Recommendation algorithm

Recommendation algorithm is an algorithm in computer science.

As follows:

1. What is the algorithm? Why do we need algorithms?

The annual data generated in the world will increase from 33ZB ten trillion bytes in 20 18 years to 175ZB, which is equivalent to generating 49 1EB data every day. Essentially, an algorithm is "an idea expressed in a mathematical way or computer code". Among them, the recommendation system is an information filtering system to help users reduce the waste of time and energy caused by browsing a large number of invalid data.

2. What are the key events in the history of recommended technology development?

Information overload is a concept that existed in 1980s and 1990s. With the rapid development of information technology and Internet, human beings have moved from the era of information scarcity to the era of information overload. By the mid-1990s, researchers began to try to solve the problem of information overload by predicting users' ratings of recommended items, contents or services. Recommendation system also appears as an independent research field.

3. What are the types of recommendation systems?

Since the development of recommendation system, its core technologies can be roughly divided into collaborative filtering-based recommendation method, content-based recommendation method and mixed recommendation method. The recommendation method based on collaborative filtering is essentially to recommend favorite items, contents or services to similar users according to their similar preferences.

According to the description information of items, labels and other related information, user-related information and users' comments, collections, likes, views, browsing and clicks on items, a recommendation algorithm model is constructed.

4. Will the recommendation algorithm narrow the information?

In the impression of the outside world, personalized recommendation is like a funnel, which will match the recommended content with users and tend to recommend content that is highly in line with their preferences, resulting in narrower and narrower recommended content.

Recommendation technology is not simply "doing what you like". In the opinion of some experts, if we can deeply stimulate and meet the potential needs of users on the basis of recommending the content that users are interested in, then the algorithm can better meet people's multi-dimensional demands for information.