Search

EP-4738099-A1 - TECHNIQUES FOR GENERATING USER INTERFACES ASSISTED BY GENERATIVE PRE-TRAINED TRANSFORMER CODE GENERATOR

EP4738099A1EP 4738099 A1EP4738099 A1EP 4738099A1EP-4738099-A1

Abstract

In some implementations, the techniques may include accessing a data file containing metadata for various data fields. The techniques may include accessing a list of predefined categories and creating a first request to a generative pre-trained transformer to categorize portions of the metadata by assigning the portions of the metadata to predefined categories of the list of predefined categories. The techniques may include receiving a first output list from the generative pre-trained transformer comprising the portions of the metadata and the assigned predefined categories. The techniques may include generating labels, features and ordering using the list of categories. The techniques may include generating application code using one or more code generation templates and the labels and features for an application. The techniques may include storing the application code. The techniques may be performed by a system or stored as a series of instructions on a computer-readable tangible medium.

Inventors

  • IRELAND, EVAN

Assignees

  • SAP SE

Dates

Publication Date
20260506
Application Date
20250311

Claims (14)

  1. A computer implemented method comprising: accessing a data file containing metadata for various data fields; accessing a list of predefined categories; creating a first request to a generative pre-trained transformer to categorize portions of the metadata by assigning the portions of the metadata to predefined categories of the list of predefined categories; receiving a first output list from the generative pre-trained transformer comprising the portions of the metadata and the assigned predefined categories; generating labels, features and ordering using the list of categories; generating application code using one or more code generation templates and the labels and features for an application; and storing the application code.
  2. The computer implemented method of claim 1 further comprising: creating a second request to a generative pre-trained transformer to determine categories to filter business data from the data file using the list of predefined categories; receiving a second output list of proposed filters from the generative pre-trained transformer; generating one or more filters based on the second output list of proposed filters; generating application code using one or more code generation templates and the one or more filters for the application; and store the application code.
  3. The computer implemented method of claim 1 or 2 further comprising: creating a third request to a generative pre-trained transformer to determine categories to sort business data from the data file using the list of predefined categories; receiving a third output list of proposed sorting terms from the generative pre-trained transformer; generating one or more sorting features to sort the metadata based on the third output list of proposed sorting terms; generating application code using one or more code generation templates and the one or more sorting features for the application; and store the application code.
  4. The computer implemented method of any one of the preceding claims, wherein the data file comprises an extensible markup language file.
  5. The computer implemented method of any one of the preceding claims, wherein the data file comprises structures and objects in JavaScript Object Notation (JSON) format.
  6. The computer implemented method of any one of the preceding claims, further comprising assigning a color to a data element in a display for an application based at least in part on the list of categories specific to the data element.
  7. The computer implemented method of any one of the preceding claims, wherein the generating application code comprises: parsing the data file to determine placeholders and elements; generating a data model containing values to be substitute in the placeholders and the elements; serializing the data model into text; and applying transformations to the text to convert into a code format.
  8. A transitory or non-transitory computer-readable medium storing a set of instructions, the set of instructions comprising: one or more instructions that, when executed by one or more processors of a device, cause the device to perform operations according to any one of the preceding claims.
  9. A system comprising: one or more processors configured to access code stored in a memory to perform operations comprising: accessing a data file containing metadata for various data fields; accessing a list of predefined categories; creating a first request to a generative pre-trained transformer to categorize portions of the metadata by assigning the portions of the metadata to predefined categories of the list of predefined categories; receiving a first output list from the generative pre-trained transformer comprising the portions of the metadata and the assigned predefined categories; generating labels, features and ordering using the list of categories; generating application code using one or more code generation templates and the labels and features for an application; and storing the application code.
  10. The system of claim 9, wherein the operations further comprise: creating a second request to a generative pre-trained transformer to determine categories to filter business data from the data file using the list of predefined categories; receiving a second output list of proposed filters from the generative pre-trained transformer; generating one or more filters based on the second output list of proposed filters; generating application code using one or more code generation templates and the one or more filters for the application; and storing the application code.
  11. The system of claim 9 or 10, wherein the operations further comprise: creating a third request to a generative pre-trained transformer to determine categories to sort business data from the data file using the list of predefined categories; receiving a third output list of proposed sorting terms from the generative pre-trained transformer; generating one or more sorting features to sort the metadata based on the third output list of proposed sorting terms; generating application code using one or more code generation templates and the one or more sorting features for the application; and storing the application code.
  12. The system of any one of claims 9 to 11, wherein the data file comprises an extensible markup language file.
  13. The system of any one of claims 9 to 12, wherein the data file comprises structures and objects in JavaScript Object Notation (JSON) format.
  14. The system of any one of claims 9 to 13, wherein the generating application code comprises: parsing the data file to determine placeholders and elements generating a data model containing values to be substitute in the placeholders and the elements serializing the data model into text; and applying transformations to the text to convert into a code format.

