Core Concepts
The core message of this paper is to develop efficient approximation algorithms with provable performance guarantees for optimizing the social sharing of fresh point-of-interest (PoI) information in location-based social networks (LBSNs).
Abstract
The paper introduces a new combinatorial optimization problem that integrates urban sensing and online social networks to enhance the social sharing of fresh point-of-interest (PoI) information. The authors prove that this problem is NP-hard and existing approximation solutions are not viable.
The key highlights are:
The authors develop a polynomial-time algorithm, Greedy-User-Select (GUS), that guarantees a (1 - m^-2/m * (k-1/k)^k) approximation of the optimum for the static crowdsensing setting, where m is the number of users and k is the budget.
For the mobile crowdsensing setting, where each selected user can move along an n-hop path to sense more PoI, the authors propose an augmentation-adaptive algorithm, Greedy-Path-Select (GPS), that achieves bounded approximation ratios ranging from 1/k(1-1/e) to 1-1/e by adjusting the augmentation factor.
The authors also present a specialized algorithm, Adjusted-GPS, for the case where the sensing graph consists of only user nodes. This algorithm guarantees an approximation ratio of at least 1 - (|V∪V'|-2)/|V∪V'| * (k-1/k)^k, where |V∪V'| is the number of nodes in the sensing graph.
The theoretical results are corroborated by simulations using both synthetic and real-world datasets across different network topologies, demonstrating the practical efficiency of the proposed algorithms.