Familiarity. Familiarity indicates how frequently a user visits a place. We define the familiarity of node x that contains k stay behaviors, motivated by the entropy in [9], as follows:
A place will have high familiarity if a user visits it with certain regularity (e.g., every day). Conversely, a place will have low familiarity if the distribution of stay at a place is biased or randomly observed. We use familiarity to determine the privateness of a place for each user. For example, employees at a restaurant exhibit high familiarity, as they regularly come to the restaurant for work. Conversely, customers at a restaurant show relatively lower familiarity, due to the randomness of their stay behavior.