Unstructured Text to Rules Engine Conversion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The current process of creating rules engine rules from regulatory authority and airline text documents is manual, time-consuming, expensive, and error-prone, requiring manual auditing and resulting in non-certifiable rules, especially when union contracts are renegotiated, leading to significant time and expense, and potential misunderstandings of pilot and cabin crew legality rules.
Innovation Solution
A method that converts unstructured text from diverse sources into structured language using tools like Annotation Query Language (AQL), allowing for automated generation and certification of rules engine rules, thereby reducing manual effort and increasing processing speed by limiting information to be processed.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If manual process is used to create rules engine rules from unstructured text documents, then flexibility in handling diverse document formats is maintained, but processing time and cost increase significantly
Solution Approach 1:
The patent introduces an intermediary structured language (AQL) that serves as a bridge between unstructured text documents and the rules engine. The conversion process transforms diverse unstructured documents into a standardized structured format, which then feeds into the rules engine. This intermediary layer enables automated processing while maintaining flexibility in handling different document sources and formats.
2Reliability
If all available unstructured text documents are processed to generate comprehensive rules, then completeness of rule coverage is improved, but time and expense increase significantly
Solution Approach 1:
The patent extracts and processes only the relevant portions of unstructured text documents that are necessary for rule generation. Rather than processing entire documents, the system identifies and extracts specific information elements, conditions, and parameters needed for rules engine rules. This selective extraction reduces processing time while maintaining rule completeness.
3Measurement precision
If manual auditing is performed to validate regulatory rules, then accuracy and certification capability are improved, but time and cost increase significantly
Solution Approach 1:
The patent implements self-service mechanisms where the system automatically validates and audits its own generated rules against the source documents and regulatory requirements. The structured language format enables automated verification of rule accuracy, eliminating or reducing the need for manual auditing while maintaining high precision and certification capability.
4Productivity
If unstructured text is directly converted to rules engine rules without structured language intermediate, then processing steps are reduced, but conversion accuracy and reliability decrease
Solution Approach 1:
The patent segments the conversion process into distinct stages: parsing unstructured text, extracting information elements, translating to structured language (AQL), and finally generating rules engine rules. This segmentation allows each stage to be optimized independently, improving both conversion efficiency and accuracy. The structured language intermediate ensures precise representation of regulatory requirements before final rule generation.
Data Source
AI summary
A method of manufacturing an aircraft. A plurality of documents, each comprising corresponding unstructured text and each being from a different source, is received at a processor, a particular document being in the plurality of documents. The particular document is selected by the processor based on a policy. The corresponding unstructured text for the particular document is converted by the processor into a structured language. The structured language is compiled by the processor into rules for a rules engine comprising software stored on a non-transitory computer readable storage medium. The rules for the rules engine are parsed by the processor into a set of rules. The aircraft is constructed according to the set of rules.