Joint Selection of Influential Users and Locations under Target Region in Location-Based Social Networks

• Published in: Sensors (Basel, Switzerland) Vol. 21; no. 3; p. 709
• Main Authors: Ali, Khurshed, Li, Cheng-Te, Chen, Yi-Shin
• Format: Journal Article
• Language: English
• Published: Switzerland MDPI AG 21-01-2021; MDPI
• Subjects: Algorithms; Approximation; Collaboration; Heuristic; influence maximization; Location based services; location-based social networks; Marketing; Network analysis; recommendation system; Site selection; Social network analysis; Social networks; viral marketing; Word of mouth advertising; influence maximization; viral marketing; location-based social networks; recommendation system; social network analysis
• ISSN: 1424-8220; 1424-8220
• DOI: 10.3390/s21030709

Influence Maximization problem, selection of a set of users in a social network to maximize the influence spread, has received ample research attention in the social network analysis domain due to its practical applications. Although the problem has been extensively studied, existing works have neglected the location's popularity and importance along with influential users for product promotion at a particular region in Location-based Social Networks. Real-world marketing companies are more interested in finding suitable locations and influential users in a city to promote their product and attract as many users as possible. In this work, we study the joint selection of influential users and locations within a target region from two complementary perspectives; general and specific location type selection perspectives. The first is to find influential users and locations at a specified region irrespective of location type or category. The second perspective is to recommend locations matching location preference in addition to the target region for product promotion. To address general and specific location recommendations and influential users, we propose heuristic-based methods that effectively find influential users and locations for product promotion. Our experimental results show that it is not always an optimal choice to recommend locations with the highest popularity values, such as ratings, check-ins, and so, which may not be a true indicator of location popularity to be considered for marketing. Our results show that not only influential users are helpful for product promotion, but suitable influential locations can also assist in promoting products in the target region.