the functional component development template, so as to generate the component configuration template, so that when the application developer constructs the AI application through the application editor, the application developer may configure the selected functional component by using the component configuration template, so that the selected functional component can play a role in the constructed application according to the configuration of the application developer.
The information about the configured functional component may include, for example, a name of the corresponding functional component, a version of the corresponding functional component, a back-end image of the corresponding functional component, a front-end UI package of the corresponding functional component, a command-line interface (CLI) of the corresponding functional component, and the like. Here, the function component corresponds to a service running in the engine device, and is a micro service organized according to a convention rule, and includes a micro service image of a back end, a function component running UI, a function component configuration custom UI, an sdk/cli package of the function component, and the like.
The information about configuring the functional component may include configuration information for configuring the corresponding functional component, and may also include information for setting configuration items for the corresponding functional component.
Step S3200, acquiring functional component information and functional component configuration information defined according to the functional component development template.
Step S3300, pack the file containing the functional component information, and generate the corresponding functional component.
In this embodiment, the packaged file may include one file or multiple files, so that the corresponding functional components are defined by the files.
The component development means may provide an operation interface for triggering a packaging command, through which a component developer may trigger a packaging request, according to which the component development means performs the packaging operation.
Step S3400, generating a component configuration template corresponding to the functional component according to the information about the configured functional component.
In this embodiment, the component development apparatus may set the configuration item for the functional component and set default configuration information for at least part of the configuration items according to the requirement for the data structure of the component configuration template, by using the acquired information about configuring the functional component, so as to generate the component configuration template corresponding to the functional component. Therefore, the generated component configuration template contains each configuration item of the corresponding functional component, and default configuration information is defined for at least part of the configuration items.
The application development device can be used for the application developer to perform personalized configuration on the selected functional component when constructing the AI application by loading the generated component configuration template.
In one embodiment, the method may further comprise: after the corresponding functional component and the component configuration template thereof are generated, the corresponding functional component and the component configuration template thereof are uploaded to a component library, and the corresponding functional component is registered in the component library.
The component library stores a database of component configuration templates corresponding to the functional components and corresponding functional components for the component providing apparatus.
In this embodiment, the component development apparatus may provide an operation interface for triggering issuance of a command, and the component developer may trigger issuance of a request through the operation interface, and execute the issuance operation according to the request.
In this embodiment, the component development apparatus may also automatically trigger the operation of uploading the corresponding functional component and the component configuration template thereof to the component library after generating the corresponding functional component and the component configuration template thereof based on the packaging request.
According to the embodiment, after the component developer registers the developed functional component to the component library, the application development device can call the functional component from the component library to be selected and configured by the application developer, so that the AI application is constructed.
Based on various functional components developed by the component developer, in one embodiment, the application developer can configure the task content information in the solution information by configuring the functional components used by the application solution and the running information of each functional component. Therefore, in this embodiment, the task content information in the scenario information may include the functional components used by the scenario and the operation information of each functional component.
In this embodiment, the application configuration information may include component configuration information configured for each functional component, in addition to the scenario information. The component configuration information is, for example, component configuration information configured by an application developer based on a component configuration template of the selected functional component. Therefore, when the engine device runs the executable file, the function components used by the scheme and the running information of each function component can be obtained through the scheme information, the action mode of the corresponding function component in the application can be obtained through the component configuration information of each function component used by the scheme, and the component configuration information is transmitted to the corresponding function component, so that when the engine device calls the function component to run the AI application, the function component can complete the corresponding task items according to the configured action mode.
In one embodiment, the scenario information in the application configuration information acquired in step S2100 may include basic information of the scenario, where the basic information includes at least one of a scenario name, a keyword of a scenario applicable field, and scenario description information. Therefore, the application user can know the application occasion of the corresponding AI application according to the basic information of the scheme, and select the executable file of the required AI application for use in the application store according to the basic information. In this embodiment, the application editor may provide basic information configuration items in the configuration interface for the application developer to enter basic information for the schema through the configuration interface.
In one embodiment, the preset template file used in step S2200 may include a schema template file corresponding to the schema information, and the application editor may set the configuration interface according to the configuration items in the schema template file to allow the application developer to configure at least the configuration items in the schema template file through the configuration interface to form the schema information.
In one embodiment, the scenario template file may include basic information configuration items of a scenario, service information configuration items of a scenario, and operation and maintenance information configuration items of a scenario.
In this embodiment, the task content information in the scenario information belongs to the service information, that is, the task content information is information formed by configuring the service information configuration item.
In this embodiment, in the case that the task content information includes the functional components used by the scheme and the operation information of each functional component, the service information may further include common configuration information of each functional component. Thus, the configuration information suitable for each selected functional component can be configured through the part of the common configuration information configuration item, and the readability of the scheme template file is improved.
The common configuration information may include, for example, at least one of a common data dictionary, public system parameters, and common initialization information.
The common data dictionary may include, for example, configuration information of the identifier key and the value, and may also include some common rule definitions run by the engine device.
In this embodiment, the service information may further include at least one of entry information of the scenario and interface information of the scenario. The application home page interface entered when the engine device runs the corresponding AI application can be configured through the portal information configuration item, the application home page interface can be configured to be a default interface of the engine device, can also be configured to be a home page interface of a certain functional component, and the like, and the configuration can be performed by setting a home page path, so that the portal information configuration item can be a home page path configuration item. The display content, the layout structure and the like of the interface can be configured through the interface information configuration item, for example, the view and the report of the interface, the setting and the display of a business intelligent dashboard (BI dashboard) and the like can be defined through the interface information configuration item.
In this embodiment, the basic information may include at least one of a scheme name, a keyword of a scheme applicable field, and scheme description information. Here, the basic information of the schema in the schema information acquired in step S2100 is information configured by being configured for the basic information configuration item.
In this embodiment, the operation and maintenance information may include operation mode information, operation resource information, and the like. For example, the operation mode of the corresponding AI application, the CPU resource required for operating the corresponding AI application, and the like may be configured by the operation and maintenance information configuration item.
In one embodiment, the application editor may provide two types of template files, one being the above-described schema template file embodying schema information and the other being a component template file embodying component configuration information, which is declared by the component developer when developing the corresponding functional components, i.e., each functional component has a component template file corresponding to each.
In this embodiment, the application configuration information configured by the application developer through the configuration interface includes, for example, schema information for configuring configuration items in a schema template file, and component configuration information for configuring configuration items in a component template file of the selected functional component.
In this embodiment, the configuration interface provided by the application editor may include a configuration interface corresponding to at least a portion of the configuration items in the solution template file and a configuration interface corresponding to at least a portion of the configuration items in each component template file.
