Integrating Data Mining Techniques and Social Network Model into Effective Framework and Its Application in Computational Biology

社会网络研究是跨学科的，需要社会学、行为科学、心理学、统计学、数学、计算机科学等方面的专业知识。信息技术的最新发展对社会网络的研究产生了很大的影响和影响，然而，如果可以将这一过程扩展到表征应用程序的大量数据，则可以构建更全面和信息丰富的模型，这一问题变得具有挑战性。我们认为，除了采用数据挖掘技术考虑显式链接之外，还可以通过考虑隐式链接来丰富模型。生成的模型将用于更有效的知识发现。定义问题和建立模型是社会网络方法论在不同领域成功应用所面临的主要问题。数据挖掘技术首次用于实现社交网络中参与者之间的交互。一旦建立了社会网络模型，除了现有的社会网络分析指标外，数据挖掘技术还可以用于进一步分析网络。本研究将利用各种数据挖掘技术来推导社交网络中链接的强度；这将通过在全球范围内考虑行为者，导致它们之间更强的相关性。我们的主要目标是：1)将数据挖掘和社会网络模型集成到一个全面和健壮的框架中。2)将所建立的模型应用于分子相互作用网络，实现其有效性。3)采用数据挖掘技术促进新分子成分的识别。4)整合海量异构数据，构建功能性社交网络。这项研究扩展了现有软件包的功能，并朝着具有社会和科学影响的综合产品迈进。社会和科学界都需要这种工具。

Social network research is multidisciplinary requiring expertise from sociology, behavioral science, psychology, statistics, mathematics, computer science, etc. Recent development in information technology has highly influenced and shaped the research in social networks, however, the problem became challenging in that a more comprehensive and informative model could be constructed if the process could be extended for the huge amount of data characterizing the application. We argue that the model could be enriched by considering implicit links in addition to the explicit ones by employing data mining techniques. The resultant model will be used for more effective knowledge discovery. Defining the problem and building the model are the main issues facing the successful application of social network methodology in different domains. Data mining techniques are firstly used to realize the interactions between the actors in a social network. Once a social network model is constructed, data mining techniques can be used to further analyze the network in addition to existing metrics in social network analysis. This research will utilize various data mining techniques to derive the strength of the links in a social network; this will lead to stronger correlations between actors by considering them globally. Our major goals are: 1) Integrating data mining and social network model into a comprehensive and robust framework. 2) Realizing the effectiveness of the developed model by applying it to molecular interaction networks. 3) Adapting data mining techniques to facilitate the recognition of new molecular components. 4) Integrating large sets of heterogeneous data to build functional social networks. This research expands the power of existing software packages and advances towards a comprehensive product of social and scientific impact. There is a need for such tool in the social and scientific communities.