Dynamic low-code page recommendation method and system

Through the dynamic low-code page recommendation method, and the automatic recommendation templates are used to use clustering and LRU algorithms to solve the problem of repetitive work in page construction in low-code platforms, improve development efficiency and user experience, and improve the competitiveness of low-code platforms.

CN119025111APending Publication Date: 2024-11-26SHANDONG INSPUR SCI RES INST CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411181148.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-27
Publication Date
2024-11-26

AI Technical Summary

Technical Problem

The low-code development platform has repetitive work when dragging a large number of pages, which affects development efficiency, especially layout adjustments, resulting in inefficiency.

Method used

Using dynamic low-code page recommendation method, using machine learning clustering algorithms (such as KMeans) and least recent use (LRU) algorithms, automatically save user-created pages as templates, and recommend rankings of the highest and most recent based on similarity and frequency of use. Use templates to reduce duplicate work.

Benefits of technology

It improves the development efficiency of low-code platforms, improves user experience and product competitiveness, reduces the repetitive work of users during page construction, and improves the efficiency of page construction.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119025111A_ABST
    Figure CN119025111A_ABST
Patent Text Reader

Abstract

The invention discloses a dynamic low-code page recommendation method and system, and belongs to the technical field of computer software and machine learning, pages created by a user are stored as templates, clustering is performed by using a clustering algorithm in machine learning, n templates with the highest ranking are searched, and the templates are recommended to the user; in addition, according to an LRU algorithm, a template which is used recently by the user is recommended; the method comprises the steps that after a user creates a page, a Dom tree arranged according to the page is converted into a json structure, content in the layout is converted into display models, and each layout corresponds to one display model; after being stored, the template is rendered into a front-end thumbnail for a user to select a corresponding template; using a KMeans clustering algorithm to extract a template of a page created by a user, and then performing clustering calculation again; and realizing an LRU algorithm: recommending according to a template recently used by a user. The development efficiency of the low-code platform can be further improved, and the competitiveness of low-code platform products is improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of computer software and machine learning technology, and specifically to a dynamic low-code page recommendation method and system. Background Art

[0002] By deploying advanced visual development tools, low-code development platforms greatly reduce the complexity of application development, allowing users with weak technical backgrounds or no background to easily design and build applications.

[0003] However, when dragging a large number of pages, there is a lot of repetitive work to be done, especially some layouts, which need to be re-dragged, affecting development efficiency. Summary of the invention

[0004] The technical task of the present invention is to address the above shortcomings and provide a dynamic low-code page recommendation method and system, which can further improve the development efficiency of the low-code platform and enhance the competitiveness of the low-code platform products.

[0005] The technical solution adopted by the present invention to solve its technical problem is:

[0006] A dynamic low-code page recommendation method saves the pages created by users as templates, uses the clustering algorithm in machine learning to cluster, finds the n highest-ranked templates, and recommends them to users; in addition, based on the LRU algorithm, recommends the templates that users have used most recently;

[0007] The implementation of this method includes:

[0008] 1) Convert page to template: After the user creates a page, the Dom tree of the page layout is converted into a json structure, and the content in the layout is converted into a display model. Each layout corresponds to a display model.

[0009] 2) Visualization module implementation: After the template is saved, it is rendered as a front-end thumbnail for users to select the corresponding template;

[0010] 3) Clustering implementation: Use the KMeans clustering algorithm to extract the template of the user-created page and then re-cluster the calculation;

[0011] 4) LRU algorithm implementation: recommend templates based on the user’s most recently used templates.

[0012] Furthermore, the page is converted to a template, and the content pattern of the json structure is: [{layoutName:layout1,layouttype:type1,content:[{widgetName:form1,wid getId:3,widgetCode:xxx,widgetColumn}]},{layoutName:layoout2,…}], in which the page layout structure is displayed in the outermost layer of the json. Each layer of the layout structure can nest layouts and page components, and a maximum of three layers of nesting are allowed.