Similar to the above-mentioned solution template file, the component template file may include a basic information configuration item of the corresponding component, a service information configuration item of the corresponding component, and an operation and maintenance information configuration item of the corresponding component.
The service information of the component includes operation rule information of the corresponding component and the like. The business information of the component may also include at least one of context declaration information and message declaration information to enable definition of an execution environment, messaging, and the like.
The basic information of the component may include at least one of a version of the corresponding functional component, a creator of the corresponding functional component, a unique identifier of the corresponding functional component, a name of the corresponding functional component, and a data group in which the corresponding functional component is located.
The component template file can also comprise configuration interface information so as to customize a configuration interface of the corresponding functional component through the configuration interface information, so that an application developer can input component configuration information through the configuration interface defined by the component template file so as to realize the configuration of the corresponding component template file.
The configuration interface information may be interface content of the configuration interface or a loading path of the configuration interface package. In this way, the application editor may render the corresponding configuration interface according to the configuration interface information in the component template file for configuration by the application developer.
In one embodiment, the method may comprise: and loading a scheme configuration interface set by the application editor in response to the operation of the configuration scheme information, and providing a configuration interface for configuring at least part of configuration items in the scheme template file on the scheme configuration interface.
In one embodiment, the method may further comprise: responding to the operation of task content information in the definition scheme information, and loading list items of the functional components in the functional component library; and detecting the editing operation of the list item to obtain the functional components used by the scheme and the running information of each functional component.
In this embodiment, the function components are registered in the component library, the application editor may be loaded to all the function components in the component library, and when an application developer triggers an operation of defining task content information, which may be triggering an operation of editing the function components, list items of the function components are provided, that is, each list item corresponds to one function component, and at least part of basic information of the function component may be displayed on the list item, so that the application developer may select a required function component according to content displayed on the list item.
The editing operation may include, for example, a selection operation, a shift operation, and the like. The functional components used by the scheme can be obtained by detecting the selection operation of the application developer on the list items, and the initialization sequence of each functional component and the like can be obtained by detecting the shift operation of the application developer on the list items, such as the operations of moving up and down and the like.
In one embodiment, the method may comprise: in response to the operation of configuring the selected functional component, analyzing a component template file corresponding to the functional component; and under the condition that the component template file defines configuration interface information, rendering a configuration interface corresponding to the functional component according to the configuration interface information. Thus, an application developer can select a configuration interface customized by the functional component to complete the configuration of the corresponding component template file.
In an example of configuring corresponding functional components through the component template file, configuring the functional components is also referred to as configuring the component template file of the functional components.
In one embodiment, the method may comprise: in response to an operation of configuring the selected functional component, file contents of a component template file of the corresponding functional component are loaded for configuration. In this embodiment, the application developer may select to fill in or modify the corresponding configuration information for the configuration item that needs to be personalized for the constructed AI application directly on the component template file.
In one embodiment, the application editor may provide a variety of ways to configure the functional components, such as configuration via a custom configuration interface for the functional components, configuration directly via a template for the functional components, configuration via a default component configuration interface for the application editor, and so forth.
In this embodiment, the method may further include: and providing a configuration entry interface in response to the operation of configuring the selected functional component, wherein the configuration entry interface comprises a first configuration entry and a second configuration entry, the first configuration entry is an entry for selecting to load the configuration interface defined by the component template file of the corresponding functional component for configuration, and the second configuration entry is an entry for selecting to load the file content of the component template file of the corresponding functional component for configuration.
In this embodiment, the configuration entry interface may further include other configuration entries, for example, an entry configured through a default component configuration interface of the application editor, and the like, which is not limited herein.
For example, as shown in FIG. 3, in an example, the configuration portal interface provides three configuration portals, respectively an "original files" configuration portal, an "official UI" configuration portal, and a "standard UI" configuration portal. In this example, the "official UI" configuration entry corresponds to the first configuration entry, and when detecting that the application developer clicks the "official UI" option, the application editor loads the configuration interface defined by the component template file of the corresponding functional component, so as to allow the application developer to configure the corresponding functional component. In this example, the "original file" configuration entry corresponds to the second configuration entry, and as shown in fig. 3, the application developer clicks on the "original file" option, so that the application editor loads the file content of the template file of the corresponding functional component at the bottom, and the developer can directly edit the file content to implement the configuration for the corresponding functional component. In this example, the "standard UI" is a portal for selecting configuration via the application editor's default component configuration interface, which is loaded by the application editor upon detecting that the application developer has selected the "standard UI" option.
In one embodiment, an operation interface for triggering an operation of configuring a corresponding functional component may be provided on a list item of the functional component documented by the application editor. The operation interface is, for example, a list item itself, and an application developer may trigger an operation of configuring a corresponding functional component by clicking the list item. The operation interface may also be, for example, an interface provided on the list item, and an application developer may trigger an operation of configuring the corresponding functional component by clicking the interface on the list item.
Continuing with FIG. 3 as an example, the application editor may display a list item of a functional component selected by the application developer in the flow layout area of the editing interface, for example, a list item of a data processing component and a list item of a model training component, and the application developer may trigger an operation of configuring the data processing component by clicking the list item of the "data processing component", so that the application editor provides a configuration entry interface of the data processing component on the right side of the flow layout area.
In the embodiment where the preset template file includes a schema template file and a component template file, the generating an executable file of an application according to the application configuration information and the preset template file in step S2200 may include the following steps S2210 to S2230:
step S2210, a first file defining the project is generated according to the project information and the project template file.
The first file may be, for example, a file in the.yaml format.
The project information includes configuration information for at least some of the configuration items in the project template file, and the step S2210 may include filling the configuration information in the corresponding configuration items of the project template file, and/or replacing existing configuration information of the corresponding configuration items of the project template file with the configuration information.
Step S2220, for each functional component, generates a second file defining the corresponding functional component according to the component configuration information of the corresponding functional component and the component template file of the corresponding functional component.
The second file may be, for example, a.yaml format file.
In this embodiment, a corresponding second file is generated for each selected functional component. For example, if the application developer selects three functional components, three second files will be generated.
The component configuration information includes configuration information for at least some configuration items in the corresponding component template file, and the step S2220 may include filling the configuration information in the corresponding configuration item of the corresponding component template file, and/or replacing existing configuration information of the corresponding configuration item of the corresponding component template file with the configuration information.
In step S2230, the first file and each second file are packaged to generate an executable file of the built application.
In this embodiment, the executable file is a packed file formed by compressing a first file generated based on a schema template file and a second file generated based on a component template file of the selected functional component, that is, the generated executable file includes a first file and at least one second file, and the second file corresponds to the functional component related in the first file one to one.
The following is the file content of the schema template file or the file content of the first file generated based on the schema template file according to an example:
id:1
version:'1.0'
kind:asol
meta:
name (scheme name) general scheme
