[目的/意义] 科学知识网络中知识单元呈现出一定的集群性与社区性，揭示科学知识网络扩散时序变化过程中的社区扩张与收敛的基本模式与特征，对于拓展、深化科学知识扩散与传递规律研究具有一定的意义。[方法/过程] 首先，基于引用关系建立邻接矩阵进而构建学科知识网络，采用复杂网络分析中的Louvain社区探测算法对领域知识网络进行社区划分；然后利用网络表示学习技术进行社区扩张与收敛特征表示与计算；最后以时间序列为逻辑线索，对不同社区的扩张、收敛演变过程进行动态跟踪建模，从而揭示科学知识网络时序变化过程中社区扩张与收敛的基本模式与特征。[结果/结论] 以医疗健康信息领域进行案例研究，研究发现社区扩张模式的发展趋势符合S形曲线函数中的Logistic模型，社区收敛模式的发展趋势符合S形曲线函数中的BiHill模型。
[Purpose/significance] Knowledge units in scientific knowledge networks show certain clustering and communality, revealing the basic patterns and rules of community expansion and convergence in the process of changing the time series of scientific knowledge networks, which has certain significance for expanding and deepening the research on the diffusion and transmission of scientific knowledge. [Method/process] Firstly, the adjacency matrix was built based on the citation relation, and then the subject knowledge network was constructed. The Louvain community detection algorithm in complex network analysis is used to divide the domain knowledge network into communities. Then, the Graph Embedding technique was used to represent and calculate the community expansion and convergence characteristics. Finally, the time series was used as the time series. Logical clues were used to dynamically track and model the process of expansion and convergence of different communities, so as to reveal the basic patterns and laws of community expansion and convergence in the process of time series change of scientific knowledge network. [Result/conclusion] A case study in the field of health information shows that the trend of community expansion conforms to the Logistic model in the S-shaped curve function and the trend of community convergence conforms to the BiHill model in the S-shaped curve function.
[1] 邱均平,李小涛. 基于引文网络挖掘和时序分析的知识扩散研究[J]. 情报理论与实践, 2014(7):5-10.
[2] 李纲,巴志超. 科研合作超网络下的知识扩散演化模型研究[J]. 情报学报, 2017(3):58-68.
[3] 岳增慧,许海云. 学科引证网络知识扩散特征研究[J]. 情报学报, 2019, 38(1):5-16.
[4] NOYONS E C M, RAAN A F J V.Monitoring scientific developments from a dynamic perspective:self-organized structuring to map neural network research[J].Journal of the Association for Information Science & Technology, 1998, 49(1):68-81.
[5] 雷迭斯多夫. 科学计量学的挑战:科学交流的发展、测度和自组织[M].乌云,译. 北京:科学技术文献出版社, 2003.
[6] LEYDESDORFF L, COZZENS S, PETER V D B. Tracking areas of strategic importance using scientometric journal mappings[J]. Research policy, 1994, 23(2):217-229.
[7] LEYDESDORFF L. Statistics for the dynamic analysis of scientometric data:the evolution of the sciences in terms of trajectories and regimes[J]. Scientometrics, 2013, 96(3):731-741.
[8] POPPER K R.The logic of scientific discovery[J]. Yinshan academic journal, 2005,12(11):53-54.
[9] 靖继鹏,马费成,张向. 情报科学理论[M].北京:科学出版社,2009.
[10] 刘则渊. 跨越学术分水岭[M]. 北京:人民出版社,2012.
[11] 万昊. 科学知识规模增长模式研究-基于数学建模和仿真论证[D]. 北京:中国科学院大学,2017.
[12] 刘自强,许海云,罗瑞,等. 基于主题关联分析的科技互动模式识别方法研究[J]. 情报学报, 2019, 38(10):997-1011.
[13] 安宁,滕广青,白淑春,等. 领域知识聚类性的动态演化分析[J]. 图书情报工作, 2018,62(10):85-93.
[14] GIRVAN M, NEWMAN M E J. Community structure in socia1and biological networks[J]. Proceedings of the National Academy of Sciences of the United States of America,2002,99(12):7821-7826.
[15] BETTENCOURT L M A, KAISER D I, KAUR J. Scientific discovery and topological transitions in collaboration networks[J]. Journal of informetrics, 2009, 3(3):210-221.
[16] 白如江,冷伏海.k-clique社区知识创新演化方法研究[J].图书情报工作,2013,57(17):94-99.
[17] 王晓光,程齐凯. 基于NEViewer的学科主题演化可视化分析[J]. 情报学报, 2013, 32(9):900-911.
[18] 滕广青. 基于频度演化的领域知识关联关系涌现[J]. 中国图书馆学报,2018,44(3):79-95.
[19] 滕广青. 关联驱动的领域知识群落生长[J]. 中国图书馆学报, 2017,43(3):58-71.
[20] KUHN T S. The structure of scientific revolutions[M]. Chicago:University of Chicago Press, 1962.
[21] PRICE D J. Science since babylon[M]. New Haven:Yale University Press, 1961.
[22] PRICE D J. Little science, big science[M]. New York:Columbia University Press, 1963.
[23] DIANA C.Invisible colleges-diffusion of knowledge in scientific communities[M]. Chicago:The University of Chicago Press, 1972.
[24] 罗双玲,张文琪,夏昊翔.基于半积累引文网络社区发现的学科领域主题演化分析——以"合作演化"领域为例[J].情报学报,2017,36(1):100-110.
[25] BLONDEL V D,GUILLAUME J L, LAMBIOTTE R, et al. Fast unfolding of communities in large networks[J]. Journal of statistical mechanics:theory and experiment, 2008, 2008(10):10008.
[26] BLONDEL V D. Louvain algorithm[EB/OL].[2019-07-17].https://perso.uclouvain.be/vincent.blondel/research/louvain.html.
[27] NEWMAN M E J, GIRVAN M. Finding and evaluating community structure in networks[J].Physical review, 2004, 69(2):108-113.
[28] 周练.Word2vec的工作原理及应用探究[J].科技情报开发与经济,2015,25(2):145-148.
[29] MIKOLOV T, SUTSKEVER I, CHEN K, et al. Distributed representations of words and phrases and their compositionality[C]//Advances in neural information processing systems 26. Cambridge:Neural Information Processing Systems Foundation Inc., 2013:3111-3119.
[30] GROVER A, LESKOVEC J. Node2vec:scalable feature learning for networks[C]//Proceeding of the 22nd ACM SIGKDD international conference on knowledge discovery and data mining. New York:ACM, 2016:855-864.
[31] LAURENS V D M, HINTON G.Visualizing data using t-SNE[J]. Journal of machine learning research, 2008, 9(11):2579-2605.
