Scalable Methods for Adaptively Seeding a Social Network

TL;DR

Algorithms for linear influence models with provable approximation guarantees that can be gracefully parallelized are developed, and adaptive seeding is shown to be scalable and to obtains dramatic improvements over standard approaches of information dissemination.

Abstract

In recent years, social networking platforms have developed into extraordinary channels for spreading and consuming information. Along with the rise of such infrastructure, there is continuous progress on techniques for spreading information effectively through influential users. In many applications, one is restricted to select influencers from a set of users who engaged with the topic being promoted, and due to the structure of social networks, these users often rank low in terms of their influence potential. An alternative approach one can consider is an adaptive method which selects users in a manner which targets their influential neighbors. The advantage of such an approach is that it leverages the friendship paradox in social networks: while users are often not influential, they often know someone who is. Despite the various complexities in such optimization problems, we show that scalable adaptive seeding is achievable. In particular, we develop algorithms for linear influence models with provable approximation guarantees that can be gracefully parallelized. To show the effectiveness of our methods we collected data from various verticals social network users follow. For each vertical, we collected data on the users who responded to a certain post as well as their neighbors, and applied our methods on this data. Our experiments show that adaptive seeding is scalable, and importantly, that it obtains dramatic improvements over standard approaches of information dissemination.

Figures (10)

• Figure 1: CDF of the degree distribution of users who liked a post by Kiva on Facebook and that of their friends.
• Figure 2: Comparison of the average degree of the core set users and the average degree of their friends.
• Figure 3: Performance of adaptive seeding compared to other influence maximization approaches. The horizontal axis represents the budget used as a fraction of the size of the core set. The vertical axis is the expected influence reachable by optimally selecting nodes on the second stage.
• Figure 4: Ratio of the performance of adaptive seeding to IM. Bars represents the mean improvement across all verticals, and the "error bar" represents the range of improvement across verticals.
• Figure 5: (a) Performance of adaptive seeding for various propagation probabilities. (b) Performance of adaptive seeding when restricted to the subgraph of users who liked HBO (red line).

Theorems & Definitions (14)

• Proposition 1
• Proposition 2
• Lemma 1
• Lemma 2
• Lemma 3
• Proposition 3
• Proposition 4
• Proposition 5
• proof : of Proposition \ref{['prop:gap']}
• proof : of Proposition \ref{['prop:cr']}