And the general scheme uses an intelligent tuning framework to solve the problem of two classification decision-making of a single behavior table, automatically completes feature generation, feature selection, algorithm selection and algorithm parameter tuning based on behavior and feedback data uploaded by a user, generates an optimal training scheme, can be automatically deployed and continuously and automatically updates the model. There is confidence that in a common scenario the modeling level of a modeling expert is close.
Figure BDA0002338106380000151
The following is the file content of a component template file or a second file generated based on a component template file according to one example:
Figure BDA0002338106380000152
Figure BDA0002338106380000161
Figure BDA0002338106380000171
< apparatus embodiment >
FIG. 4 is a functional block diagram of an apparatus for building artificial intelligence applications, according to one embodiment. As shown in fig. 4, the apparatus 4000 may include an application configuration module 4100 and a document generation module 4200.
The application configuration module 4100 is configured to obtain application configuration information configured for an application, where the application configuration information includes scenario information for implementing an artificial intelligence scenario of the application. The file generating module 4200 is configured to generate an executable file of the application according to the application configuration information and a preset template file.
In one embodiment, the recipe information includes task content information for the recipe.
In one embodiment, the task content information includes content information of each task item and association information between the task items.
In one embodiment, the task content information includes functional components used by the solution and operation information of each functional component; the application configuration information further includes component configuration information configured for each of the functional components, respectively.
In one embodiment, the operation information includes at least one of operation phase information to which the corresponding functional component belongs in the solution, initialization information of the corresponding functional component, and operation condition information of the corresponding functional component.
In one embodiment, the file generation module 4200, when generating the executable file of the application according to the application configuration information and the preset template file, may be configured to: generating a first file defining the scheme according to the scheme information and the scheme template file; aiming at each functional component, generating a second file defining the corresponding functional component according to the component configuration information of the corresponding functional component and the component template file of the corresponding functional component; and packaging the first file and each second file to generate the executable file.
In one embodiment, the component template file may include a basic information configuration item of a corresponding component, a service information configuration item of the corresponding component, and an operation and maintenance information configuration item of the corresponding component, where the service information of the component includes operation rule information of the corresponding component.
In one embodiment, the basic information of the component may include at least one of a version of the corresponding functional component, a creator of the corresponding functional component, a unique identifier of the corresponding functional component, a name of the corresponding functional component, and a data set in which the corresponding functional component is located.
In one embodiment, the set of priced service information may further include at least one of context declaration information and message declaration information.
In one embodiment, the component template file may also include configuration interface information.
In one embodiment, the application configuration module 4100 may be further configured to: analyzing a component template file corresponding to the functional component in response to the operation of configuring the functional component; and under the condition that the component template file defines configuration interface information, rendering a configuration interface corresponding to the functional component according to the configuration interface information.
In one embodiment, the application configuration module 4100 may be further configured to: and responding to the operation of configuring the functional component, and loading the file content of the component template file of the corresponding functional component for configuration.
In one embodiment, the application configuration module 4100 may be further configured to: and providing a configuration entry interface in response to the operation of configuring the functional component, wherein the configuration entry interface comprises a first configuration entry and a second configuration entry, the first configuration entry is an entry for selecting to load the configuration interface defined by the component template file of the corresponding functional component for configuration, and the second configuration entry is an entry for selecting to load the file content of the component template file of the corresponding functional component for configuration.
In one embodiment, the application configuration module 4100 may be further configured to: in response to an operation of defining the task content information, loading list items of the functional components in the functional component library; and detecting editing operation on the list items, and acquiring the functional components used by the scheme and the running information of each functional component according to the editing operation.
In one embodiment, the editing operation includes a selection operation and a shift operation.
In one embodiment, the application configuration module 4100 may be further configured to: and providing an operation interface for triggering the operation of configuring the corresponding functional component on the list item.
In one embodiment, the functional components used by the scheme may include any one or more of the following: the system comprises a data preprocessing component, a data splitting component, a data merging component, a feature extraction component, a model training component, a model evaluation component, a self-learning component, a real-time estimation component and a batch estimation component.
In one embodiment, the apparatus 4000 may further include a component generation module, which may be configured to: providing a functional component development template; acquiring functional component information defined according to the functional component development template and information related to configuration functional components, wherein the functional component information comprises function implementation information of corresponding functional components; packing the file containing the functional component information to generate a corresponding functional component; and generating a component configuration template corresponding to the functional component according to the information about the configured functional component.
In one embodiment, the apparatus 4000 may further include a component publishing module configured to: and responding to the operation of issuing the corresponding functional component, uploading the corresponding functional component and the component configuration template of the corresponding functional component to a component library, and registering the corresponding functional component in the component library.
In one embodiment, the preset template file may include a scheme template file, where the scheme template file includes a basic information configuration item of a scheme, a service information configuration item of the scheme, and an operation and maintenance information configuration item of the scheme, where the service information includes the task content information.
In one embodiment, the task content information may include functional components used by the solution and operation information of each of the functional components; the service information also includes common configuration information of each of the functional components.
In one embodiment, the common configuration information may include at least one of a common data dictionary, public system parameters, and common initialization information.
In one embodiment, the business information of the scenario may further include at least one of entry information of the scenario and interface information of the scenario.
In one embodiment, the operation and maintenance information of the solution may include operation mode information and operation resource information.
In one embodiment, the scenario information may further include basic information of the scenario, where the basic information of the scenario includes at least one of a scenario name, a keyword of a scenario applicable field, and scenario description information.
In one embodiment, the application configuration module 4100, when obtaining application configuration information configured for the application, may be configured to: and acquiring the application configuration information input through the provided configuration interface.
In one embodiment, the arbitrary configuration interface may include at least one of a configuration item providing a check option and a configuration item providing an input box.
In one embodiment, the application configuration module 4100 may be further configured to: searching whether the template file has configuration information corresponding to the configuration items of the configuration interface; and, in the event of having the configuration information, providing the corresponding configuration information in the corresponding configuration item of the configuration interface.
In one embodiment, the template file has defined corresponding configuration information for at least some of the contained configuration items.
In one embodiment, the file generation module 4200, when generating the executable file of the application according to the application configuration information and the preset template file, may be configured to: and updating the template file according to the application configuration information to generate an executable file of the application.
In one embodiment, the apparatus may further include an application publishing module, which may be configured to: after the file generation module generates the executable file of the application, loading an engine matched with the template file to run the executable file; and, in the event of a pass through, publishing the executable file for download.
In one embodiment, the application publishing module may be to: after the file generation module 4200 generates the executable file of the application, the executable file is released to the application providing apparatus for downloading.
< System embodiment >
In one embodiment, referring to fig. 5, there is further provided a system AIS running an artificial intelligence scheme, the system AIS comprising an engine apparatus 5000 and an apparatus for constructing artificial intelligence application 4000 of any of the above embodiments.
In this embodiment, the apparatus 4000 may include the application development apparatus 4000 a. The application development apparatus 4000a may include the application configuration module 4100, the file generation module 4200, the application release module, and the like, which belong to the application development apparatus 4000 a.
In this embodiment, the apparatus 4000 for constructing an artificial intelligence application is configured to generate an executable file of the application, where the executable file has a data structure that can be parsed by the engine apparatus 5000.
