C-TRAIL: A Program Comprehension Approach for Leveraging Learning Models in Automated Code Trail Generation

摘要

With society's increasing utilization of (embedded) software, the amount of program source code is proliferating while the skilled human resources to maintain and evolve this code remain limited. Therefore, software tools are needed that can support and enhance program code comprehension. This paper focuses on program concept location and cognitive learning models, and contributes an automatic code trail generator approach called a Code Trail Recommender Agent Incorporating Learning models (C-TRAIL). Initial empirical results applying the prototype on obfuscated code show promise for improve program comprehension efficiency and effectiveness.