Abstract
Location-Based Social Networks (LBSNs) have been used in many applications including crowd estimation and event detection. Snapchat is an LBSN that offers real-time activity monitoring around the world through its Snap Map using aggregation of user-submitted snaps. In this paper, we used Snap Map to study the crowd behavior at the Grand Holy Mosque (GHM) in Mecca during the holy month of Ramadan. We tracked activities in specific locations around the GHM for a month. We then used this temporal/spatial data to study crowd density and the behavior of GHM's visitors and compared them with the reported figures published by GHM's officials. Initial results show that Snap Map can be useful in understanding crowd behavior and detecting certain events, and may potentially be used in crowd size prediction.