基于派系演化的Folksonomy模式下的知识网络变化研究

发布时间：2018-04-09 20:07

本文选题：Folksonomy　切入点：知识网络　出处：《东北师范大学》2017年硕士论文

【摘要】：作为Web2.0典型应用的Folksonomy知识组织模式开辟了网络社群知识组织新的路径。然而,无论是朴素结构主义的研究思想,还是基于主观动态因素的研究思路,都距离Folksonomy模式中知识组织变化趋势的研究尚存一定的间隙。随着复杂网络分析理论与技术的引入,研究思路突破了传统“树形”结构的模型,社群知识间纷繁复杂的关联关系得以揭示。然而,现有基于网络分析的研究多从静态截面上对Folksonomy模式中社群知识组织变化趋势展开研究,或者有针对动态演化规律的研究但是不够深入,对于知识组织过程中产生的知识涌现更是鲜有问津。特别是针对派系演化作为研究并且基于动态网络的视角寥寥无几。有鉴于此,本文从社群知识网络结构动态演化的视角,基于派系在时间序列上的演化来研究Folksonomy模式中社群知识组织变化趋势问题,以及整个网络的变化趋势性问题。本文研究的第一部分通过构建“文章-标签”的隶属矩阵,然后在此基础上构建了“标签-标签”邻接矩阵。并且对邻接矩阵基于复杂网络分析技术进行可视化。第二部分通过网络引入派系作为研究指标,派系在不同时间刻度上的值我们就可以推断出骨架上点的数据量。第三部引入Logistic方程并且对Logistic方程进行改进,若我们能预测网络骨架上的点变化趋势是可预测的,那么整个知识网络变化的规律也是可预测的。最后实验结果也表明基于派系演化的Folksonomy模式下知识网络变化是稳定的。
[Abstract]:As a typical application of Web2.0, Folksonomy knowledge organization model has opened up a new path of network community knowledge organization.However, there is still a gap between the research ideas of naive structuralism and the research ideas based on subjective dynamic factors, which are still far from the research on the changing trend of knowledge organization in Folksonomy model.With the introduction of the theory and technology of complex network analysis, the research ideas break through the traditional "tree" structure model, and the complex and complex relationship between community knowledge can be revealed.However, the existing researches based on network analysis mainly focus on the trends of community knowledge organization in Folksonomy model from static cross-section, or on the dynamic evolution law, but it is not deep enough.There is little interest in the emergence of knowledge in the process of knowledge organization.In particular, there are few perspectives on factional evolution based on dynamic networks.In view of this, from the perspective of dynamic evolution of community knowledge network structure, based on the evolution of clique in time series, this paper studies the changing trend of community knowledge organization in Folksonomy model and the trend of the whole network.In the first part of this paper, we construct the membership matrix of "article-label", and then construct the "label-label" adjacency matrix.And the adjacent matrix is visualized based on complex network analysis technology.In the second part, we can infer the data amount of the point on the skeleton by introducing the clique as the research index through the network, and the value of the clique in different time scale.In the third part, the Logistic equation is introduced and the Logistic equation is improved. If we can predict the trend of point change on the network skeleton, then the law of the change of the whole knowledge network is also predictable.The experimental results also show that the knowledge network changes are stable in the Folksonomy mode based on factional evolution.
【学位授予单位】：东北师范大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：O157.5

【相似文献】