This study (.pdf) delves into programmatic methodology that can be used to predict strong and weak ties between users of a social network. From a community manager’s perspective, this is important because predictive activities can alleviate some oversight tasks while intelligently satisfying the needs of community members. As an example of practical implications of their research, the authors note that:
When users make privacy choices, a system could make educated guesses about which friends fall into trusted and untrusted categories.
Consider a politician or company that wants to broadcast a message through the network such that it only passes through trusted friends.
From a marketing perspective, it’s important to understand this as way to drive customer loyalty because as social networks continue to grow, predictive systems that deliver more relevant information in meaningful ways will drive overall customer loyalty. This could be a huge value add for such social network marketing/branding services like Facebook pages.
Related post: Peering Under the Hood at Facebook