With the popularity of location-serving social networks, user
behavior analysis of location service social network is one of the
research priorities. Literature concludes that the number of times a
user has checked in and the location of the visit follow a normal
logarithmic distribution. At the same time, the analysis also shows
that the number of users’ check-in does not increase as the
number of friends of the user increases. Literature mainly makes
analysis of the usage habits of users of location service social
networks, such as their usage time, their main access area, and so
on. Furthermore, some literature study the similarity of users in
behavioral trajectories through a large amount of geographic
location data. The literature takes advantage of the Global
Positioning System (GPS) log to make analysis of user's behavior
trajectory. Since the GPS log can record the user's trajectory in
detail, it is suitable for analyzing the user's behavior pattern. The
literature identifies the user community through the user's access
trajectory as well as calculate the similarity of the user's activity
trajectory through the GPS log, At first, it identifies the
geographical location points once visited by the user, and then
clusters them. User similarity is calculated by matching the user’s
access sequence at the location after clustering. Literature used the
Foursquare data in the location service social networking site to
examine the semantic similarity of users in terms of location.
Since Foursquare has semantic classification of geographic
location data, the article calculate the semantic similarity of user
access behavior through location classification information. Since
the GPS log can continuously track the user's behavior track, the
user in the social network of the location service, only signs in
after arriving at a certain location, and does not continuously track
the user's behavior track. Furthermore, the user‘s sign-in has