There exist correlations between editing behaviors and hyperlinks structure in Wikipedia
|There exist correlations between editing behaviors and hyperlinks structure in Wikipedia|
|Author(s)||Zhang H., Huang L., Li D., Zheng B.|
|Published in||Proceedings - 2011 IEEE International Conference on Granular Computing, GrC 2011|
|Keyword(s)||human behavior dynamics, social network analysis, Wikipedia (Extra: Collective behavior, Human behaviors, Hyperlink structure, Hyperlinks, Initial stages, Random network, Sample dataset, Social Network Analysis, Structure similarity, Two layers, Two-layer network, Wikipedia, Granular computing, Social networking (online), Hypertext systems)|
|Article||BASE, CiteSeerX, Google Scholar|
|Web||Ask, Bing, Google (PDF), Yahoo!|
|Download and mirrors|
|Local copy||Not available|
|Remote mirror(s)||Not available|
|Export and share|
|BibTeX, CSV, RDF, JSON|
|Browse properties · List of conference papers|
There exist correlations between editing behaviors and hyperlinks structure in Wikipedia is a 2011 conference paper written in English by Zhang H., Huang L., Li D., Zheng B. and published in Proceedings - 2011 IEEE International Conference on Granular Computing, GrC 2011.
The co-editing in Wikipedia is a typical and complex collective behavior with lots of voluntary editors' participation, while the relationship between the edit behaviors and the hyperlink structure of articles remains unknown until now. In this paper, we try to explore the correlation between them via a novel two-layer network. In this two-layer network, we model the articles in Wikipedia as nodes, and model the edits and hyperlinks as the edges of two layers respectively. Here, the correlation is suggested to be measured by a structure similarity metric. By analyzing the structure similarity of two layers via a method named partially ordered ranking, we find that there exist significant and stable correlations: in our sample dataset composed of four Wikipedia categories, the structure similarity is around 0.6, which is two times than that of a theoretical random network. Furthermore, if turn back to the initial stage of categories, i.e., take the evolution into consideration, the correlation is evolving too. Usually the evolution undergoes a sharp decline stage from the initial high value, and at last it tends to the stable value around 0.6.
- This section requires expansion. Please, help!
Probably, this publication is cited by others, but there are no articles available for them in WikiPapers.