基于主观逻辑方法的消费者多源信任融合模型
发布时间：2018-11-09 13:32
【摘要】：社会化商务中,消费者依赖在线口口相传建立感知信任,其本质是复杂网络上的多源信任融合问题。国内外学者对多源信任融合问题进行了大量研究,并以主观逻辑方法为代表形成了信任融合方法的研究体系。然而,由于社交网络中消费者感知信任的多源性和高度主观性以及用户生成内容的海量化,给多源信任融合模型带来量化难、实时处理难和融合难等问题。针对上述难题,提出"先聚类、后融合"的研究思路,先对海量推荐信息进行聚类,再融入感知信任的主观因素构建多源信任融合模型。首先,将推荐信息间的相似性作为节点关系,从社交网络中抽取出推荐信息相似性网络,用谱平分方法对聚簇进行划分,实现对推荐信息的聚类;其次,用网络属性度量感知信任的影响因素,从复杂网络视角出发,提出消费者感知信任定性因素的量化方法;再次,以多属性决策方法为基础,改进主观逻辑方法构建多源信任融合模型,从而将感知信任的影响因素融入主观逻辑方法,突破主观逻辑方法只考虑推荐信息和网络路径的局限性;最后,通过仿真实验对推荐信息实验数据进行聚类,并分析主观因素对感知信任意见空间的调节作用,验证模型的可行性。研究结果表明,研究模型能够快速划分推荐信息相似性网络,客观地量化感知信任的影响因素,使其融入信任度计算之中,且能够体现消费者感知信任的主观性和异质性。从仿真实验结果看,该模型能够有效解决大规模社会网络中推荐信息海量化问题,权威程度、从众行为和主体间亲密度等影响因素对信任度计算结果起调节作用。该模型将信任融合模型扩展到社会化商务领域,可以帮助商家评价已有消费群体对新消费者感知信任的影响力,为大规模网络中消费者感知信任的度量和预测提供新视角,为商家实时分析消费者感知信任意向和制定营销策略提供方法支持。对于难以建立新消费者信任的商家,可以通过制定激励机制来提高用户生成内容量,培养或引入权威人士作为明星节点,建立有主题的小社区等形式来促进主体间交流,增强消费者之间的亲密度,从而提高消费者感知信任。
[Abstract]:In social commerce, consumers rely on online word of mouth to establish perceived trust, which is essentially a multi-source trust fusion problem in complex networks. Scholars at home and abroad have done a lot of research on the problem of multi-source trust fusion and formed a research system of trust fusion method represented by the subjective logic method. However, because of the multi-source and highly subjective consumer perception trust and the user-generated content quantization in social network, it brings many problems such as quantization, real-time processing and fusion. Aiming at the above problems, this paper puts forward the research idea of "first clustering, then fusion". Firstly, we cluster the massive recommended information, and then integrate the subjective factors of perceived trust to construct a multi-source trust fusion model. Firstly, the similarity of recommendation information is taken as the node relation, the similarity network of recommendation information is extracted from social network, and the cluster is divided by the method of spectral partition to realize the clustering of recommendation information. Secondly, using network attributes to measure the influencing factors of perceived trust, from the perspective of complex network, a quantitative method of qualitative factors of consumer perceived trust is proposed. Thirdly, based on the method of multi-attribute decision, we improve the subjective logic method to construct the fusion model of multi-source trust, so that the influencing factors of perceived trust can be integrated into the subjective logic method. Break through the limitation of subjective logic method only considering recommendation information and network path; Finally, the data of recommendation information experiment are clustered through simulation experiments, and the effect of subjective factors on perceived trust opinion space is analyzed to verify the feasibility of the model. The results show that the research model can quickly divide the similarity network of recommendation information, quantify objectively the influencing factors of perceived trust, and integrate it into the calculation of trust degree, and can reflect the subjectivity and heterogeneity of consumer perceived trust. The simulation results show that the model can effectively solve the problem of recommendation information sea quantization in large-scale social networks. The influence factors such as authority degree, herd behavior and inter-agent affinity play an important role in the calculation of trust degree. This model extends the trust fusion model to the field of social commerce, which can help businesses evaluate the influence of existing consumer groups on new consumers' perceived trust, and provide a new perspective for the measurement and prediction of consumer perceived trust in large-scale networks. Provides the method support for the merchant to analyze the consumer perception trust intention and to formulate the marketing strategy in real time. For businesses that are difficult to establish new consumer trust, they can promote communication among subjects by establishing incentive mechanisms to increase the capacity of users to generate, train or introduce authoritative individuals as star nodes, and establish small communities with themes. Enhance the affinity between consumers, thereby enhancing consumer perceived trust.
