Related Experiment Videos
Signatures of small-world and scale-free properties in large computer programs
Alessandro P S de Moura1, Ying-Cheng Lai, Adilson E Motter
1Instituto de Física, Universidade de São Paulo, Caixa Postal 66318, 05315-970 São Paulo, Brazil.
Summary
Large computer programs form scale-free networks with small-world properties due to their growth and optimization. This reveals inherent structural characteristics of complex software systems.
Area of Science:
- Computer Science
- Network Science
- Software Engineering
Background:
- Large computer programs consist of numerous interconnected units.
- The logical connections form a network representing program structure and information flow.
Purpose of the Study:
- To analyze the network structure of large computer programs.
- To identify scale-free and small-world properties in software networks.
Main Methods:
- Analysis of network properties in large computer programs.
- Investigating the impact of program growth and optimization on network topology.
Main Results:
- Software networks exhibit scale-free characteristics, with node link probabilities following a power-law distribution.
- Performance optimization leads to small-world network structures in computer programs.
Conclusions:
- The network structure of large computer programs naturally displays scale-free and small-world features.
- These findings are likely generic across large software systems, extending network studies to software.