[0013] Furthermore, the visualization module realizes that after the page is saved, a corresponding template is generated, rendered as a picture and uploaded to the file server. When the user creates a page, the corresponding picture is downloaded for reference.

[0014] Furthermore, the clustering implementation specifically includes:

[0015] 3.1) Layout encoding, counting the current layout type and encoding it using bit data;

[0016] 3.2) Component encoding: count the current page components and perform data encoding;

[0017] 3.3) Display template coding, uniformly give coding to page components in the layout, and store component coding in the display template for a series of organization;

[0018] 3.4) The maximum number of outermost layouts is limited to 4, the maximum depth of layouts is limited to 3, and style layout encoding is unified;

[0019] 3.5) Define the similarity of visualization templates, use cosine distance to represent the similarity between two templates, use Kmeans algorithm to classify, and rank according to quantity.

[0020] Furthermore, the layout code is encoded using 8 bits, 00000000~01111111;

[0021] The component code is encoded using 8 bits, 00000000~01111111;

[0022] The display template encoding is represented by 16 bits.

[0023] Furthermore, the unified style layout encoding is performed;

[0024] Using binary 000000010000000000000000000000000,

[0025] Indicates that there is only one layout at the outermost level;

[0026] Binary concatenation 00000001000000021000000110000011,

[0027] Indicates a nested layout within a layout, and a page display template;

[0028] Binary concatenation 10000001100000110000000000000000,

[0029] Indicates that there are two display templates under the upper layout component.

[0030] Furthermore, the LRU algorithm implements that after a user creates a new page, the latest template creation time of the page is recorded and updated, and the number of creation times of the page template is accumulated; when the threshold given by the recommended template is exceeded, the template with the longest time and the least number of times used is deleted, and a new template is added for use.

[0031] The present invention also claims protection for a dynamic low-code page recommendation system, comprising:

[0032] Implement the page conversion template module, which is used to convert the Dom tree of the page layout into a json structure after the user creates a page, and convert the content in the layout into a display model. Each layout corresponds to a display model.

[0033] The visualization module is used to render the template as a front-end thumbnail after it is saved, so that the user can select the corresponding template;

[0034] The clustering module uses the KMeans clustering algorithm to extract the template of the user-created page and then re-cluster the calculation;

[0035] The LRU algorithm module is used to recommend templates that users have recently used.

[0036] The system specifically implements page recommendation through the above-mentioned dynamic low-code page recommendation method.

[0037] The present invention also claims a dynamic low-code page recommendation device, comprising: at least one memory and at least one processor;

[0038] The at least one memory is used to store a machine-readable program;

[0039] The at least one processor is used to call the machine-readable program to implement the above method.

[0040] The present invention also claims protection for a computer-readable medium having computer instructions stored thereon, which, when executed by a processor, causes the processor to perform the above method.

[0041] Compared with the prior art, the dynamic low-code page recommendation method and system of the present invention has the following advantages:

[0042] Beneficial effects:

[0043] For applications developed using low-code, the application pages are similar and there are not many types of pages. The present invention automatically saves the pages created by users as templates, uses the clustering algorithm in machine learning to cluster, finds the n highest-ranked templates, and recommends them to users. According to the LRU algorithm, the templates that users have used most recently are recommended. Improve the user experience. Further improve the development efficiency of the low-code platform and enhance the competitiveness of the low-code platform products. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 It is a flowchart of a dynamic low-code page recommendation method provided by an embodiment of the present invention;

[0045] Figure 2 It is a diagram illustrating the clustering implementation process provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0046] The present invention will be further described below in conjunction with specific embodiments.

[0047] An embodiment of the present invention provides a dynamic low-code page recommendation method, which saves the pages created by users as templates, uses the clustering algorithm in machine learning to cluster them, finds the n highest-ranked templates, and recommends them to users; in addition, based on the LRU algorithm, recommends the templates that users have used most recently.

