Flowgen helps create detailed software use case flows automatically
Relationally Guided Use Case Modeling with LLMs
Software EngineeringComputation and Language
Summary
Creating detailed use case flows for software is important but takes a lot of time and expert knowledge. The authors designed a method called FlowGen that uses advanced language models to understand and build these flows more accurately. FlowGen works by analyzing the meaning of requirements, predicting decision points, and generating alternative pathways in the software’s behavior. Tests show it performs better than other tools at producing precise and complete use case flows. This can help software teams save effort and improve early software design and testing.
What this means in practice
- •For software requirements engineers: Generate detailed and semantically consistent use case flows from requirement texts to aid analysis and design.
- •For test automation teams: Automatically produce alternative test case scenarios from generated use case flows to improve coverage.
Authors
Guangyu Wang, Bangqi Li, Ji Wu, Zhijun Shao
Abstract
Use case flows are important elements of use case modeling because they support downstream software engineering activities, including requirements analysis, architectural and detailed design, and test case generation. However, constructing them manually is costly and expertise-intensive, while existing automated approaches still struggle to preserve semantic consistency, control-flow logic, data-flow logic, and the intended system boundary, especially when identifying branch points and generating alternative flows. To address this problem, we propose FlowGen for complete use case flow construction. FlowGen uses LLM-based Semantic Information Processing (SIP) to extract semantic elements, constructs a Semantic Relational Graph (SRG) encoded by an enhanced R-GAT for basic flow generation (BFGen), and further supports branch point prediction through BPP and branch-conditioned alternative flow generation through AFGen. Evaluations on 13 public and 7 industrial datasets show that FlowGen consistently outperforms competitive baselines in all three core components. In particular, BFGen improves over the best baseline by 14% in Precision, 7-25% in Recall, 11-30% in F1, and 10-19% in AUC; BPP improves Precision by 30-110%, Recall by 33-91%, and F1 by 32-117%; AFGen improves Precision by 8-23%, F1 by 5-18%, and AUC by 0.6-2.5%. Moreover, we validate the effectiveness of the LLM-based SIP module and the attention preservation factor in BFGen, analyze the impact of requirement completeness on BFGen, and examine how different scopes of branch-related context affect AFGen.