In this embodiment, the engine 5000 is configured to run the application according to the executable file.
In this embodiment, since the data structure of the executable file generated by the device 4000 is a data structure that can be analyzed by the engine device 5000, when the engine device 5000 receives a request for running the application, the engine device 5000 can run the application by analyzing the executable file of the application.
In one embodiment, the system AIS may further include a dedicated application providing device 6000, the device 4000 may be further configured to publish the generated executable file to the application providing device 6000 for downloading, and the engine device 5000 may be configured to run the application according to the executable file downloaded from the application providing device 6000. In this embodiment, the application providing apparatus 6000 is equivalent to an application store, and the application user can download an executable file of an application satisfying the use requirement of the application user in the application providing apparatus 6000 and start the engine apparatus 5000 to run the application.
In one embodiment, an application developer may edit an executable file of an application through some functional components previously developed by a component developer, and therefore, the system AIS may further include a component providing apparatus 7000, and after the component developer completes development of the functional components, the component developer may send a package file of the functional components to the component providing apparatus 7000 for use by the developer in developing an AI application through the apparatus 4000, and therefore, the component providing apparatus 7000 is used for providing the functional components to which the executable file relates.
In this embodiment, in the case that the executable file of the application relates to a functional component, when the application is executed according to the executable file, the engine apparatus 5000 is configured to call the related functional component according to the executable file to execute the application.
In this embodiment, referring to fig. 5, the device 4000 may include the component development device 4000b, and the component development device 4000b is used by a component developer to develop a functional component for an application developer to develop an AI application. The component development apparatus 4000b includes, for example, the component generation module and the component release module.
< Engine apparatus embodiment >
In one embodiment, an Engine apparatus 5000 is also provided, as shown in fig. 6, the Engine apparatus 5000 includes an Engine management (Engine manager) module 5100 and an Engine execution (Engine Core) module 5200.
The engine management module 5100 is used to parse an executable file of an AI application, which is obtained by defining the application based on a template file adapted to the engine apparatus 5000.
The template files include, for example, a plan template file and a component template file.
The engine execution module 5200 is used to run the application according to the executable file.
In one embodiment, the engine executing module 5200 may load and display the home interface of the application according to the home interface information of the application configured in the executable file when the application is run according to the executable file. The home interface information is, for example, configuration information under a service information configuration item in the first file.
In one embodiment, as shown in fig. 6, the engine execution module 5200 may include, for example, a Process coordinator module (Process coordinator) 5201, a Component LifeCycle (Component LifeCycle) management module 5202, a Message (Message) management module 5203, a run environment Context (Context) management module 5204, an access point (EntryPoint) management module 5205, an operation and maintenance lateral Support (Ops Support) module 5206, a Business lateral Support (Business Support) module 5207, and a routing (Router) module 5208.
The component lifecycle management module 5202 can be used to manage and control the whole lifecycle of starting, stopping, updating, monitoring, etc. of the processing function component. The flow coordination module 5201 may be configured to perform a transformation process of migration of each operation stage according to preset arrangement information of each operation stage, and coordinate ordered work of each functional component. The access point management module 5205 may be used to perform a built-in access point (service Component, task Component), and to perform management, triggering, and operation of a custom access point. The message management module 5203 may be used to perform internal message management of the engine apparatus 5000, providing a subscription (Topic-Subscribe) mode. The runtime environment context management module 5204 can be used to store and exchange the built-in variables and public information for the full lifecycle of the engine apparatus 5000. The operation and maintenance lateral support module 5206 may be used to perform operation and maintenance related public capacity support of the engine device 5000 and the functional components, such as Log, monitoring, tracking, and other data collection and analysis. The business lateral support module 5207 can be used to perform common capacity support related to the engine apparatus 5000 and the functional component business, such as consistency, effect, performance, efficiency analysis, and the like.
The routing module 5208 may be used to forward and proxy the frame of the engine apparatus 5000 to the view, api, cli, etc. of the functional component, but is not responsible for service forwarding of the functional component's exposure to the external port.
In one embodiment, the executable file may include a first file and a second file corresponding to the functional components used one to one; the first file comprises scheme information defining a scheme, wherein the scheme information comprises task content information; each second file contains component information defining a corresponding functional component.
In one embodiment, the scenario information includes basic information of the scenario, service information of the scenario, and operation and maintenance information of the scenario, where the service information includes task content information.
In one embodiment, the business information of the scenario may further include at least one of entry information of the scenario and interface information of the scenario.
In one embodiment, the basic information of the solution may include at least one of a solution name, a keyword of a solution applicable field, and solution description information.
In one embodiment, the component information of each second file may include basic information of a corresponding functional component, business information of the corresponding functional component, and operation and maintenance information of the corresponding functional component, where the business information of the corresponding functional component includes business rule information of the corresponding functional component, and the like.
In one embodiment, the executable file may be a compressed file obtained by compressing the files contained therein. Thus, the engine management module 5100, when parsing the executable file of the application, may be configured to: decompressing the executable file; and storing all the files obtained by decompression into a directory area according to the set storage path. The engine execution module 5200, when running the application from the executable file, may be configured to: the application is run according to the files of the directory area.
For example, the executable file includes a compressed file of a first file and at least one second file, so that when the executable file is parsed, the executable file may be first decompressed to obtain the first file and the at least one second file, and the files may be saved in the directory area for the engine execution module 5200 to read.
The set save path is the default save path of the engine apparatus 5000.
For example, the engine management module 5100 may include a file parsing (aslexplorel) module, through which the engine management module 5100 may parse an executable file.
In one embodiment, the engine execution module 5200, when running the application according to the files of the directory area, may be configured to: an engine instance module 5200 'is obtained from the file initialization engine execution module 5200 of the directory area, including the components of the initialization engine execution module 5200, that is, the engine execution module is made to be an engine instance module 5200' dedicated to running the application; and, the engine instance module 5200' is launched to run the application.
In this embodiment, the Engine device 5000 is a general device capable of running various AI applications, and the Engine execution module 5200 has a general configuration for running an application according to an executable file, but since different AI applications may have respective running requirements, the Engine execution module 5200 may initialize the Engine execution module 5200 according to configuration information of each configuration item in a file (for example, including a first file and a second file) of a directory area when it is required to run one AI application, so that the Engine execution module 5200 forms an Engine Core-Instance (ECI) module capable of being dedicated to running the AI application, and runs the AI application through the Engine Instance module.
In one embodiment, the engine management module 5100, when parsing the executable file of the application, may be configured to: in response to a request to run the application, the executable file of the application is obtained and the engine execution module 5200 is notified to run the application according to the executable file.
The request may be triggered by an application user, and after receiving the request, the engine device may parse an executable file of the AI application corresponding to the request to run the application.
In one embodiment, the engine management module 5100, when parsing the executable file of the application, may be configured to: detecting whether the executable file can be identified; in the event that the executable file can be identified, the engine execution module 5200 performs an operation of running the application according to the executable file.
In this embodiment, the engine management module 5100 detecting whether the executable file can be identified may include, for example: the engine management module 5100 searches the record data of the template file for the version number of the template file used by the executable file, and if the version number is found, it indicates that the executable file is generated using the template file that can be identified by the engine device 5000, and the detection result indicates that the engine device 5000 can identify the executable file.
In one embodiment, the executable file defines task content information for implementing the application's schema. In this embodiment, the engine executing module 5200 may be configured to execute each task item included in the solution according to the task content information when running the application according to the executable file, so as to run the application.
