Software execution processes as an evolving complex network☆

Cai, Kai-Yuan; Yin, Beibei

doi:10.1016/j.ins.2009.01.011

Cited by 58 publications

(45 citation statements)

References 53 publications

(107 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We can find that the curves in the loglog plot roughly fit a piecewise power law. This coincides with the observation presented in reference [10] that the out-degree distribution may be better treated as a piecewise power law than a single power law. Note here that a piecewise power law distribution is also a reflection of the scale-free property.…”

supporting

confidence: 92%

“…Scale-free distribution obeys a right-screwed straight-line form on the doubly logarithmmic scale. Unlike the existing studies, our previous work [10] treated the dynamic execution process of software as an evolving network which is called software mirror graph in our work, and found that software mirror graph might also be small-world and scale-free.…”

Section: Introductionmentioning

confidence: 87%

“…For example, from the directed graph it is not clear how often a method is executed and how often a pair of methods is executed consecutively, although information of this kind is important for identifying method importance and software reliability. In our previous work [10], we proposed software mirror graph, which introduced a set of attributes to directed graph to record the dynamic information. More specifically, various attributes can be defined to convey dynamic information of interest and then be associated with nodes and edges in the directed topological graph.…”

Section: Software Mirror Graphmentioning

confidence: 99%

“…Here we note that besides the average path length, average temporal distance was also defined in our previous work [10]. It is defined to measure how many steps are required from an arbitrary method to another arbitrary method on average, which is calculated by the second attribute  in the attribute set A.…”

Section: Average Path Lengthmentioning

confidence: 99%

“…In order to run a subject program, a test suit comprising various test cases was required to represent the input domain of the subject program. A test case is a sequence of primitive actions applied to the GUI program under operation [10]. Related statistics of the software operation platform are tabulated in Table 1.…”

Section: Properties Of the Software Mirror Graphmentioning

confidence: 99%

See 4 more Smart Citations

Generating Mechanisms for Evolving Software Mirror Graph

Zhu¹,

Yin²,

Cai³

2012

JMP

Self Cite

View full text Add to dashboard Cite

Following the growing research interests in complex networks, in recent years many researchers treated static structures of software as complex networks and revealed that most of these networks demonstrate small-world effect and follow scale-free degree distribution. Different from the perspectives adopted in these works, our previous work proposed software mirror graph to model the dynamic execution processes of software and revealed software mirror graph may also be small world and scale-free. To explain how the software mirror graph evolves into a small world and scale free structure, in this paper we further proposed a mathematical model based on the mechanisms of growth, preferential attachment, and walking. This model captures some of the features of the software mirror graph, and the simulation results show that it can generate a network having similar properties to the software mirror graph. The implications are also discussed in this paper.

show abstract

supporting

confidence: 92%

Section: Introductionmentioning

confidence: 87%

Section: Software Mirror Graphmentioning

confidence: 99%

Section: Average Path Lengthmentioning

confidence: 99%

Section: Properties Of the Software Mirror Graphmentioning

confidence: 99%

See 3 more Smart Citations