Abstract
In network science several topology-based link prediction methods have been developed so far. The classic social network link prediction approach takes as an input a snapshot of a whole network. However, with human activities behind it, this social network keeps changing. In this paper, we consider link prediction problem as a time-series problem and propose a hybrid link prediction model that combines eight structure-based prediction methods and self-adapts the weights assigned to each included method. To test the model, we perform experiments on two real world networks with both sliding and growing window scenarios. The results show that our model outperforms other structure-based methods when both precision and recall of the prediction results are considered.
| Original language | English |
|---|---|
| Title of host publication | Proceedings of the 2016 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining, ASONAM 2016 |
| Publisher | Institute of Electrical and Electronics Engineers Inc. |
| Pages | 1221-1228 |
| Number of pages | 8 |
| ISBN (Print) | 9781509028467 |
| DOIs | |
| Publication status | Published - 21 Nov 2016 |
| Event | 2016 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining, ASONAM 2016 - San Francisco, United States Duration: 18 Aug 2016 → 21 Aug 2016 |
Conference
| Conference | 2016 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining, ASONAM 2016 |
|---|---|
| Country/Territory | United States |
| City | San Francisco |
| Period | 18/08/2016 → 21/08/2016 |