Recommendation System Types

Hey guys 😀 !!

We all love to watch YouTube, whether its T-Series or Mr. Beast, we consume content of every genre irrespective of their origin.

But have you wondered how a recommendation engine works 🤔🤔!!

So, let's dive into a developer's mind to understand this 🔍.

Generally Recommendation System algorithms are majorly classified as :

1.Content Based Filtering

2.Collaborative Filtering

3.Hybrid Methods

There are other complex systems too, but these are the building blocks.

✔ Content-Based Filtering

Suppose we are the developers, if our consumer consumes some content ,say he/she might be watching some IPL highlights 🏏, then we will recommend him/her similar content of say, other IPL teams highlights.

So we are recommending items to a user based on the attributes or features of the items and the user's preferences.

✔ Collaborative Filtering

Suppose there are 2 users A and B, Both A and B have similar taste of content as it is observed that A watches and likes the same videos as B does. So, it can be concluded that whenever A likes a new video, it must be recommended to B.

E.g - Suppose Ram and Raj loves to watch IPL highlights, after sometime Raj also started watch BBL(Big Bash League) highlights. So , the very next day Ram will have BBL highlights recommendations on its YouTube home page.

So , this approach recommends items to a user based on the preferences and behaviors of other users who are similar to them. This is also called User-based collaborative filtering.

✔ Hybrid Methods

Hybrid system might combine collaborative filtering and content-based filtering to provide more accurate and diverse recommendations.

Initially, YouTube used content-based filtering for recommendations to its user. But now it has opted Hybrid methods to achieve more accurate recommendations.