A Factorization Machines-based Participant Recruitment Approach in Mobile Crowdsensing
Hanyang Ning, Miao Ma, Bo Yang · 2022
Mobile Crowdsensing employs mobile devices to get massive data efficiently and economically, which is a people-centric sensing paradigm followed with the development of mobile communication technology. The key issue of mobile crowdsensing is to recruit suitable participants. Here we propose a new approach on participant recruitments that introduces factorization machine to cross features of participants and tasks in pairs to reduce the number of weight parameters and accelerates the process of model training. In this approach, both the quality of historical task implements and the constraints of time and location are integrated to mine the ability of each potential participant. Compared with some state-of-the-art approaches on Gowalla dataset, experimental results show that our approach is the best in term of precision, recall, hit rate, normalized discounted cumulative gain and mean reciprocal rank.