论文标题
在社交网络中寻找早期的创新采用者
Finding Early Adopters of Innovation in Social Network
论文作者
论文摘要
社交网络通过同龄人对收养的影响而在创新的扩散中起着基本作用。因此，已使用包括广泛网络中心度措施在内的网络地位来描述个人的亲和力，以采用创新及其传播扩散的能力。然而，就易感性和影响力以及网络中心方面而言，社交网络具有分类性。这使得对影响者的识别很困难，尤其是因为敏感性和中心性并不总是齐头并进。在这里，我们提出了专家建议方法的顶级候选算法，以根据其感知的专业知识对个人进行排名，这与创新者和早期采用者的分类性质产生了很好的共鸣。利用两个在线社交网络的采用数据，这些数据在采用方面具有分类性，但代表了网络中心的不同水平，我们证明，与其他广泛使用的指数相比，顶级候选人排名在捕获早期采用者方面更有效。与其他方法强调的节点相比，顶级候选人节点提前采用，并在创新者，早期采用者和早期多数中具有更高的影响力。这些结果表明，顶级候选方法可以识别出好种子，以影响社交网络上的最大化运动。
Social networks play a fundamental role in the diffusion of innovation through peers' influence on adoption. Thus, network position including a wide range of network centrality measures have been used to describe individuals' affinity to adopt an innovation and their ability to propagate diffusion. Yet, social networks are assortative in terms of susceptibility and influence and in terms of network centralities as well. This makes the identification of influencers difficult especially since susceptibility and centrality does not always go hand in hand. Here we propose the Top Candidate algorithm, an expert recommendation method, to rank individuals based on their perceived expertise, which resonates well with the assortative nature of innovators and early adopters. Leveraging adoption data from two online social networks that are assortative in terms of adoption but represent different levels of assortativity of network centralities, we demonstrate that the Top Candidate ranking is more efficient in capturing early adopters than other widely used indices. Top Candidate nodes adopt earlier and have higher reach among innovators, early adopters and early majority than nodes highlighted by other methods. These results suggest that the Top Candidate method can identify good seeds for influence maximization campaigns on social networks.
