The design of query data correlation under mysql-php

The problem is that all the other movies seen by a movie user have taken the first 10 hits by the click-through rate of each movie.
The problem is as good as JD's "bought AAA people also bought product cc ...".

Currently ready to use Php+mysql to get, but feel this is not PHP capable, ask Daniel Guide!

Reply content:

The problem is that all the other movies seen by a movie user have taken the first 10 hits by the click-through rate of each movie.
The problem is as good as JD's "bought AAA people also bought product cc ...".

Currently ready to use Php+mysql to get, but feel this is not PHP capable, ask Daniel Guide!

It should be said that this is not what MySQL is capable of.

Maybe you think group by you order by can do it, but after all you'll find that you can't do it.

PS. From IMDB, until now, a total of 2,877,662 films, 5,899,065 cast, the number of people do not have accurate figures, but how to say there are many billion, imagine your database will put a number of records, and then think about when you join between several tables, What the database will look like ...

Pps. I'll give you some references.
Http://zh.wikipedia.org/wiki/%E5%8D%94%E5%90%8C%E9%81%8E%E6%BF%BE
http://www.quuxlabs.com/blog/2010/09/matrix-factorization-a-simple-tutorial-and-implementation-in-python/

In fact, this can not be real-time statistics, processing all the logic in the background, if not from the front desk request and then to calculate, any language will have problems

This problem is not related to the language and database, because you can not query in real-time, only to find the right time to cache the relevant statistics

