Analysis of influence contribution in social advertising

Online Social Network (OSN) providers usually conduct advertising campaigns by inserting social ads into promoted posts. Whenever a user engages in ad, she may further propagate the ad to her followers recursively and propagation process is known as word-of-mouth effect. In order spread promotion cascade widely efficiently, OSN provider often tends select influencers, who normally have large audiences over network, initiate campaign. This marketing model, also termed influencer marketing, has been gaining increasing traction investment rapidly becoming one of most widely-used channels digital marketing. this paper, we formulate problem for derive influence contributions influencers given campaign result, considering viral ads, namely contribution allocation (ICA) . We make connection between ICA concept Shapley value cooperative game theory reveal rationale behind ICA. A naive method obtain solution enumerate all possible cascades delivering resulting an exponential number potential cascades, which computationally intractable. Moreover, generating producing exact result non-trivial. Facing challenges, develop linear time under threshold (LT) devise fully polynomial-time randomized approximation scheme (FPRAS) independent (IC) model. Specifically, IC propose efficient approach estimate expected probabilistic graphs modeling OSNs designing scalable sampling with provable accuracy guarantees. extensive experiments show that our algorithms yield solutions remarkably higher quality several baselines improve efficiency significantly.