Content-based filtering

[rochon.a1.119]

It's like a kind of database segmentation to match users with the same tastes and create a general profile.

This grouping can reveal new information that will allow for improved marketing actions for each segment in the future.

This is the type of filtering that we can see in some e-commerce sites.


It is a filtering based on the keywords used by the user and the items that the user may like.

Based on this information, algorithms analyze the elements and offer more personalized content on future visits.

3.- Filtering based on demographic profile
This type of filtering does not require user history to work on germany phone number list recommendations; instead, suggestions are made based on different demographic profiles.

It may be the least accurate, but it still works.

I will demonstrate this later.

4.- Hybrid filtering
As its name suggests, this type of filtering seeks to mix the attributes of all the previous ones, to provide a much more complete and personalized experience.

An example is Netflix.

You can see that the content shown to a user in Spain is not the same as that shown to a user in LATAM.