In this paper we present a contrasting view of a social network, one that focuses on the aggregate messaging patterns of over four million members of Facebook. This approach allows us to characterize users’ behavior on a large scale. And while personal idiosyncrasies and the massive scale of these social networks might lead one to conclude that one is unlikely to discover any strong global patterns of interaction, our analysis discovered a number of strong regularities across the whole network. Most noteworthy, messaging
within facebook exhibits temporal rhythms that are robust and consistent cross campuses and across seasons, resulting from the myriad of individual choices that members make on when and with whom to communicate. For example, among other insights, the data strongly suggest that college students follow two patterns, a “weekend” pattern between midday Friday and midday Sunday, and a “weekday” pattern at all other times. Further, our analysis uncovers a grouping effect whereby students in the same university tend to have similar temporal messaging habits, even when the times of day in question do not appear to be a direct effect of the school schedule.
Read the report Rhythms of social interaction: messaging within a massive online network (pdf) by Scott Golder, Dennis Wilkinson, and Bernardo Huberman of HP Labs. Commentary from SmartMobs.