You are on page 1of 22

“Getting information off the


Internet is like taking a drink
from a fire hydrant!”
-Mitchell Kapor
The World is an over-crowded place!
We are overloaded!
Recommender
Systems
Presented by,
Bhanu Prakash
B141895
Contents :
What are Recommender Systems ?
Working of Recommender Systems
Data Collection

Data Storage
Filtering the Data
Content based Filtering

Collaborative Filtering
Applications
Conclusion
References
What are Recommender Systems?
Automated form of a “shop counter guy”.

Filters the data using different algorithms.

Recommends the most relevant items to users.


It first captures the past behavior of a customer
and based on that, recommends products which
the users might be likely to buy.
Working :

Step 2 Algorithm 1
Data Storage Content based
Filtering

Step 1 Step 3 Algorithm 2


Data Collection Filtering the Data Collaborative
Filtering
Step 1 : Data Collection
Explicit data Implicit data

Data provided Gathered from


intentionally. available data
streams.
Step 2 : Data Storage
Standard SQL Database

NoSQL Database
Step 3 : Filtering the Data
To extract the relevant information.

Required to make the final recommendations.

Algorithm 1 : Content based Filtering.


Algorithm 2 : Collaborative Filtering.
Algorithm 1 : Content based Filtering
Recommends products which are similar to the ones that a
user has liked in the past.
EXAMPLE OF NETFLIX :
Profile Vector : Contains past behavior of the user.
Item Vector : Contains all information related to movies.
Cosine Similarity : Similarity between two vectors.
Algorithm 1 Continued...
Movies are arranged in descending order based on cosine
values.
One of the two below approaches is used for
recommendations:
Top-n Approach : n-value
Rating scale Approach : threshold value
DRAWBACK:It will never recommend
products which the user has
not bought or liked in the past
Algorithm 2 : Collaborative Filtering
Finds a subset of users who have similar tastes and
preferences to the target user.
Uses this subset for offering recommendations.

Two approaches :
(a)User Based
(b)Item Based
Algorithm 2 : (a)User Based
This algorithm first finds the similarity score between users.

Picks out the most similar users.

Recommends products which these similar users


have liked or bought previously.

Prediction :

Pu,i is the prediction of an item.


Rv,i is the rating given by a
user v to a movie I.
Su,v is the similarity between
users.
Algorithm 2 : (a)User Based Contd...
Useful only when number of users is less.

Not effective when number of users are more.


Algorithm 2 : (b)Item Based
Compute the similarity between each pair of items.

Looks for items that are similar to the articles that user
has already rated and recommend most similar articles.

Quite stable in itself as compared to User based


collaborative filtering.
Similarity between Items :

Prediction :
Algorithm 2 : (b)Item Based Contd...
Far less resource consuming than user-user collaborative
filtering.
Applications :

Facebook

Google
E-Commerce Sites(Amazon,Flipkart...)
Music – (Saavn , Gaana...)
Food – (Swiggy , Zomato...)
Movie – (BookMyShow , Netflix...)
Conclusion :

Recommender systems open new opportunities of


retrieving personalized information on the Internet. It
also helps to alleviate the problem of information
overload which is a very common phenomenon with
information retrieval systems and enables users to
have access to products and services which are not
readily available to users on the system.
References :
https://www.analyticsvidhya.com/blog/2018/06/comprehensive-
guide-recommendation-engine-python/
https://en.wikipedia.org/wiki/Recommender_system

https://www.youtube.com/watch?v=9gBC9R-msAk

https://medium.com/@cfpinela/recommender-
systems-user-based-and-item-based-collaborative-
filtering-5d5f375a127f
Thank You!
Any questions?

You might also like