Recommendation System in Machine Learning

In today's world, we are surrounded by machines and different types of gadgets. Machine learning is a technique which prepares machines to think and learn. A recommendation system is an important part of machine learning. Sometimes we get shocked when we find some videos on YouTube related to the topics which we recently searched on Google. It is not any magic. It is the recommendation system. We can define a recommendation system as a system which can predict users' choices based on their previous activities. But it is not appropriate to say that the recommendation system can recommend correct every time. Let's explore some examples for better understanding:

Example 1: It is obvious that when you type in the Google search bar, it shows you some recommended items with your typed word. Google uses your previous searches and other users' searches to recommend this.

Example 2: When we go to Amazon or Flipkart for shopping, we find products similar to our previous purchases. Here also recommendation system is used to predict your required item.

Example 3: In music apps or video apps like Spotify or YouTube, also a recommendation system is used to provide more satisfying service to their users by recommending their intended songs or videos.

Classification of Recommendation System

There are different types of recommendation systems:

1. Collaborative filtering systems: You may have noticed that in some ads they said that the people who have watched the cinema which you recently viewed are also watching other cinemas and they give the list of cinemas. This technique is called a collaborative filtering system. In this technique, it is supposed that if user1 and user2 have liked one item, then they both may like another item also.
   Ex: YouTube shows the content on the basis of your subscribed channels. It means you watch the things which your co-subscribers watch.
   
   Pros and cons: Its main advantage is its simplicity. But it has also disadvantages like it has no opportunity to give new suggestions.
   
2. Content-based systems: For example, we have discussed shopping apps. These types of apps use content-based systems. In this technique, recommendations are provided based on the user's previous activities. The app shows similar products when the user buys a product in his feed.
   
   Pros and cons: It has the advantage of providing the user with a new recommendation depending on its previous activities. But here also exists disadvantages like if a user has never interacted with a product's properties, then it is difficult to show the recommendation of the same.