A Classification and Clustering Analysis of Nodes on Knowledge Diffusion Paths——A Case from Knowledge-Sharing Platform"ZHIHU"

Ying WEI,Guan LIU,Feng LI

DOI: https://doi.org/10.13833/j.issn.1007-7634.2018.05.013

2018-01-01

Abstract:[Purpose/significance]In order to better understand and model the knowledge diffusion process, this paper addresses characteristics and pattern of nodes on the diffusion paths and network. [Method/process]Through tracing back a typical knowledge diffusion paths from one of the most popular Chinese knowledge-sharing social media platform-"ZHIHU (知乎)"network, a knowledge diffusion network and related social network are represented. Briefly, two kinds of analysis are done: one is to mine the characteristics of nodes based on their label on both the knowledge diffusion network and the social network, the other is to cluster the out-linked nodes of the same node, the son nodes of a same father node in the social network named. [Result/conclusion]According to our cross-correlation analysis, the results of K-Means clustering of nodes on the knowledge diffusion paths are significant. More importantly, the son nodes of a same father node show great similarity and cluster well. These results imply a novel research direction for modeling the real knowledge diffusion process.