Mining the network value of customers
Proceedings of the seventh ACM SIGKDD international conference on Knowledge discovery and data mining
Mining knowledge-sharing sites for viral marketing
Proceedings of the eighth ACM SIGKDD international conference on Knowledge discovery and data mining
Maximizing the spread of influence through a social network
Proceedings of the ninth ACM SIGKDD international conference on Knowledge discovery and data mining
The dynamics of viral marketing
ACM Transactions on the Web (TWEB)
Cost-effective outbreak detection in networks
Proceedings of the 13th ACM SIGKDD international conference on Knowledge discovery and data mining
Efficient influence maximization in social networks
Proceedings of the 15th ACM SIGKDD international conference on Knowledge discovery and data mining
Extracting influential nodes for information diffusion on a social network
AAAI'07 Proceedings of the 22nd national conference on Artificial intelligence - Volume 2
Tractable models for information diffusion in social networks
PKDD'06 Proceedings of the 10th European conference on Principle and Practice of Knowledge Discovery in Databases
Hi-index | 0.00 |
Finding the most influential nodes is an important issue in social network analysis. To tackle this issue, Kempe et al. proposed the natural greedy strategy, which, although provides a good approximation, suffers from high computation cost on estimating the influence function even if adopting an efficient optimization. In this paper, we propose a simple yet effective evaluation, the expectation, to estimate the influence function. We formulate the expectation of the influence function and its marginal gain first, then give bounds to the expectation of marginal gains. Based on the approximation to the expectation, we put forward a new greedy algorithm called Greedy Estimate-Expectation (GEE), whose advantage over the previous algorithm is to estimate marginal gains via expectation rather than running Monte-Carlo simulation. Experimental results demonstrate that our algorithm can effectively reduce the running time while maintaining the influence spread.