Automated Database Schema Generation via Vector Space Field Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current WYSIWYG editors for database schema generation are limited in automatically creating and modifying schemas, requiring predefined templates and relying on database programmers, and do not effectively handle user-generated fields or relationships, especially in semi-structured document databases like XML and JSON.
Innovation Solution
A system that uses a graphical user interface to infer information from user-generated fields, guide proper labeling, and analyze relationships between fields to generate database schemas, enabling integration with various database types, including relational, semi-structured, and NoSQL databases, through a process of field validation and vector space representation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If predefined templates are used for schema generation, then the system can provide automated schema creation, but it cannot handle user-generated fields outside the templates
Solution Approach 1:
The system performs self-service by automatically analyzing user-generated field labels, inferring their meanings through vector space representation, and generating appropriate database schema elements without requiring predefined templates or manual programmer intervention. The system serves itself by autonomously completing the entire schema generation process from raw user input.
Solution Approach 2:
The system changes the parameter of field representation from fixed template categories to dynamic vector space embeddings. By converting field labels into vector representations and analyzing their semantic relationships, the system adapts to handle any user-generated field regardless of whether it fits predefined templates, thus resolving the contradiction between automation and versatility.
2Extent of automation
If the system infers relationships from field labels, then it can automatically determine field relationships, but it struggles with ambiguously labeled fields
Solution Approach 1:
The system uses feedback mechanisms by analyzing the vector space relationships between field labels and continuously refining its understanding of field relationships. When fields are ambiguously labeled, the system can leverage contextual information from other fields in the form to infer correct relationships, effectively using feedback from the overall form structure to disambiguate individual field labels.
Solution Approach 2:
The system combines multiple sources of information (field labels, field positions, data types, and contextual relationships) to create a composite understanding of field relationships. This composite approach allows the system to accurately determine relationships even when individual field labels are ambiguous, by synthesizing information from multiple facets of the form structure.
3Ease of operation
If WYSIWYG editors generate forms with database connections, then they simplify form creation, but they lack integration with semi-structured and NoSQL databases
Solution Approach 1:
The system achieves universality by designing a database-agnostic schema generation approach. The vector space representation method and relationship inference mechanisms are applicable regardless of the target database type. The system can generate schemas for relational databases, semi-structured document databases like XML, and NoSQL databases, all through the same automated process, thus providing multi-functionality without compromising ease of operation.
Data Source
AI summary
A system and method of generating a database schema from a graphical user interface used to create a form. The embodiments discloses the system that utilizes a drag and drop application that allows for configuration of a plurality of forms. These forms can then be placed in a graphical flow that will dictate the order of the forms. Through its graphical user interface, the system is able to gather information on field structure, flow among form elements, element identification, among other embodiments. This information allows the system to automate the creation of the database schema without user intervention.