In one embodiment, the executable file defines task content information for implementing a schema of the application, the task content information including functional components used by the schema and execution information for each of the functional components. In this embodiment, the engine execution module 5200, when running the application according to the executable file, can be configured to call the functional components used to run the application according to the executable file. In this embodiment, for example, the flow coordination module of the engine execution module 5200 may call the used functional component to complete the corresponding task item according to the executable file to run the application.
In one embodiment, the operation information includes at least one of operation phase information to which the corresponding functional component belongs in the scheme, initialization information of the corresponding functional component, and operation condition information of the corresponding functional component. In this embodiment, the engine execution module 5200 may start (or initialize) the used functional component according to the execution information of the executable file to execute the application. In this embodiment, for example, the component lifecycle management module of the engine execution module 5200 may start a functional component or the like to be used based on the execution information.
In one embodiment, to enable the functional components involved by the executable file to function in a configured manner when called by the engine arrangement 5000, the engine execution module 5200 is further configured to, prior to calling the functional components from the executable file to run the application: initializing a corresponding functional component according to the running information of each functional component defined by the executable file; acquiring and registering an external communication interface returned by each functional component after initialization is completed; and pushing the component information of each functional component defined by the executable file to the corresponding functional component through the external communication interface. In this embodiment, the relevant operations may be performed, for example, by a component lifecycle management module of the engine execution module 5200.
In this embodiment, the engine executing module 5200 initializes the functional component according to the configuration information of the functional component configuration item scComponents under the business information configuration item businessSetting, for example, according to the first file, where the configuration information of the functional component configuration item scComponents includes the functional component used by the corresponding scheme and the operation information of the corresponding functional component.
In this embodiment, the operation information includes initialization information, which reflects whether the corresponding functional components need to be initialized and an initialization sequence, and the engine executing module 5200 may sequentially initialize the functional components according to the initialization sequence determined by the initialization information of the functional components to complete the starting of the functional components, and for the functional components that do not need to be initialized, the starting is skipped in the initialization stage, and the corresponding functional components are started after the corresponding operation conditions are detected.
In this embodiment, after the initialization is completed, the functional components return their external communication interfaces to the engine execution module 5200, so that the engine execution module 5200 can push the component information of the corresponding functional component defined in the executable file to the corresponding functional component through the external communication interface to complete the configuration of the corresponding functional component for the AI application, and further, the functional component can complete the corresponding task items in the executable file configuration manner when the engine execution module 5200 runs the application. In this embodiment, for example, the function component information defined in the second file is pushed to the corresponding function component to complete the configuration of the corresponding function component for the application.
The engine execution module 5200 (or the engine instance module 5200') reports its own status to the engine management module 5100 after starting, so that the engine management module 5100 marks and displays the engine status, for example, during initialization of each functional component, the engine management module 5100 sets its own status to "being started", after completing initialization of each functional component, sets its own status to "running", and if the initialization of each functional component is abnormal, sets its own status to "abnormal".
In one embodiment, the engine execution module 5200 is further configured to, prior to calling a functional component from the executable file to run the application: acquiring at least one item of state information, running environment context information and subscription message information returned by each functional component after initialization is completed, wherein the state information can comprise starting state information, running state information and the like; and, the information returned by the function component is registered in the engine execution module 5200. In this embodiment, the engine execution module 5200 may control the performance of the corresponding functional component according to the information during the running of the application, for example, subscribe to related messages, periodically check common context information, and the like. In this embodiment, the relevant operations may be performed, for example, by a message management module of the engine execution module, a runtime environment context management module.
In one embodiment, the executable file defines the operating condition information of the functional component, and the engine execution module 5200, when calling the functional component to run the application according to the executable file, can be configured to: and calling the functional components to execute corresponding tasks according to the running condition information of each functional component defined in the executable file so as to run the application.
In this embodiment, for example, the running condition information of the functional component that needs to be used may be configured in a first file of the executable file, for example, the functional component configuration item under the service information configuration item of the first file configures three functional components, and the running condition information of the three functional components is configured as follows: after the first functional component completes the corresponding task, the second functional component executes the corresponding task based on the output data of the first functional component, and the third functional component executes the corresponding task based on the output data of the second functional component.
In one embodiment, the running information configured in the executable file may not include the running condition information, but only include running phase information, initialization information, and the like. In this embodiment, the engine execution module 5200 may preset an operation sequence of a common operation phase of the AI application, so that, when the operation information configured by the executable file includes the operation phase information, the engine execution module 5200 may call each functional component to execute a corresponding task according to the preset operation sequence of different operation phases, which is beneficial to reducing configuration information required to construct the AI application. In this embodiment, the relevant operations may be performed by, for example, a flow coordination module of the engine execution module 5200.
In this embodiment, the engine execution module 5200, when calling the functional component to run the application according to the executable file, can be configured to perform the following steps S5211 to S5214:
in step S5211, the operation phase information of each functional component defined by the executable file in the scheme is obtained.
The operating phase information is, for example, configuration information under the functional component configuration item scComponents of the first file.
Step S5212, preset arrangement information reflecting the operation sequence of each operation stage is acquired.
For example, an AI application generally includes a data processing stage, a model investigation stage, a model prediction stage, a self-learning stage, etc., and in any AI application, the operation sequence of these operation stages can be basically determined, for example, the operation sequence of the model investigation stage is arranged after the data processing stage, the model prediction stage is arranged after the model investigation stage, and the self-learning stage is arranged after the model prediction stage, etc., so that, according to the possible operation sequence relationship of different operation stages, the arrangement information about the operation sequence of each operation stage can be obtained, and according to the arrangement information, the operation sequence relationship among these function components related to the application to be operated can be determined, and further, the operation condition information of these function components can be obtained.
Step S5213, obtaining the operating condition information of each functional component defined by the executable file according to the operating phase information and the arrangement information.
Step S5214, according to the running condition information, calling the functional component to execute the corresponding task to run the application.
In this embodiment, when the executable file defines the operation phase information of the functional component, the engine execution module 5200 may determine the operation condition information of each functional component according to the operation phase information and the preset arrangement information, without requiring an application developer to configure the operation condition information when constructing the AI application, thereby improving the usability of the system AIs.
In one embodiment, the engine execution module 5200, when calling the functional component from the executable file to run the application, may be configured to: in response to an event of monitoring a task completion message returned by a functional component in the current operation stage of the scheme, checking whether the current operation stage is completed according to an executable file; and under the condition of completion, providing the output data of the current operation stage for the functional component of the next operation stage of the scheme, and calling the functional component of the next operation stage to execute the corresponding task. In this embodiment, the relevant operations may be performed by, for example, a flow coordination module of the engine execution module 5200.
In this embodiment, after the functional component executes the corresponding task, a task completion message is returned to the engine device, and after the engine device receives the task completion message, in response to an event that the task completion message is received, it may check whether all the functional components in the current operating phase have returned any completion message according to the operating phase information of the functional component configured in the executable file, if yes, it indicates that the current operating phase has been completed, and if no, it indicates that the current operating phase is still operating.