[0048] The implementation of this method includes:

[0049] 1. Convert page to template: After the user creates a page, the Dom tree of the page layout is converted into a json structure, and the content in the layout is converted into a display model. Each layout corresponds to a display model.

[0050] 2. Visualization module implementation: After the template is saved, it is rendered as a front-end thumbnail for users to select the corresponding template;

[0051] 3. Clustering implementation: Use KMeans clustering algorithm to extract the template of the user-created page and then re-cluster the calculation;

[0052] 4. LRU algorithm implementation: recommend based on the template most recently used by the user.

[0053] The specific implementation process is as follows:

[0054] To implement page conversion template:

[0055] After the user creates a page, the Dom tree of the page layout is converted into a json structure, and the content in the layout is converted into a display model. Each layout corresponds to a display model. The content mode of the json structure is: [{layoutName:layout1,layouttype:type1,content:[{widgetName:form1,wid getId:3,widgetCode:xxx,widgetColumn}]},{layoutName:layout2,…}], in the outermost layer of the json, the page layout structure is displayed, and each layer of the layout structure can nest layouts and page components, and the layout allows up to three layers of nesting.

[0056] Visualization module implementation:

[0057] After the page is saved, a corresponding template is generated, rendered as an image and uploaded to the file server. When the user creates a page, the corresponding image is downloaded for reference.

[0058] Clustering implementation:

[0059] (1) Layout encoding: count the current layout type and use bit data for encoding; use 8 bits for encoding, 00000000~01111111.

[0060] (2) Component encoding: count the components of the current page and perform data encoding in the same way; use 8 bits for encoding, 00000000~01111111.

[0061] (3) Display template encoding: Uniformly give encoding to page components within the layout, using 16 bits to represent it. The component encoding is stored in the display template for a series of organization.

[0062] (4) The maximum number of outermost layouts is limited to 4, the maximum depth of layouts is limited to 3, and style layout encoding is performed uniformly.

[0063] Using binary 000000010000000000000000000000000,

[0064] Indicates that there is only one layout at the outermost level;

[0065] Binary concatenation 00000001000000021000000110000011,

[0066] Indicates a nested layout within a layout, and a page display template;

[0067] Binary concatenation 10000001100000110000000000000000,

[0068] Indicates that there are two display templates under the upper layout component.

[0069] (5) Define the similarity of visualization templates, use cosine distance to represent the similarity between two templates, use Kmeans algorithm to classify, and rank according to quantity.

[0070] LRU algorithm implementation:

[0071] After a user creates a new page, the latest template creation time of the page is recorded and updated, and the number of page template creations is accumulated; if the threshold given by the recommended template is exceeded, the template with the longest time and the least number of uses is deleted, and a new template is added for use.

[0072] The embodiment of the present invention also provides a dynamic low-code page recommendation system, including:

[0073] Implement the page conversion template module, which is used to convert the Dom tree of the page layout into a json structure after the user creates a page, and convert the content in the layout into a display model. Each layout corresponds to a display model.

[0074] The visualization module is used to render the template as a front-end thumbnail after it is saved, so that the user can select the corresponding template;

[0075] The clustering module uses the KMeans clustering algorithm to extract the template of the user-created page and then re-cluster the calculation;

[0076] The LRU algorithm module is used to recommend templates that users have recently used.

[0077] The system specifically implements page recommendation through the dynamic low-code page recommendation method described in the above embodiment:

[0078] The page conversion template module is implemented.

[0079] After the user creates a page, the Dom tree of the page layout is converted into a json structure, and the content in the layout is converted into a display model. Each layout corresponds to a display model. The content mode of the json structure is: [{layoutName:layout1,layouttype:type1,content:[{widgetName:form1,wid getId:3,widgetCode:xxx,widgetColumn}]},{layoutName:layout2,…}], in the outermost layer of the json, the page layout structure is displayed, and each layer of the layout structure can nest layouts and page components, and the layout allows up to three layers of nesting.

