원문정보
피인용수 : 0건 (자료제공 : 네이버학술정보)
초록
영어
Label Propagation Algorithm is a kind of community discovery method. This algorithm contains large numbers of random selections, which made the result uncertain and reduced the stability of the algorithm. In order to solve these problems, this paper proposed Label Cohesion Algorithm (LCA). In LCA algorithm the label propagation process is divided into two steps. The first step is taking pretreatments on the original labels. The second step is label updating. In the first step we change node label though node centripetal. In the second step the paper use label Cohesion as the judgement to choose the new label. Finally the experimental result shows that the accuracy of the algorithm has been improved.
목차
Abstract
1. Introduction
2. Basic Concept
3. Label Cohesion Algorithm
3.1. The Thought of LCAAlgorithm
3.2. The Description of LCA Algorithm
4. Experiments and Analysis
4.1. Karate Club Network Analysis
4.2. American College Football Network Analysis
5. Conclusions
Acknowledgements
References
1. Introduction
2. Basic Concept
3. Label Cohesion Algorithm
3.1. The Thought of LCAAlgorithm
3.2. The Description of LCA Algorithm
4. Experiments and Analysis
4.1. Karate Club Network Analysis
4.2. American College Football Network Analysis
5. Conclusions
Acknowledgements
References
저자정보
참고문헌
자료제공 : 네이버학술정보