In one embodiment, the engine execution module 5200, when calling the functional component from the executable file to run the application, may further include: and loading and displaying the corresponding running interface according to the running interface information of the corresponding functional component configured by the executable file. For example, the run interface information can be configured in a second file that configures the functional component. Through the operation interface, an application user can see the operation condition of the corresponding functional component, and related information, state, data and the like can be inquired through an entrance and/or an interface provided by the operation interface.
Taking an online pre-estimation service application as an example, the following describes a process of the engine device 5000 running the application, according to the definition of an executable file of the application, as shown in fig. 7, a scenario of the application involves four running phases, namely a data processing phase, a scenario exploration phase, a model iteration phase and a model application phase, each phase includes at least one functional component, for example, the data processing phase includes a data processing component, the scenario exploration phase includes a scenario exploration component, the model iteration phase includes a self-learning component, the model application phase includes an online pre-estimation component, and the like, and the process may include, for example:
1) the engine management module 5100 checks whether the executable file can be recognized, and in case of recognition, decompresses the executable file of the application, and saves the decompressed first file and second file in the temporary directory area.
2) The engine execution module 5200 initializes the engine execution module 5200 based on the temporary target zone file (including the first file and the second file), and obtains an engine instance module 5200' capable of running the application.
3) The engine instance module 5200 'notifies the functional components to perform initialization according to the operation information of the functional components defined in the executable file, and after the functional components complete initialization, the engine instance module 5200' pushes the configuration information of each functional component to the corresponding functional component. The engine instance module 5200' reports its own status to the engine management module 5100 after completion of the startup, and the engine management module 5100 sets the operating status of the engine device 5000 to "running".
4) When the engine instance module 5200' operates, the operating condition information of the four operating phases is determined to be: the scheme exploration phase runs after the data processing phase is completed, the model iteration phase runs after the scheme exploration phase is completed, and the model application phase provides pre-estimation service based on the output of the model iteration phase.
5) The flow coordination module of the engine instance module 5200 'invokes the data processing component in the data processing stage to import data according to the schema definition, and the data processing component will trigger a data import completion message after completing data import and report the completion to the engine instance module 5200'.
The engine instance module 5200' performs corresponding stage processing at the data processing stage including: monitoring or regularly calling an interface between the data processing component through the component lifecycle management module 5202; updating the operation information of the stage through the flow coordination module 5201; updating the runtime environment context (context) at this stage by the runtime environment context management module 5204; the change information of this stage is issued by the message management module 5203; an entry point (EntryPoint) logic, a binding trigger event and the like are set through the entry point management module 5205; and, the stage outcome of this stage is handled by the flow coordination module 5201, etc.
6) The engine instance module 5200' checks whether the data processing stage has been completed, and if so, sets the output of the stage to the imported data set address, according to the definition of the executable file.
7) The engine instance module 5200' enters a scenario exploration phase after the inspection data processing phase is completed, and in the scenario exploration phase, automatic scenario exploration is performed by calling a scenario exploration component, scenario investigation is completed, and after the initial model is generated by the investigation, the scenario exploration component triggers an exploration completion message and reports the generated model to the engine instance module.
8) After the engine instance module 5200' checks that the scheme exploration phase is completed, the engine instance module enters a model iteration phase, in the model iteration phase, self-learning training is performed by calling the self-learning component according to the model generated in the scheme exploration phase, and after the updated model is generated by training, the self-learning component triggers a self-learning completion message and reports the generated updated model to the engine instance module.
9) After the model iteration stage is completed, the engine instance module 5200 'enters a model application stage, and in the model application stage, the engine instance module calls the updated model generated in the model iteration stage to provide the pre-estimation service, and reports the state of executing the pre-estimation service to the engine instance module 5200'.
< media examples >
In one embodiment, a computer-readable storage medium is also provided, wherein the computer-readable storage medium stores a computer program readable and executable by a computer, the computer program being configured to perform a method of building an artificial intelligence application according to any method embodiment when the computer program is read and executed by the computer.
The present invention may be an apparatus, method and/or computer program product. The computer program product may include a computer-readable storage medium having computer-readable program instructions embodied therewith for causing a processor to implement various aspects of the present invention.
The computer readable storage medium may be a tangible device that can hold and store the instructions for use by the instruction execution device. The computer readable storage medium may be, for example, but not limited to, an electronic memory device, a magnetic memory device, an optical memory device, an electromagnetic memory device, a semiconductor memory device, or any suitable combination of the foregoing. More specific examples (a non-exhaustive list) of the computer readable storage medium would include the following: a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a Static Random Access Memory (SRAM), a portable compact disc read-only memory (CD-ROM), a Digital Versatile Disc (DVD), a memory stick, a floppy disk, a mechanical coding device, such as punch cards or in-groove projection structures having instructions stored thereon, and any suitable combination of the foregoing. Computer-readable storage media as used herein is not to be construed as transitory signals per se, such as radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating through a waveguide or other transmission medium (e.g., optical pulses through a fiber optic cable), or electrical signals transmitted through electrical wires.
The computer-readable program instructions described herein may be downloaded from a computer-readable storage medium to a respective computing/processing device, or to an external computer or external storage device via a network, such as the internet, a local area network, a wide area network, and/or a wireless network. The network may include copper transmission cables, fiber optic transmission, wireless transmission, routers, firewalls, switches, gateway computers and/or edge servers. The network adapter card or network interface in each computing/processing device receives computer-readable program instructions from the network and forwards the computer-readable program instructions for storage in a computer-readable storage medium in the respective computing/processing device.
The computer program instructions for carrying out operations of the present invention may be assembler instructions, Instruction Set Architecture (ISA) instructions, machine-related instructions, microcode, firmware instructions, state setting data, or source or object code written in any combination of one or more programming languages, including an object oriented programming language such as Smalltalk, C + + or the like and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The computer-readable program instructions may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or the connection may be made to an external computer (for example, through the Internet using an Internet service provider). In some embodiments, aspects of the present invention are implemented by personalizing an electronic circuit, such as a programmable logic circuit, a Field Programmable Gate Array (FPGA), or a Programmable Logic Array (PLA), with state information of computer-readable program instructions, which can execute the computer-readable program instructions.
Aspects of the present invention are described herein with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and/or block diagrams, and combinations of blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer-readable program instructions.
These computer-readable program instructions may be provided to a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions/acts specified in the flowchart and/or block diagram block or blocks. These computer-readable program instructions may also be stored in a computer-readable storage medium that can direct a computer, programmable data processing apparatus, and/or other devices to function in a particular manner, such that the computer-readable medium storing the instructions comprises an article of manufacture including instructions which implement the function/act specified in the flowchart and/or block diagram block or blocks.
The computer readable program instructions may also be loaded onto a computer, other programmable data processing apparatus, or other devices to cause a series of operational steps to be performed on the computer, other programmable apparatus or other devices to produce a computer implemented process such that the instructions which execute on the computer, other programmable apparatus or other devices implement the functions/acts specified in the flowchart and/or block diagram block or blocks.
The flowchart and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present invention. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of instructions, which comprises one or more executable instructions for implementing the specified logical function(s). In some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions. It is well known to those skilled in the art that implementation by hardware, by software, and by a combination of software and hardware are equivalent.
Having described embodiments of the present invention, the foregoing description is intended to be exemplary, not exhaustive, and not limited to the embodiments disclosed. Many modifications and variations will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The terminology used herein is chosen in order to best explain the principles of the embodiments, the practical application, or improvements made to the technology in the marketplace, or to enable others of ordinary skill in the art to understand the embodiments disclosed herein. The scope of the invention is defined by the appended claims.

