How Netflix used the big data to hold the poker house on fire.

&http://www.aliyun.com/zixun/aggregation/37954.html ">nbsp;

One of the most popular American dramas of all time. When it comes to counting the cards, Netflix spends 100 million of dollars on copyright, David Fincher and old puppet Kevin Spacey, who first hit the original series and became the hottest online series in the US and more than 40 countries. So, before filming, did Netflix know that the House of Cards would fire?

In the American television industry, nothing is certain. Maybe you've got the gold director, the Power actor and the popular theme script, and the results are still out. In any business, it would be scary to foresee the future, and Netflix could have approached that level in the Card House war.

As the world's largest online film rental nen service provider, Netflix knows what people like to see almost as much as anyone. It already knows that the user likes Fincher (social network, seven Sin director), also know Spacey starring film performance is Good, also know that the British play version of the "card House" is very popular, the intersection of the three shows that it is worth gambling on this matter.

Netflix has 27 million subscribers in the United States and 33 million in the world, and it knows better than anyone what kind of movies and TV they like. Research has shown that daily peak-hour web downloads are from Netflix's streaming services, and last year people watched streaming video online more time than physical DVDs. Every day users generate more than 30 million behaviors on Netflix, such as when you pause, replay, or fast-forward, and the Netflix subscriber gives you 4 million ratings a day, and 3 million search requests, asking about the time and equipment for the episodes to play.

The success of the card house has benefited from the huge amount of user data accumulated and analyzed by Netflix, but some people think that big data mining and analysis can only tell you what the audience liked before, can't tell you what the audience will like in the future, which algorithm can predict the success of "South Park" such a dark horse film?

Of course, in terms of specific ideas, Netflix relies on people rather than algorithms. Recruit the right people, give them enough freedom and a budget, and then OK. The two guest performers invited by Netflix's next original play, "arrested Development", are not algorithmic.

The digging of big data is increasingly seeping into every aspect of life, from the Obama campaign's use of data analysis fundraising to Nate Silver, who successfully predicted the results of the 50 state elections in 2012, recently Microsoft Research also said it had successfully predicted most of the Oscar awards. In the near future, the results of large data mining may be more accurate than the intuitive judgment of an industry veteran.

(Responsible editor: Lu Guang)