What is collaborative filtering?

Cooperation filtering is a data processing method that relies on the use of data from many sources to develop the profiles of people who are associated with similar taste and expenditure habits. This technique is used in a number of different settings. Some of the most famous collaborative filtering applications can be seen on the Internet, where it is used for marketing, to predict the taste of users and curator sites that rely on the input from users to work.

In a simple example of how cooperating filtering works, the site might want to set up a recommendation system for TV shows. Web users provide data when they log in and report programs they like. This data is used to identify users with similar taste. If 75% of people who like shows like show B can conclude that people who like one show probably like the other. Therefore, when the user logs in and identifies as a fan that shows how they look for designs, the system can recommend SHOw b.

for collaborative filtering to work needs a lot of data. The larger the population from which the data is drawn, the more useful and efficient the data will be. A small amount of data is more likely to end up results that do not have meaningful, such as false connections that result in poor predictions of taste. Such systems often suffer from a problem with a cold start in which they are slowly evolving, as the database must be filled first. Soon the adoptors can frustrate the system because it makes bad recommendations because it does not have enough data.

Cooperation filtering is also widely used on social networks and websites that provide tools such as enterprise bookmarking in which users share and promote links to sites that they consider interesting. As users add to the data file in the system, the system can begin to issue recommendations that are designed to the taste of every user. For example, a social bookmark site can generate random N linksand basis of links and users that someone has expressed in the past.

Merchants can provide users very targeted marketing using collaborative filtering. This personalized marketing can be highly effective because users feel that they are personally solved, and it is more likely to accept recommendations. The huge amount of data provided voluntarily on the website such as social networks are a hot commodity among traders that buys data from such pages for the development of adapted campaigns.