Automated Software Design Specification Generation from Natural Language

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current methods for software design from natural language requirements involve multiple interpretations and translations, leading to a loss of knowledge and deviations between requester descriptions and final products, as they require manual interpretation and translation by industry analysts and designers, resulting in inaccuracies and increased costs.

Innovation Solution

A system that automatically generates software design specification documents directly from natural language descriptions using a deterministic model that identifies conceptual components and applies analytical rules, eliminating the need for manual translation and ensuring complete and accurate design representation.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If manual interpretation and translation by industry analysts and designers is used, then human expertise and flexibility are utilized, but knowledge loss and deviations between requester descriptions and final products occur

Engineering Contradiction:
Improveaccuracy of design representationVSAvoidknowledge loss in translation
Core Design Contradiction:
ReliabilityVSLoss of information

Solution Approach 1:

The patent introduces an intermediary system comprising a parser, natural language processing module, and deterministic model that acts as a mediator between the requester's natural language description and the final software design. This intermediary automatically translates and structures the requirements without human intervention, preserving the original intent while converting it into formal design specifications.

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent replaces the mechanical human interpretation process with an automated computational system. The deterministic model and natural language processing algorithms substitute the manual analysis and translation performed by industry analysts and designers, eliminating the knowledge loss inherent in human-to-human communication chains.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

2Productivity

If manual interpretation and translation processes are used, then human expertise is applied, but processing time and costs increase

Engineering Contradiction:
Improvesoftware design speedVSAvoidtime for manual translation
Core Design Contradiction:
ProductivityVSLoss of time

Solution Approach 1:

The system enables self-service by allowing the software design process to be automatically generated from natural language descriptions without requiring manual intervention from analysts or designers. The deterministic model autonomously parses, interprets, and produces design specifications, significantly reducing the time and labor required for software design.

Inventive Principle:
Principle #25Self-service

3Manufacturing precision

If multiple translation steps are used, then comprehensive analysis is achieved, but the gap between requester intentions and final products increases

Engineering Contradiction:
Improvedesign accuracyVSAvoidcomplexity of translation process
Core Design Contradiction:
Manufacturing precisionVSDevice complexity

Solution Approach 1:

The patent merges multiple translation steps into a single integrated deterministic model. Instead of separate phases for parsing, analysis, and design generation performed by different humans, the system combines these functions into one automated process that directly transforms natural language into software design specifications, reducing the cumulative complexity and error propagation of multiple translation steps.

Inventive Principle:
Principle #5Merging (Combining)

Data Source

PatentEP3364295B1Method and system for automatically generating functional architecture documents and documents for software design and analysis specification
Publication Date: 2020.08.05 HUEBRA NADIA ANALIA

AI summary

The present invention is a computer-implemented system and method for automatic generation of Functional Architecture, business, analysis and software design documents based on phrases expressed in natural language. Eligible languages and their syntactic and grammatical features can be entered into the system. The system receives the description of the case in natural language, in one of the eligible languages, through an input/output device. The processor automatically analyses the description, automatically extracts the functional components from its morphosyntactic structure and automatically generates functional architecture documents for use in industry. Based on the functional components of the text, the processor automatically generates design components in order to produce business, analysis, and software design specification documents.