中文摘要：

随着Web2．0技术的迅速发展，社会网络开始在我们的生活中扮演着重要的角色，越来越多的人在网络中发表言论、互相交流、共享信息．然而，在社会网络中，信任关系是用户间进行交互的基础，不同用户之间的信任关系强度不同，相同用户在不同领域内的信任关系强度也存在差异，信任关系的不确定性是信任评估的最大挑战．针对以上问题提出了一种基于改进D-S证据理论的灵活直观的评估方法，该方法综合考虑用户被关注度、用户信誉度、用户活跃度和用户相似度4个方面，将这4个方面作为4个属性证据，同时根据模糊理论中的隶属度原理获取基本信任分配，然后基于以上4个属性证据构建多源属性证据信任关系强度融合模型，在领域内对其信任关系强度进行评估，最后采用Epinions中真实的数据集进行实验．实验结果验证了该方法的可行性和优势，为复杂的社会网络环境中信任关系强度评估的研究提供了有价值的新思路．

英文摘要：

With the rapid development of Web 2.0 technologies,social networks play an importantrole in our daily life,for example,more and more people deliver speech,communicate with eachother,and share information in social networks.However,trust relationship is the basis of theinteraction between users in social networks,and the trust relationship strength is also differentbetween different users.In addition,the same users in different fields do not all have the samerelationship strength.What’s more,the uncertainty of trust relationship is still the biggestchallenge of trust evaluation.Having given the issues which have been discussed above,this paperpresents a flexible and intuitive method for the evaluation of trust relationship strength on thebasis of improved D-S evidence theory.Overall,this method considers the following fouraspects：the attention of the user,the credibility of the user,the vitality of the user,and thesimilarity of the users,which are regarded,subsequently,as four attribute evidences to access to distribution of basic trust based on membership degree principle of fuzzy theory at the same time.In order to evaluate the strength of trust relationship in a specific field,we build multi-sourceattribute evidences fusion model based on the four attribute evidences.Finally the model is built onthe Epinions real-life dataset,experimental results reveal that using D-S evidence theory instrength prediction of trust relationship is feasible and advantageous,which provides valuablenew ideas for trust relationship strength research in complex social networks.