Ask, the movie website "You also may like the following movie" This function how to realize?

Ask, the movie website "You also may like the following movie" This function how to realize?
What points of knowledge are involved?

Reply content:

Ask, the movie website "You also may like the following movie" This function how to realize?
What points of knowledge are involved?

Point of interest search, such as when you search for a movie, a certain type of time, the system will be based on the keywords you search, categories and other key tags to search for relevant results, to find the most matching the most similar information to you;

The site will record the keywords you searched for and then sort them out, giving you a similar name for your movie.

This is the use of big data. What types of movies do you often search for or watch, who are the actors, who are the directors, the player or the browser, the data, and finally the data analysis to recommend the movies you like.
Big data is almost ubiquitous in our lives and is widely used ...

Use the movie category or tag to achieve

give you a reference. Write the presentation ID of the movie type to the cookie when the user browses to your site's movie

So when a user browses to another page: Can have more of his own cookies to push the videos of interest that he wants to see

can refer to this paper: http://webpages.uncc.edu/sakella/courses ...

It should involve referral systems, calculate correlations based on recommended systems, and then match your search terms.

Recommendation System.
In short, it is to extract some of the characteristics of the film, such as the type, age, starring and so on. Then find the most relevant movies based on these features that users might like.
For the recommendation system has a special algorithm, is a discipline, you can search.

This is a very interesting question and of course it can be answered.
For example, you say the movie, the movie has many types, the love movie, the literary film, the disaster film Ah, what and so on. Imagine that there should be a movie type field in the database, which can be as simple as a number type, such as (1 disaster Movie 2 zombie Movie 3 comedy ...) And so on
Then is to achieve your function, click on the movie, go to the movie Details page, this time the database has two steps, the first is to read the information of this movie, the second step is based on the type of this video field values to find the same type of movie, and finally displayed on the page.
OK, that's simple.

This is a typical collaborative filtering scenario, the basic idea that you grab all the bangs
User's browsing data, and then find the user who is most similar to the current user, and then recommend his browsing to the current user. This is based on the user dimension of the filter, but also based on the similarity of the film, give a bunch of results, similar calculations can be quantified with the number of common user visits.

You can refer to the books related to data mining.

Depending on the movie properties such as director actor type.

You can think of the n properties of the movie (actor, type, rating, etc.) as n dimensions (and possibly plus a user dimension), a total of n (n+1) dimensions.
A movie is equivalent to a vector of this multidimensional space in mathematics. What we need to find is a vector group that has the smallest angle between the dimensions and the vector, which seems to be the cosine value. It's also a movie you might like.

