"In fact, I’d contend that the friend and follower models can be elegantly replaced by a frequency and engagement model." Quite honestly, I think FB (as an exa

"In fact, I’d contend that the friend and follower models can be elegantly replaced by a frequency and engagement model." Quite honestly, I think FB (as an example) tries to do this already. When I "Like" friend A's feeds more often than friend B's feeds, I see more of friend A's feeds. As far as "engagement" goes, I assume the author means "quality of engagement". How do you begin to come up with a way of quantifying the quality of an engagement for one person versus another? Time of engagement? Number of words spoken/written? "Based on current smartphone technology, social networks can and should leverage location, time, frequency of interaction, and behavioral similarities." Smartphone, laptop, whatever, they're just data collection vehicles. The difficulty is developing a model that maps values to people. Someone optimizing a social network for behavioral similarities, for example, fails if the individuals in the network thrive on behavioral differences. There is plenty of room for improvement with social networks. I don't think the way forward is for social networks to improve on network building optimizations.