【作者】 冯博；

【导师】 樊治平；

【作者基本信息】 东北大学 ， 管理科学与工程， 2009， 博士

【摘要】 在经济全球化和信息网络技术迅猛发展的背景下,组织(企业)内部以及组织之间的协同网络快速发展,并成为一种必然趋势。依托于网络组织开展的合作是优化配置经济资源、科技资源、生产力要素,以及推动知识创新的最有效的方式之一。协同网络运作和管理中存在着大量的决策问题,例如协同网络的构建、协同网络的运营绩效评价等。关于协同网络环境下的决策问题研究被国家自然科学基金委管理科学部列为管理科学与工程学科“十一五”期间(2006-2010年)的重要研究课题。在协同网络中,由于网络主体之间的密切合作和交互,使得网络主体之间形成了丰富的协同信息。在很多面向协同网络的现实决策问题中,不仅需要使用以往传统决策中使用的关于网络主体个体表现的信息,还需要同时考虑协同主体之间协作表现的协同信息。这种需要同时考虑网络主体的个体信息和协同网络主体间协同信息的决策问题,称为基于协同网络信息的决策问题。该问题具有大量的实际背景,例如新产品研发团队的伙伴选择问题,交叉功能团队的伙伴选择问题等等。目前,协同网络信息在决策问题中的重要性已经引起一些学者的关注,但是,关于如何使用协同网络信息进行决策,并针对决策问题给出科学有效的决策方法的研究很少被发现。因此,对基于协同网络信息的决策问题进行提炼,并给出具有针对性的决策方法,是一个具有重要意义、迫切需要解决的研究课题。本文对基于协同网络信息的决策问题进行了探讨,针对现有研究的薄弱之处,主要开展了以下创新性工作：(1)提炼了基于协同网络信息的研究问题,并提出了问题研究的理论框架。由于基于协同网络信息的决策问题具有大量的实际背景,并且对该类问题的研究具有很好的理论意义和实际应用价值,所以针对基于协同网络信息的决策问题进行了提炼,划分为四大类问题：网络主体排序与优选问题、大网络选择子网络的优化决策问题、子群选择的多指标决策问题和网络协同表现评价问题,并进一步分解为若干典型决策问题。在此基础上,为解决基于协同网络信息的决策问题,给出了问题研究的理论框架。主要包括,①针对协同网络、协同网络信息进行了描述和分类；②对基于协同网络信息的决策问题进行了网络结构描述和形式化描述,并给出了一般性的问题结构模型；③给出了进行基于协同网络信息决策问题分析的矩阵模型。这使得基于协同网络信息决策问题的研究形成一个理论体系。同时,为相关问题的研究提供了一般性的理论指导框架和分析框架,并为研究问题的扩展与应用提供了理论基础。(2)针对网络主体排序与优选问题,提出了基于协同网络信息的团队成员选择的多指标决策方法。具体地,研究了两类典型决策问题,分别提出了：①基于协同网络信息的合作研发联盟成员选择方法；②基于协同网络信息的项目团队成员选择方法。(3)针对大网络选择子网络的优化决策问题,提出了基于协同网络信息的团队成员选择的多目标决策方法。具体地,按照由简单到复杂的情形研究了三类典型决策问题,分别提出了：①考虑协同信息的团队成员选择方法；②基于协同网络信息的R&D团队成员选择方法；③基于个体与协同信息的交叉功能团队成员选择方法。(4)针对子群选择的多指标决策问题,提出了基于协同网络信息的优秀团队选择的决策方法。具体地,选择最具代表性的优秀创新研究群体项目选择问题作为典型研究问题,提出了一种基于协同网络信息的创新研究群体项目选择方法。(5)针对网络协同表现评价问题,提出了基于协同网络信息的团队表现评价方法。具体地,按照由简单到复杂的情形研究了两类典型问题,分别提出了：①新产品研发团队的协作满意度评价方法；②基于信誉与合作的虚拟团队相互信任评价方法。上述提出的理论与方法不仅能够解决合作研发联盟的伙伴选择、R&D团队成员选择、交叉功能团队成员选择、知识创新团队伙伴选择、优秀创新研究群体选择、虚拟团队相互信任评价、NPD团队协作满意评价等大量实际问题；并可扩展到考虑技术兼容性的技术组合方案选择、考虑部件兼容性的产品配置、基于服务要素的服务设计以及基于组合效用的产品组合方案选择等其他问题。提出的方法对于考虑关联信息的决策问题的研究提供了方法层面的借鉴和指导。更多还原