【作者单位】： 大连理工大学系统工程研究所;
【基金】：国家自然科学基金(71431002) 国家创新研究群体科学基金(71421001)~~
【分类号】：F713.55
[Abstract]:In social commerce, consumers rely on online word of mouth to establish perceived trust, which is essentially a multi-source trust fusion problem in complex networks. Scholars at home and abroad have done a lot of research on the problem of multi-source trust fusion and formed a research system of trust fusion method represented by the subjective logic method. However, because of the multi-source and highly subjective consumer perception trust and the user-generated content quantization in social network, it brings many problems such as quantization, real-time processing and fusion. Aiming at the above problems, this paper puts forward the research idea of "first clustering, then fusion". Firstly, we cluster the massive recommended information, and then integrate the subjective factors of perceived trust to construct a multi-source trust fusion model. Firstly, the similarity of recommendation information is taken as the node relation, the similarity network of recommendation information is extracted from social network, and the cluster is divided by the method of spectral partition to realize the clustering of recommendation information. Secondly, using network attributes to measure the influencing factors of perceived trust, from the perspective of complex network, a quantitative method of qualitative factors of consumer perceived trust is proposed. Thirdly, based on the method of multi-attribute decision, we improve the subjective logic method to construct the fusion model of multi-source trust, so that the influencing factors of perceived trust can be integrated into the subjective logic method. Break through the limitation of subjective logic method only considering recommendation information and network path; Finally, the data of recommendation information experiment are clustered through simulation experiments, and the effect of subjective factors on perceived trust opinion space is analyzed to verify the feasibility of the model. The results show that the research model can quickly divide the similarity network of recommendation information, quantify objectively the influencing factors of perceived trust, and integrate it into the calculation of trust degree, and can reflect the subjectivity and heterogeneity of consumer perceived trust. The simulation results show that the model can effectively solve the problem of recommendation information sea quantization in large-scale social networks. The influence factors such as authority degree, herd behavior and inter-agent affinity play an important role in the calculation of trust degree. This model extends the trust fusion model to the field of social commerce, which can help businesses evaluate the influence of existing consumer groups on new consumers' perceived trust, and provide a new perspective for the measurement and prediction of consumer perceived trust in large-scale networks. Provides the method support for the merchant to analyze the consumer perception trust intention and to formulate the marketing strategy in real time. For businesses that are difficult to establish new consumer trust, they can promote communication among subjects by establishing incentive mechanisms to increase the capacity of users to generate, train or introduce authoritative individuals as star nodes, and establish small communities with themes. Enhance the affinity between consumers, thereby enhancing consumer perceived trust.
【作者单位】： 大连理工大学系统工程研究所;
【基金】：国家自然科学基金(71431002) 国家创新研究群体科学基金(71421001)~~
【分类号】：F713.55
【参考文献】
相关期刊论文 前9条
1 卢云帆;鲁耀斌;林家宝;亓小林;;社会化商务中顾客在线沟通研究:影响因素和作用规律[J];管理评论;2014年04期
2 徐彪;张媛媛;张s，
本文编号：2320553
本文链接：https://www.wllwen.com/guanlilunwen/yingxiaoguanlilunwen/2320553.html
