Thematic subscriptions in social networks
In modern social networks of general purpose, such as Facebook, we have a situation where, roughly speaking, everyone writes about everything. To stand out in such conditions, you must 1) either be / become an actual journalist, able to find the most popular topics and interesting to write about them, or 2) to bring interest in yourself from the real, to be popular or to be successful in something in traditional society, or 3) to specialize in something and “beat at one point”, gradually forming the thematic community around it.
But imagine a situation where you could subscribe to a topic, tag, interest. Let's say the theme “start-ups” was liked and you get into the tape from completely different sources, people, organizations, media. Here, of course, the automatic ranking mechanism will begin to play a significant role, determining what is worthy of getting into your tape in the first place. This is at least a more meaningful task than what Facebook is now doing with its EdgeRank, trying to decide for me which of my forty friends is more interesting to me. In this version, the strategy of recruiting popularity would probably change. You would have a chance to stand out by writing even one post. Such a system would probably be more curious and productive than what it is now.
Until now, it was about reading mode. It would be interesting to also experiment with the recording mode. For example, in some social network you could not write about anything except posts on a strictly limited set of topics. This is not exactly the same as what we now have with specialized thematic social networks - the latter set only a general direction. The mentioned choice of topics does not necessarily belong to the general direction. It may be quite different. This is a matter of experiment. What will give a restriction in writing? - perhaps just a simplified ranking. After all, for example, Google should solve simultaneously two different non-trivial tasks on a search query - in order for the links found to be, firstly, the most relevant, and secondly, among the relevant ones, ranked by quality. If you remove or simplify the task of determining relevance, then you can focus more on the task of ranking by quality.
But imagine a situation where you could subscribe to a topic, tag, interest. Let's say the theme “start-ups” was liked and you get into the tape from completely different sources, people, organizations, media. Here, of course, the automatic ranking mechanism will begin to play a significant role, determining what is worthy of getting into your tape in the first place. This is at least a more meaningful task than what Facebook is now doing with its EdgeRank, trying to decide for me which of my forty friends is more interesting to me. In this version, the strategy of recruiting popularity would probably change. You would have a chance to stand out by writing even one post. Such a system would probably be more curious and productive than what it is now.
Until now, it was about reading mode. It would be interesting to also experiment with the recording mode. For example, in some social network you could not write about anything except posts on a strictly limited set of topics. This is not exactly the same as what we now have with specialized thematic social networks - the latter set only a general direction. The mentioned choice of topics does not necessarily belong to the general direction. It may be quite different. This is a matter of experiment. What will give a restriction in writing? - perhaps just a simplified ranking. After all, for example, Google should solve simultaneously two different non-trivial tasks on a search query - in order for the links found to be, firstly, the most relevant, and secondly, among the relevant ones, ranked by quality. If you remove or simplify the task of determining relevance, then you can focus more on the task of ranking by quality.
')
Source: https://habr.com/ru/post/144922/
All Articles