Maximising influence in sensed heterogeneous social network with privacy preservation. (2018)

• Record Type: Journal Article
• Authors: Han, Meng; Han, Qilong; Li, Lijie; Li, Ji; Li, Yingshu
• Is Part Of: International journal of sensor networks. Volume 28:Number 2(2018)
• Page Start: 69
• Page End: 79
• Publication Date: 2018
• Subjects: sensed data -- location -- social -- data privacy -- influence maximisation; Sensor networks -- Periodicals; 681.2
• Languages: English
• ISSNs: 1748-1279

Abstract

Maximising influence to improve marketing performance has a significant impact on targeted advertisements and viral product promotion, which has become a fundamental problem in social data analysis. Most existing works neglect the fact that location data could also play an important role in the influence prorogation. This paper considers maximising influence towards both sensed location data and online social data with privacy concern. We merge location data from cyber-physical networks and relationship data from online social networks into a unified, then propose an efficient algorithm to solve the influence maximisation problem. Furthermore, our privacy-preserving mechanism could protect the sensitive location and link information during the whole process of data analysis. Real-life datasets are empirically tested with our framework and demonstrate the power of sensed and online data combination to influence maximisation. The experiment results suggest that our framework is outperforming most existing alternative resolutions and succeeds in preserving privacy.