Claims (10)

1. A method of building an artificial intelligence application, comprising:
acquiring application configuration information configured for the application, wherein the application configuration information comprises scheme information for realizing an artificial intelligence scheme of the application;
and generating an executable file of the application according to the application configuration information and a preset template file.
2. The method of claim 1, wherein the recipe information includes task content information for the recipe.
3. The method of claim 2, wherein the task content information includes content information of each task item and association information between the respective task items.
4. The method of claim 2, wherein the task content information includes functional components used by the solution and running information of each of the functional components;
the application configuration information further includes component configuration information configured for each of the functional components, respectively.
5. The method according to claim 4, wherein the operation information includes at least one of operation phase information to which the corresponding functional component belongs in the scheme, initialization information of the corresponding functional component, and operation condition information of the corresponding functional component.
6. The method of claim 4, wherein generating the executable file of the application according to the application configuration information and a preset template file comprises:
generating a first file defining the scheme according to the scheme information and the scheme template file;
aiming at each functional component, generating a second file defining the corresponding functional component according to the component configuration information of the corresponding functional component and the component template file of the corresponding functional component;
and packaging the first file and each second file to generate the executable file.
7. An apparatus for building artificial intelligence applications, comprising:
the application configuration module is used for acquiring application configuration information configured for the application, wherein the application configuration information comprises scheme information for realizing an artificial intelligence scheme of the application; and the number of the first and second groups,
and the file generation module is used for generating the executable file of the application according to the application configuration information and a preset template file.
8. An apparatus comprising at least one computing device and at least one storage device, wherein the at least one storage device is to store instructions for controlling the at least one computing device to perform the method of any of claims 1 to 6; alternatively, the apparatus implements the apparatus of claim 7 through the computing device and the storage device.
9. A computer-readable storage medium, on which a computer program is stored which, when being executed by a processor, carries out the method according to any one of claims 1 to 6.
10. A system for running an artificial intelligence scheme comprising engine means and the means for constructing an artificial intelligence application of claim 7, wherein,
the device for constructing the artificial intelligence application is used for generating an executable file of the application;
the engine device is used for running the application according to the executable file.
CN201911364703.1A 2019-12-26 2019-12-26 Method, device and system for constructing artificial intelligence application Pending CN111158651A (en)