Description

BACKGROUND Software applications have the capability to generate other software applications up to and including the generation of the specific program code for those applications. However, many of these applications that can be developed using these developer tools produce user interfaces that are not initiative to a user and are not organized and ordered in a way that is helpful to a user. The user interfaces for these applications may present data in a boilerplate fashion or using traditional rules such as displaying fields in alphabetical order. The data may be presented in such a way that is unreadable or illogical to a user. Further, editing these applications created using the developer tools can be an arduous task by developers to correct the deficiencies of these development tools. Application developers can use improved tools using artificial intelligence to assist in the development of more user-friendly applications, including user interfaces for those applications, using data and efficient ways to customize those applications. SUMMARY A system of one or more computers can be configured to perform particular operations or actions by virtue of having software, firmware, hardware, or a combination of them installed on the system that in operation causes the system to perform the actions. One or more computer programs can be configured to perform particular operations or actions by virtue of including instructions that, when executed by data processing apparatus, cause the apparatus to perform the actions. In one general aspect, a computer implemented method may include accessing a data file containing metadata for various data fields. The computer implemented method may include accessing a list of predefined categories. The method may include creating a first request to a generative pre-trained transformer to categorize portions of the metadata by assigning the portions of the metadata to predefined categories of the list of predefined categories. This method may include receiving a first output list from the generative pre-trained transformer comprising the portions of the metadata and the assigned predefined categories. The method may moreover include generating labels, features and ordering using the list of categories. The method may also include generating application code using one or more code generation templates and the labels and features for an application. The method may furthermore include storing the application code. Other embodiments of this aspect include corresponding computer systems, apparatus, and computer programs recorded on one or more computer storage devices, each configured to perform the actions of the methods. Implementations may include one or more of the following features. In various embodiments, the computer implemented method may include creating a second request to a generative pre-trained transformer to determine categories to filter business data from the data file using the list of predefined categories. The computer implemented method may include receiving a second output list of proposed filters from the generative pre-trained transformer. The computer implemented method may include generating one or more filters based on the second output list of proposed filters. The computer implemented method may include generating application code using one or more code generation templates and the one or more filters for the application. The computer implemented method may include generating the application code. The computer implemented method may include storing the application code. Computer implemented methods may include creating a third request to a generative pre-trained transformer to determine categories to sort business data from the data file using the list of predefined categories. The computer implemented method may include receiving a third output list of proposed sorting terms from the generative pre-trained transformer. The computer implemented method may include generating one or more sorting features to sort the metadata based on the third output list of proposed sorting terms. The computer implemented method may include generating application code using one or more code generation templates and the one or more sorting features for the application. The computer implemented method may include storing the application code. In various embodiments, the data file may include an extensible markup language file. In various embodiments, the data file may include structures and objects in JavaScript Object Notation (JSON) format. In various embodiments, the computer implemented method may include assigning a color to a data element in a display for an application based at least in part on the list of categories specific to the data element. In various embodiments, the process of generating application code may include parsing the data file to determine placeholders and elements. The process may include generating a data model containing values to be substitute