[0080] The visualization module,

[0081] After the page is saved, a corresponding template is generated, rendered as an image and uploaded to the file server. When the user creates a page, the corresponding image is downloaded for reference.

[0082] The clustering module is implemented as follows:

[0083] (1) Layout encoding: count the current layout type and use bit data for encoding; use 8 bits for encoding, 00000000~01111111.

[0084] (2) Component encoding: count the components of the current page and perform data encoding in the same way; use 8 bits for encoding, 00000000~01111111.

[0085] (3) Display template encoding: Uniformly give encoding to page components within the layout, using 16 bits to represent it. The component encoding is stored in the display template for a series of organization.

[0086] (4) The maximum number of outermost layouts is limited to 4, the maximum depth of layouts is limited to 3, and style layout encoding is performed uniformly.

[0087] Using binary 000000010000000000000000000000000,

[0088] Indicates that there is only one layout at the outermost level;

[0089] Binary concatenation 00000001000000021000000110000011,

[0090] Indicates a nested layout within a layout, and a page display template;

[0091] Binary concatenation 10000001100000110000000000000000,

[0092] Indicates that there are two display templates under the upper layout component.

[0093] (5) Define the similarity of visualization templates, use cosine distance to represent the similarity between two templates, use Kmeans algorithm to classify, and rank according to quantity.

[0094] The LRU algorithm module,

[0095] After a user creates a new page, the latest template creation time of the page is recorded and updated, and the number of page template creations is accumulated; if the threshold given by the recommended template is exceeded, the template with the longest time and the least number of uses is deleted, and a new template is added for use.

[0096] An embodiment of the present invention also provides a dynamic low-code page recommendation device, comprising: at least one memory and at least one processor;

[0097] The at least one memory is used to store a machine-readable program;

[0098] The at least one processor is used to call the machine-readable program to implement the dynamic low-code page recommendation method described in the above embodiment.

[0099] The embodiment of the present invention also provides a computer-readable medium, on which computer instructions are stored, and when the computer instructions are executed by a processor, the processor executes the dynamic low-code page recommendation method described in the above embodiment. Specifically, a system or device equipped with a storage medium can be provided, on which software program codes that implement the functions of any of the above embodiments are stored, and a computer (or CPU or MPU) of the system or device reads and executes the program code stored in the storage medium.

[0100] In this case, the program code itself read from the storage medium can realize the function of any one of the above-mentioned embodiments, and thus the program code and the storage medium storing the program code constitute a part of the present invention.

[0101] The storage medium embodiments for providing the program code include a floppy disk, a hard disk, a magneto-optical disk, an optical disk (such as CD-ROM, CD-R, CD-RW, DVD-ROM, DVD-RAM, DVD-RW, DVD+RW), a magnetic tape, a non-volatile memory card, and a ROM. Alternatively, the program code can be downloaded from a server computer by a communication network.

[0102] In addition, it should be clear that the functions of any of the above embodiments can be implemented not only by executing the program code read by the computer, but also by enabling an operating system operating on the computer to complete part or all of the actual operations based on instructions from the program code.

[0103] In addition, it can be understood that the program code read from the storage medium is written to a memory provided in an expansion board inserted into the computer or written to a memory provided in an expansion unit connected to the computer, and then based on the instructions of the program code, a CPU installed on the expansion board or the expansion unit is enabled to perform part or all of the actual operations, thereby realizing the functions of any of the above-mentioned embodiments.

[0104] The present invention is shown and described in detail above through the accompanying drawings and preferred embodiments. However, the present invention is not limited to these disclosed embodiments. Based on the above multiple embodiments, those skilled in the art can know that the code review methods in the above different embodiments can be combined to obtain more embodiments of the present invention, and these embodiments are also within the protection scope of the present invention.

