Recommender System

 In today's digital age, recommender system are essential for helping people find relevant material products and services based on their tastes, since there are so many options available to them that can be overwhelming. Our personalised internet experiences are powered by these clever algorithms which can do anything like suggesting films in different websites and provide shopping recommendations in different online store.

This topic was very different from other lessons and it was also interesting to learn new things. In this lesson we didn't use C language like in other lessons rather we worked with Python. The fundamental idea is that users and customers' interest and behaviours are scanned by computers, which subsequently provide recommendations based on the interest of other specific consumers in groups.

This is also known as collaborative filtering. It makes the use of collective knowledge gathered form users' interaction with objects to forecast a users preferences or interest. Collaborative filtering makes recommendations based on past user behaviour and interaction with things, as opposed to content based filtering which focuses on the attributes of items.

 Another way to view these relationships is based on their similarities and differences as shown in above Venn Diagram.

As part of this study, we created a python project with Anaconda that imports libraries from the internet for movies, users and ratings. Based on the users ratings of other movies and similarities between each movie, an algorithm is then used to generate recommendations for these users. It was a different experience to see how algorithms are made and how they can be used for a variety of different things. Although the code was pretty difficult, it was great to see how these algorithms were used in recommendation system.