Ahsan Hussain, Bettahally N. Keshavamurthy, Ravi Prasad K. Jagannath
Intelligent Automation & Soft Computing, Vol.25, No.3, pp. 473-486, 2019, DOI:10.31209/2018.11007092
Abstract Prediction space of distinct check-in locations in Location-Based Social Networks
is a challenge. In this paper, a thorough analysis of Foursquare Check-ins is
done. Based on previous check-in sequences, next location of social-users is
accurately predicted using multinomial-Hidden Markov Model (mHMM) with
Steady-State probabilities. This information benefits security-agencies in
tracking suspects and restaurant-owners to predict their customers’ arrivals at
different venues on given days. Higher accuracy and Steady-State venuepopularities obtained for location-prediction using the proposed method,
outperform various other baseline methods. More >