Priority Applications (2)

Application Number Priority Date Filing Date Title
CN201911364703.1A CN111158651A (en) 2019-12-26 2019-12-26 Method, device and system for constructing artificial intelligence application
PCT/CN2020/139466 WO2021129812A1 (en) 2019-12-26 2020-12-25 Method and system for running artificial intelligence application, and device

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201911364703.1A CN111158651A (en) 2019-12-26 2019-12-26 Method, device and system for constructing artificial intelligence application

Publications (1)

Publication Number Publication Date
CN111158651A true CN111158651A (en) 2020-05-15

Family

ID=70558058

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201911364703.1A Pending CN111158651A (en) 2019-12-26 2019-12-26 Method, device and system for constructing artificial intelligence application

Country Status (1)

Country Link
CN (1) CN111158651A (en)

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2021129812A1 (en) * 2019-12-26 2021-07-01 第四范式(北京)技术有限公司 Method and system for running artificial intelligence application, and device
WO2022037627A1 (en) * 2020-08-20 2022-02-24 第四范式(北京)技术有限公司 Integrated management method, apparatus and system for artificial intelligence applications
CN115964027A (en) * 2023-03-16 2023-04-14 杭州实在智能科技有限公司 Desktop embedded RPA flow configuration system and method based on artificial intelligence
CN116107561A (en) * 2023-04-14 2023-05-12 湖南云畅网络科技有限公司 Low-code-based action node rapid construction method, system and storage medium
CN116561075A (en) * 2023-05-15 2023-08-08 北京百度网讯科技有限公司 Method for generating dynamic link library file, method and device for calling operator

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20080127052A1 (en) * 2006-09-08 2008-05-29 Sap Ag Visually exposing data services to analysts
CN103425778A (en) * 2013-08-16 2013-12-04 国家电网公司 Intelligent development platform for database application system
CN103543993A (en) * 2012-07-11 2014-01-29 深圳市华高软件科技有限公司 Application program building method and system
CN104216691A (en) * 2013-05-31 2014-12-17 华为技术有限公司 Application creating method and device
CN107908399A (en) * 2017-11-14 2018-04-13 山东浪潮通软信息科技有限公司 A kind of page development method and device

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20080127052A1 (en) * 2006-09-08 2008-05-29 Sap Ag Visually exposing data services to analysts
CN103543993A (en) * 2012-07-11 2014-01-29 深圳市华高软件科技有限公司 Application program building method and system
CN104216691A (en) * 2013-05-31 2014-12-17 华为技术有限公司 Application creating method and device
CN103425778A (en) * 2013-08-16 2013-12-04 国家电网公司 Intelligent development platform for database application system
CN107908399A (en) * 2017-11-14 2018-04-13 山东浪潮通软信息科技有限公司 A kind of page development method and device

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2021129812A1 (en) * 2019-12-26 2021-07-01 第四范式(北京)技术有限公司 Method and system for running artificial intelligence application, and device
WO2022037627A1 (en) * 2020-08-20 2022-02-24 第四范式(北京)技术有限公司 Integrated management method, apparatus and system for artificial intelligence applications
CN115964027A (en) * 2023-03-16 2023-04-14 杭州实在智能科技有限公司 Desktop embedded RPA flow configuration system and method based on artificial intelligence
CN116107561A (en) * 2023-04-14 2023-05-12 湖南云畅网络科技有限公司 Low-code-based action node rapid construction method, system and storage medium
CN116561075A (en) * 2023-05-15 2023-08-08 北京百度网讯科技有限公司 Method for generating dynamic link library file, method and device for calling operator
CN116561075B (en) * 2023-05-15 2024-01-30 北京百度网讯科技有限公司 Method for generating dynamic link library file, method and device for calling operator

Similar Documents

Publication Publication Date Title
CN111158651A (en) Method, device and system for constructing artificial intelligence application
EP3605324B1 (en) Application development method and tool, and storage medium thereof
KR102092722B1 (en) Tools and methods for real-time dataflow programming language
WO2018228211A1 (en) Application conversion method, apparatus and device
US20130226555A1 (en) Multilingual Build Integration for Compiled Applications
US10891115B2 (en) Model configuration using partial model data
US11755461B2 (en) Asynchronous consumer-driven contract testing in micro service architecture
US20210208854A1 (en) System and method for enhancing component based development models with auto-wiring
CN111158797A (en) Method, system and engine device for operating artificial intelligence application
US20200326914A1 (en) Creating an app method and system
US10572278B2 (en) Smart controls for user interface design and implementation
JP2021170335A (en) Application construction method, device, electronic facility, storage medium, and program
CN114064024A (en) Method, apparatus, device, storage medium, and program product for developing micro-application
WO2021129812A1 (en) Method and system for running artificial intelligence application, and device
US9996344B2 (en) Customized runtime environment
CN101295382A (en) Method and device for implementing service logical relation between business object properties
Miravet et al. Framework for the declarative implementation of native mobile applications
CN111435328B (en) Application testing method and device, electronic equipment and readable storage medium
JP7014960B2 (en) Information processing equipment, servers, their processing methods and programs
CN115469849B (en) Service processing system, method, electronic equipment and storage medium
US11061664B2 (en) Code management system and code management method
CN111443905B (en) Service data processing method, device and system and electronic equipment
CN113703638A (en) Data management page processing method and device, electronic equipment and storage medium
CN112882698A (en) Development environment generation method and device, computer storage medium and electronic device
US10802844B2 (en) Distributed application based off of stateful desktop application

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination