社交网话题信息传播影响簇发现谱系挖掘方法及其应用研究

发布时间：2018-03-02 13:23

本文选题：话题信息传播　切入点：频繁传播序列　出处：《合肥工业大学》2015年硕士论文　论文类型：学位论文

【摘要】：在社交网中,话题信息能否引起网民的关注和转发,与网民对该话题信息的关注度、兴趣度和参与度等有密切关系。具有相同兴趣和偏好的网民,为了吸引更多的人关注他们讨论的话题内容,会用各种方法聚焦眼球、频繁互动和积极推广,以增强话题信息的传播影响作用。企业可借此展开营销信息推广活动,通过设置营销话题,借助高影响力的网民人群传播和放大营销信息的内容,以实现低成本的口碑营销。利用社交网传播信息一般都有特定的营销主题话题,如何找到对推广该营销信息有高影响力的贡献人群(影响簇)是企业网络营销策划决策的关键问题。本文研究在社交网中发现对话题信息传播有重要影响力人群的方法,主要通过分析社交网中有相同兴趣和爱好的个体及其之间频繁互动关系,采用DBSCAN聚类方法,构建频繁传播序列和影响簇谱系发现算法,以此形成为社交网话题信息传播影响簇发现谱系挖掘方法,并通过设计面向广告投放的影响簇群用户发现应用,为企业网络营销决策支持提供理论依据和实用方法。
[Abstract]:In social networks, whether or not topic information can attract the attention and transmission of Internet users is closely related to the attention, interest and participation of netizens to the topic information. In order to attract more people to pay attention to the topics they are discussing, they will use various methods to focus their eyeballs, interact frequently and actively promote the topic information, so as to enhance the communication and influence of topic information. Enterprises can use this to launch marketing information promotion activities. By setting up marketing topics, spreading and amplifying the content of marketing information with the help of high-impact netizens, we can achieve low-cost word-of-mouth marketing. The use of social networks to spread information generally has a specific topic of marketing theme. How to find the people (influence cluster) who have a high influence on the promotion of the marketing information is the key problem of the enterprise's network marketing planning and decision. This paper studies the method of finding a group of people who have important influence on the topic information dissemination in the social network. Based on the analysis of individuals with the same interests and hobbies in social networks and their frequent interactions, DBSCAN clustering method is used to construct the frequent propagation sequences and the algorithms for discovering the influence cluster pedigree. In this way, the method of discovering the cluster discovery pedigree for the topic information transmission of social network is formed, and through the design of the application of the influence cluster user discovery for advertising, it provides the theoretical basis and practical method for the enterprise network marketing decision support.
【学位授予单位】：合肥工业大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：G206

【相似文献】