【作者】 徐美凤；

【导师】 叶继元；

【作者基本信息】 南京大学 ， 图书馆学， 2011， 博士

【摘要】 随着互联网的普及与深入发展,学术虚拟社区已经成为学术交流与知识共享的重要场所。尽管学术社区在知识共享范围与速度等方面有着诸多优势,但是仍然存在着用户参与不足、知识共享意愿低等问题。深入探索学术社区的知识共享特征与规律,有利于社区管理者提供针对性的激励与管理,有利于引导虚拟环境下的学术交流活动,推动虚拟环境下的学术知识共享,也有利于合理地使用虚拟社区服务于知识管理。目前关于虚拟社区知识共享的研究涉及多学科领域,包括社会学、心理学、行为科学、激励理论等。本文通过对大量文献的研究发现已有研究存在以下不足：(1)研究中较少考虑不同社区的学术领域特征；(2)忽视个人主观行为对学术社区交流环境的影响。本文以复杂适应系统理论为基础,以学术虚拟社区作为研究对象；界定了学术虚拟社区、社区知识共享的基本概念；对学术社区知识共享主体进行了分类；阐明了社区成员知识共享行为的动机与影响因素；揭示了知识共享主体的互动规律；对学术社区的专业特征进行了系统分析；并进一步通过对社区成员行为的仿真模拟,探寻学术社区知识共享的动态演化规律。本文研究内容主要包括以下几个方面：1.利用问卷调查和社区网页信息收集的方法,获取丁香园社区神经科学专业讨论版的基本数据,分析学术社区的知识共享主体特征,并结合知识共享基尼系数和社会网络分析研究主体间的互动特征。结果显示,即使目前运营比较成功的学术社区,成员间知识共享的参与度也普遍不高,但宽松自由的交流氛围仍吸引着大量成员加入社区。2.根据主体间的相互作用是系统演化动力的观点,从知识共享主体间相互影响的角度出发,整合已有研究成果,分析学术社区知识共享行为的影响因素。同时,区别于已有研究中仅将发贴、回帖作为社区知识共享行为的观点,将成员的单纯浏览行为与发贴、回帖一起视为共享行为,构建学术社区知识共享行为影响因素的结构方程模型。通过问卷调查,对模型进行了验证与分析。发现：成员之间的熟悉程度、成员的级别对单纯的浏览行为有积极影响；信息性动机、成员的利他心理、知识共享自我效能对发帖行为具有显著影响；知识共享自我效能、成员对社区管理的信任、社区激励以及互惠规范对回帖行为具有显著影响。同时成员对社区管理的信任、社区激励对成员的信息性动机、基于认同因素的共享动机有显著的积极影响。3.对小木虫、丁香园、中国学术论坛和国学论坛四个学术社区进行发回帖内容收集,分析发现,学术社区的知识共享主体具有身份相对稳定性、交流内容的专业性、交流态度的严谨理性等共性。同时,人文管理类社区与理工类社区在话题来源、语言风格以及对知识权威的认同等方面存在差异。引入基尼系数对社区主体间知识共享不均衡程度进行分析,发现人文管理类社区主体间不均衡程度明显高于理工类社区；使用社会网络分析方法对社区知识共享主体构成进行分析,发现两类社区都存在中心势过低的现象,同时人文管理类社区中更容易形成社区核心。基于各观测指标数据的描述性统计以及结构方程模型分析,可以发现两类社区成员的知识共享行为偏好、知识共享动机、影响因素均有所区别。4.通过上述关于学术社区知识共享的主体构成、层次结构、影响元素等的理论分析与实证研究,构建学术社区知识共享行为的概念模型,模拟学术社区主体间知识共享的动态过程,归纳出理工学科和人文管理学科社区具有问答式和发布式两种不同的交流组织形式。本研究创新之处主要体现在两个方面：(1)理论视角：提出学术虚拟社区知识共享是复杂适应系统,结合现有研究成果,构建了学术社区知识共享行为影响因素模型,较为深入地研究了学术社区中知识共享的影响因素；同时对学术社区知识共享的专业特征进行了系统分析。(2)研究路径：引入基于多主体的建模仿真,构建学术社区知识共享的概念模型,并进行仿真模拟,以探寻学术社区知识共享的动态演化过程,模拟结果不仅验证了实际调查获得的某些结论,而且归纳出了人文管理类社区与理工类社区知识共享组织形式的差异。更多还原

