Efficient Budget-Distance-Aware Influence Maximization in Geo-Social Network

Abstract

With the popularity of location-based information sharing in social networks, more and more businesses are trying to advertise their products online. In reality, users are more likely to choose a shopping location that is affordable in price and near in distance, while businesses hope to maximize the number of potential users through a geo-social network. Here the affordable ability for each user is defined as budget. There is an urgent need to fast find a certain number of influencers, so as to influence as many users as possible, which is called the Budget-Distance-Aware Influence Maximization (BDAIM) problem. To overcome this challenge, we propose a BDAIM model and design four pruning rules that can remove low-influence users to reduce time complexity. Then, we develop an algorithm with a $1-1/e$ approximation ratio to find high-influence users by combining these pruning rules. Finally, extensive experimental results on real-world datasets have demonstrated the efficiency and effectiveness of proposed methods.