system
The system uses generative AI to automatically detect and correct inconsistencies between source code and specification documents, ensuring the specification reflects the latest changes, improving specification management efficiency.
JP2026106984APending Publication Date: 2026-06-30SOFTBANK GROUP CORP
Patent Information
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- SOFTBANK GROUP CORP
- Filing Date
- 2024-12-18
- Publication Date
- 2026-06-30
AI Technical Summary
Technical Problem
Inconsistent change history of source code with specification documents is difficult to detect manually, leading to inefficient specification management.
Method used
A system comprising an acquisition unit, extraction unit, and suggestion unit that uses generative AI to automatically acquire, analyze, and suggest changes to specification documents based on detected inconsistencies between source code and specification documents.
Benefits of technology
Automatically detects and corrects inconsistencies, ensuring the specification document always reflects the latest source code changes, enhancing specification management efficiency and reducing communication discrepancies.
✦ Generated by Eureka AI based on patent content.
Smart Images
Figure 2026106984000001_ABST
Abstract
The system according to this embodiment aims to automatically detect inconsistencies between the source code change history and the specification document, thereby streamlining the management of the specification document. [Solution] The system according to the embodiment comprises an acquisition unit, an extraction unit, a detection unit, and a suggestion unit. The acquisition unit automatically acquires the change history of the source code. The extraction unit analyzes the change history acquired by the acquisition unit and extracts important requirements and specifications. The detection unit detects differences and inconsistencies between the requirements extracted by the extraction unit and the implementation of the source code. The suggestion unit suggests changes to the specifications based on the differences and inconsistencies detected by the detection unit.
Need to check novelty before this filing date? Find Prior Art