Claims

1. A dynamic low-code page recommendation method, characterized in that: Save the pages created by users as templates, use the clustering algorithm in machine learning to cluster them, find the n highest-ranked templates, and recommend them to users; In addition, according to the LRU algorithm, the templates most recently used by the user are recommended; The implementation of this method includes: 1) Convert page to template: After the user creates a page, the Dom tree of the page layout is converted into a json structure, and the content in the layout is converted into a display model. Each layout corresponds to a display model. 2) Visualization module implementation: After the template is saved, it is rendered as a front-end thumbnail for users to select the corresponding template; 3) Clustering implementation: Use the KMeans clustering algorithm to extract the template of the user-created page and then re-cluster the calculation; 4) LRU algorithm implementation: recommend templates based on the user’s most recently used templates.

2. A dynamic low-code page recommendation method according to claim 1, characterized in that: The page is converted to a template, and the content pattern of the json structure is: [{layoutName:layout1,layouttype:type1,content:[{widgetName:form1,wid getId:3,widgetCode:xxx,widgetColumn}]},{layoutName:layoout2,…}], in which the page layout structure is displayed in the outermost layer of the json. Each layer of the layout structure can nest layouts and page components, and a maximum of three layers of nesting are allowed.

3. A dynamic low-code page recommendation method according to claim 1, characterized in that: The visualization module realizes that after the page is saved, a corresponding template is generated, rendered as a picture and uploaded to the file server. When the user creates a page, the corresponding picture is downloaded for reference by the user.

4. A dynamic low-code page recommendation method according to claim 1, characterized in that: The clustering implementation specifically includes: 3.1) Layout encoding, counting the current layout type and encoding it using bit data; 3.2) Component encoding: count the current page components and perform data encoding; 3.3) Display template coding, uniformly give coding to page components in the layout, and store component coding in the display template for a series of organization; 3.4) The maximum number of outermost layouts is limited to 4, the maximum depth of layouts is limited to 3, and style layout encoding is unified; 3.5) Define the similarity of visualization templates, use cosine distance to represent the similarity between two templates, use Kmeans algorithm to classify, and rank according to quantity.

5. A dynamic low-code page recommendation method according to claim 4, characterized in that: The layout code is encoded using 8 bits, 00000000~01111111; The component code is encoded using 8 bits, 00000000~01111111; The display template encoding is represented by 16 bits.

6. A dynamic low-code page recommendation method according to claim 4 or 5, characterized in that: The unified style layout encoding; Using binary 000000010000000000000000000000000, Indicates that there is only one layout at the outermost level; Binary concatenation 00000001000000021000000110000011, Indicates a nested layout within a layout, and a page display template; Binary concatenation 10000001100000110000000000000000, Indicates that there are two display templates under the upper layout component.

7. A dynamic low-code page recommendation method according to claim 1, characterized in that: The LRU algorithm is implemented such that after a user creates a new page, the latest template creation time of the page is recorded and updated, and the number of creation times of the page template is accumulated; when the threshold given by the recommended template is exceeded, the template with the longest time and the least number of times used is deleted, and a new template is added for use.

8. A dynamic low-code page recommendation system, characterized in that: include: Implement the page conversion template module, which is used to convert the Dom tree of the page layout into a json structure after the user creates a page, and convert the content in the layout into a display model. Each layout corresponds to a display model. The visualization module is used to render the template as a front-end thumbnail after it is saved, so that the user can select the corresponding template; The clustering module uses the KMeans clustering algorithm to extract the template of the user-created page and then re-cluster the calculation; The LRU algorithm module is used to recommend templates that users have recently used. The system specifically implements page recommendation through the dynamic low-code page recommendation method described in any one of claims 1 to 7.

9. A dynamic low-code page recommendation device, characterized in that: include: at least one memory and at least one processor; The at least one memory is used to store a machine-readable program; The at least one processor is used to call the machine-readable program to implement the method described in any one of claims 1 to 7.

10. A computer-readable medium, characterized in that The computer readable medium stores computer instructions, which, when executed by a processor, cause the processor to execute the method according to any one of claims 1 to 7.