There’s been a lot of talk about how important metadata is to Social TV. Obviously, without it videos cannot be found. Poor metadata will return poor results. Garbage in garbage out. However, metadata is but one piece of the Social TV puzzle. Video recommendations are a result of a complex intersection of metadata (about the videos) and user profiles. This concept is called adaptation in web science. It’s worth a look if you have time.
Social TV relies on user modelling. Who are you? Accurate video recommendations rely on an exhaustive record of your tastes, mood, interests, etc. And this portrait must be dynamically updated in realtime. Unlike videos, we change and so must our profile. Most second screen social TV applications continually harvest user data (in most cases with your permission) from applications such as Facebook, Twitter, Google, etc. The more data about you these systems can collect, the more accurate your user profile will be. This’ll lead to better recommendations (assuming the recommendation engine is good).
What happens when your user profile is attached to the closed system of the social TV application? In all likelihood it is stuck there in its proprietary format and cannot be shared or exported. Furthermore the full profile is likely hidden and unalterable by you, its creator. Alternatively, there are efforts within the semantic web community to enable users to create, track, edit, and share one decentralized online identity. This is an exciting area that is full of potential to the way we interact and find all types of information on the web. However, this Linked Open Data relies on keeping our profiles distributed and open for all to see. Hence, privacy becomes a major issue.
Who should own your online profile? Should you have the option of keeping it private?