Effective social promotion networks require reliable algorithms for discovering events and targeting their locations to online users. But what are the underlying factors that affect ties among online social network (OSN) users? The authors examine the relationship between user characteristics and their OSN connections.

The user traits investigated in the research include user weight (number of friends), number of check-ins, lifespan, “the maximum amount of movements made by each user,” and density (number of check-ins per day). The inside and outside fraction (IO-fraction)--“density outside the home is divided by density inside the home”--is used to gauge the association between distance, user activities, and user attributes.

Experiments performed on data from three large-scale, location-based social networks examine the effects of user characteristics on OSN connections; the results are used to develop an algorithm for computing the likelihood of friendship materialization in OSNs. The IOF is used to exhibit the associations between mobility and user characteristics.

Data analysis results reveal that the movement of users is relevant to their OSN activities; the probability of a pair of users forming a friendship within a distance can be rationally computed; and friendship probability and user weight, activity, lifespan, and movement can be estimated.

The authors clearly recognize that the datasets used for the various research investigations do not contain “a time labeled friendship graph and location information in each time for every user.” Consequently, the study fails to show the establishment of social connections. In the face of this limitation, I call on computational statisticians and online marketing research analysts to read this interesting paper and to help identify more reliable datasets and accurate models for advertising specific events to customers based on individual behavior.