【Abstract】 With the popularity and in-depth development of Internet, academic virtual communities have become the important academic knowledge and exchange sharing place. Although academic communities have lots of advantages such as knowledge updating and exchanging rate, but there are still some preblems including users’lack of participation and low willingness of knowledge sharing. Exploring the knowledge sharing feature and law in academic communities is significant. It will be helpful for community managers provide pertinent incentives and effective management. And it will help guide the academic activities and promote academic knowledge sharing under virtual environment. Meanwhile, it is also conducive to the rational use of virtual communities for knowledge management.Present research on the knowledge sharing of virtual communities is refering to in multidisciplinary areas. Based on the extensive literature study, the shortages of present study are found:(1) the discipline characteristics of virtual communities were ignored; (2) personal behaviors’impact on the environment of virtual communities was neglected.Based on Complex Adaptive System (CAS) theory, this dissertation taking academic virtual communities as research objects, study the characteristics and laws of knowledge sharing. In this study, the conception, process and pattern of knowledge sharing are defined firstly through literature research.The the constitute of knowledge sharing subjects in academic communities is analyzed, the motivation and impact factors of knowledge-sharing behavior are set forth, the interaction laws of knowledge-sharing is revealed, the professional characteristics of academic communities are systematically analyzed;and with the simulation of members’behavior, the dynamic evolution of knowledge-sharing procession in academic virtual community is explored.The main study content of this dissertion includes the following aspects:1. In this paper, with survey and the web page information collection, the basic data of neural science forum of Dingxiangyuan community are gathered, the main characteristic of knowledge-sharing subjects is analyzed, and Gini coefficient of knowledge sharing and social network analysis method is employed to study the the interaction between the knowledge-sharing subjects of academic communties. The results shows that even in the more successful academic virtual communities, members’participation are generally not high, but the relaxed atmosphere of free knowledge exchange is still attracting a large number of members to the community.2. According to the view that interactions among the agents are the driving force of evolution, the impact factors of knowledge sharing behavior are analyzed based on the existing research. At the same time, different from the existing research in which posting and replying acts are taken as knowledge-sharing behavior, browsing is also regared as knowledge-sharing behavior in this dissertion and the structural equation model of academic community sharing knowledge is subsequently established. Through questionnaires, the model is verified and analyzed. The result shows that:the familiarity between members and member’s class has a positive impact on scanner; information motivation, the members altruism, knowledge-sharing self-efficacy significantly affected posting behavior; knowledge sharing self-efficacy, the members trust on the community management, motivation and reciprocity norms have a significant impact on repliers’behavior. At the same time the members confidence of the community management and the incentive of community significantly positively impact on the members information motivation and sharing motivation based on recognition factors.3. The emuch.net, dxy.cn, frchina.net and guoxue.com are selected as sample communities. The research found the members’common characteristics of the academic communities with the relative stability of identity, professional communication content, strict rationality of communication attitude. However, some differences between the communities of humanity-management and engineering-science are also found, such as the source of topics, the language style and the acceptance of knowledge authority. Gini coefficient is introduced to measure the level of imbalance in the knowledge sharing activities, and the value shows that the imbalance of humanity-management community is higher than that of-engineering-science community. The subject construction is analyzed by social network analysis and it is found that both the communities have the low network centralization and the community core is more easily formed in the humanity-managenrent community. Based on the descriptive statistical analysis and structural equation modeling, it is found that there are some differences between the two kind communities such as the preference knowledge sharing, knowledge sharing motivation, impact factors.4. A conceptual model of knowledge sharing behavior is built and the dynamic procession of knowledge sharing is simulated based on the above research. The result of simulation shows that the science engineering community and human management community have different types of knowledge sharing, that is, accordingly Q/A (question/answer) type and release type.The innovative points of this dissertation are mainly embodied in the following aspects:(1) Theoretical perspective:it is proposed that knowledge sharing in virtual community academic is the complex adaptive system. With the existing research results, the impact factors model of knowledge sharing behavior in academic communities is built. The In-depth research of influencing factors of knowledge sharing is done; meanwhile, the professional characteristics of knowledge sharing in academic communities are systematically analyzed. (2)Study path: multi-agent based modeling is used to explore the dynamic evolution process of knowledge sharing in academic communities. The simulation results not only verify some conclusion of the above investigation, but also sums up the differences of knowledge sharing organization form between humanity-management communities and engineering-science ones.更多还原