ABSTRACT

Social networks are online platforms that people use for interaction, information sharing and propagation of new ideas. Finding influential users in online social networks is a significant research problem due to its vast research applications in information diffusion, marketing and advertising. The relevant literature presents several models proposed for identifying influential users in social networks. In this survey, we present a review of the most relevant studies on influential users mining in microblog networks. First, we propose a new taxonomy by classifying the influence finding algorithms into five main categories based on their underlying framework and baseline methods. Second, each study is analysed according to the proposed framework, experimental datasets, validation approaches and evaluation results. Finally, the survey concludes with discussion on applications from the relevant literature, exploring open research challenges and presenting possible future research directions. The findings of this survey indicate that influential users mining in microblogs has many applications in marketing, advertising and information diffusion. In addition, this survey can be used as a guideline, particularly by young researchers, for establishing a baseline before initiating a research or identifying attractive as well as relevant research insights.

摘要

社交网络是人们用于互动、信息共享和传播新思想的在线平台。由于其在信息传播、营销和广告方面的广泛研究应用，在在线社交网络中寻找有影响力的用户是一个重要的研究问题。相关文献提出了几种用于识别社交网络中有影响力的用户的模型。在本次调查中，我们回顾了微博网络中影响力用户挖掘的最相关研究。首先，我们提出了一种新的分类法，根据其基础框架和基线方法将影响发现算法分为五个主要类别。其次，根据提出的框架、实验数据集、验证方法和评估结果对每项研究进行分析。最后，调查最后讨论了相关文献中的应用，探索了开放的研究挑战并提出了未来可能的研究方向。本次调查结果表明，微博中影响力用户挖掘在营销、广告和信息传播方面有很多应用。此外，这项调查可用作指南，尤其是年轻研究人员，用于在开始研究或确定有吸引力的相关研究见解之前建立基线。