【Abstract】 In the background of economic globalization and rapid development of information network technology, the collaboration networks within and among organizations (enterprises) develop rapidly. It becomes an inevitable trend. The cooperation relying on network organizations could achieve the opitimal configuration of science and technology resource, economic rescource and procuctivity factors. It is also one of the most effective means to put forword knowledge innovation. Furthermore, there are a lot of decision problems in the operation and management of collaboration networks, such as the construction and performance assessment of collaborative networks. The decision problem under collaboration network environment has been ranked as one of the most significant research issues by management science department of National Nature Science Foundation of China (NSFC) in the Eleventh Five-Year Plan (2006-2010) for management science and enginnering.Whithin a collaboration network, abundant collbaborative information produces due to the close interaction and cooperation among agents of the network. In a number of decision problems oriented to collaboration networks, not only the individual information asscociating with the agent of network commonly used in the existing decision-making methods, but also the collaborative information between a pairwise agents of network should be considered. We call this type of decision problem as the one based on collaboration information. It exists extensively in the real wold. Now, a number of schorlars have focused on the decision problem based on individual and collaborative information, namely collaboration network information. However, few studides are conducted to make decision using collaboration network information by scientific and effective decision-making methods. Therefore, it is a significant, necessary and urgent research issue to summarize the problem based on collaboration network information and propose the decision-making methodology according to a particular problem.The purpose of this thesis is to address the decision problem based on collaboration network information. With regard to the limitations in the existing research, a series of research works are conducted and the contributions are as follows:First, a new problem is presented and the theoretical framework for the research problem is proposed. The decision problem based on collaboration network information has extensive practical backgrounds. The study of this problem has profound theritical and pratical significanc. Thus, in this thesis, the decision problem with collagoration network information is presented based on the emerging problems. Moreover, the decision problems are divided into four categories, i.e., the problem of ranking and selecting agents of networks, the optimization deicision problem of selecting a sub-netwok from a parent network, the multiple criteria deicision problem of clinique selection, and the problem of performance evaluation of collaboration networks, and then each category problems are subdivided into several typical decision problems. Based on that, to solve the problem based on collaboration network information, the theoretical framework for the research problem is proposed, i.e.,ⅰ) The collaboration network and collaboration information are described and classified.ⅱ) The decision problem based on collaboration network information is described by network graph and is represented in the formalization, and then a structure model is preseted to character the general problem.ⅲ) A matrix model is proposed to analyze the decision problem based on collaboration network information. All of these works make the studies of the addressed problem to be theoretically systemic. Moreover, that provides a general theoretical guidance framework and analytical framework to the related problems and also lays a theoretical foundation to the extended issues and related applications.Second, with regard to the problem of ranking and selecting agents of networks, the multiple attribute decision making method for team member selection based on collaboration network information is proposed. Concretely, two type classical deicision problems are studied, and then two methods are proposed, i.e.,ⅰ) a method for member selection of codevelopment alliances based on collaboration network information, andⅱ) a method for member selection of program team based on collaboration network information.Third, with regard to the optimization deicision problem of selecting a sub-netwok from a parent network, the muti-objective decision-making method for team member selection based on collaboration network information is proposed. Concretely, three type problems are studied according to the research thought from the simple problem to the more complicated one, three methods are proposed as follows.ⅰ) A method for team member selection considering collaboration information, ii) A method for team member selection of R&D teams based on collaboration network information, iii) A method for team member selection of cross-funcational teams based on the individual and collaborative information.Fourth, with regard to the multiple criteria deicision problem of clinique selection, the decision-making method for excellent team selection based on collaboration network information is proposed. Concretely, a most representative problem, namely, the problem of found program selection of excellent innovation research group is studied. Then, a method for foud program selection of excellent innovation research group is proposed.Fifth, with regard to the problem of performance evaluation of collaboration networks, the method for evaluating team performance based on collaboration network information is proposed. Concretely, two type problems are studied according to the research thought from the simple problem to the more complicated one. Then, two methods are pososed, i.e.,ⅰ) a method for evaluating collaboration satisfaction of NPD teams using mutual-evaluation information, andⅱ) a method for mutual trust evaluation of virtual teams based on reputation and collaboration.The proposed method can solve many practicle problems. For instance, partner selection for cooperative R&D alliances, member selection for R&D teams, member selection for cross-functional teams, member selection for knowledge innovation teams, scientific fund project selection of innovation research groups, assessment on mutual trust of virtual teams, evaluation on collaborative satisfaction of NPD teams, and so on. The proposed mthods can further extend to solve other problems such as selecting technology combination alternative considering technology compatibility, configuring products considering component compatibility, sevice design based on sevice factors as well as product packge selection based on combinational utility and so on. The proposed methods in this thesis provide guidance and reference in methodology to futher studies on the decision problem with relative information.更多还原