Audiometry
发表于 2025-3-23 10:53:21
Pablo Curiel,Ivan Pretel,Ana B. Lagonderstanding and provided some initial evidence that this approach scales. However, there is additional experimental work to be done before moving our approach to program understanding out of the lab and into commercial practice. Despite its shortcomings, however, our work is an example both of the
Constituent
发表于 2025-3-23 15:53:12
Constraint-Based Design Recovery for Software Reengineering978-1-4615-5461-5Series ISSN 1384-6469
neutralize
发表于 2025-3-23 20:05:54
http://reply.papertrans.cn/24/2360/235941/235941_13.png
单纯
发表于 2025-3-23 22:53:00
http://reply.papertrans.cn/24/2360/235941/235941_14.png
Hearten
发表于 2025-3-24 03:24:00
http://reply.papertrans.cn/24/2360/235941/235941_15.png
FOLD
发表于 2025-3-24 09:35:50
Introduction,ngineering but is difficult to obtain due to the complexity of software, the specialized knowledge used to construct it, and the lack of existing understanding aids. As a result, there is a need for tools that can help programmers obtain this understanding. It is particularly crucial to have tools t
synchronous
发表于 2025-3-24 12:28:22
http://reply.papertrans.cn/24/2360/235941/235941_17.png
indices
发表于 2025-3-24 17:03:07
Program Understanding and Constraint Satisfaction,straint satisfaction problems (CSPs) to program understanding. One way to do so involves breaking program understanding into plan matching (finding all instances of a single plan in code) and plan ordering (determining an ordering for trying the plans in a plan library). Plan matching can be directl
外科医生
发表于 2025-3-24 21:59:44
Initial Experiments with Concept Recovery, algorithms tend to have different behavior in different problem domains, we have run a series of experiments to try to find a solving strategy whose performance is reasonable on small programs (those containing less than 1,500 lines of source code). We have then taken the best strategy and explored
FRET
发表于 2025-3-24 23:24:17
Additional Experiments with Concept Recovery,ollection of artificially generated programs, and they focused primarily on variants of FCDR, a particular, general-purpose constraint satisfaction approach. We now report on experiments with programs based on real-world C code. While not definitive, these new experiments support our earlier results