appledora/mlpr-amazon_product_recommendation

Product Recommendation System using Amazon Electronics Dataset

Product Recommendation System using Amazon Electronics Dataset

Introduction to Recommendation systems

In this modern world we are overloaded with data and this data provides us the useful information. But it's not possible for the user to extract the information which interest them from these data. In order to help the user to find out information about the product , recommedation systems where developed.

Recommeder system creates a similarity between the user and items and exploits the similarity between user/item to make recommendations.

What recommeder system can solve ?

1. It can help the user to find the right product.
2. It can increase the user engagement. For example, there's 40% more click on the google news due to recommendation.
3. It helps the item providers to deliver the items to the right user.In Amazon , 35 % products get sold due to recommendation.
4. It helps to make the contents more personalized.In Netflix most of the rented movies are from recommendations.

Types of recommendations

There are mainly 6 types of the recommendations systems :-

1. Popularity based systems :- It works by recommeding items viewed and purchased by most people and are rated high.It is not a personalized recommendation.

2. Classification model based:- It works by understanding the features of the user and applying the classification algorithm to decide whether the user is interested or not in the prodcut.

3. Content based recommedations:- It is based on the information on the contents of the item rather than on the user opinions.The main idea is if the user likes an item then he or she will like the "other" similar item.

4. Collaberative Filtering:- It is based on assumption that people like things similar to other things they like, and things that are liked by other people with similar taste. it is mainly of two types: a) User-User b) Item -Item

5. Hybrid Approaches:- This system approach is to combine collaborative filtering, content-based filtering, and other approaches .

6. Association rule mining :- Association rules capture the relationships between items based on their patterns of co-occurrence across transactions.

Learning Outcomes:

Online E-commerce websites like Amazon, Flipkart uses different recommendation models to provide different suggestions to different users. Amazon currently uses item-to-item collaborative filtering, which scales to massive data sets and produces high-quality recommendations in real-time.

● Exploratory Data Analysis

● Creating a Recommendation system using real data

● Benchemarking Popularity based models, Matrix factorization model and Collaborative filtering

Data Description:

Link : https://www.kaggle.com/saurav9786/amazon-product-reviews?fbclid=IwAR26xfJJK1c5zEeXh8yK9tTgDZrXSkdZIIfjGOhnegCobgskrUEYFcky_cw

The dataset here is taken from the below website.

Source - Amazon Reviews data (http://jmcauley.ucsd.edu/data/amazon/) The repository has several datasets. For this case study, we are using the Electronics dataset.

Domain:

E-commerce

Attribute Information:

● userId : Every user identified with a unique id

● productId : Every product identified with a unique id

● Rating : Rating of the corresponding product by the corresponding user

● timestamp : Time of the rating ( ignore this column for this exercise)

Steps and tasks:

1. Read and explore the given dataset. (Rename column/add headers, plot histograms, find data characteristics)

2. Take a subset of the dataset to make it less sparse/ denser. ( For example, keep the users only who has given 50 or more number of ratings )

3. Split the data randomly into train and test dataset. ( For example, split it in 70/30 ratio)

4. Build Popularity Recommender model.

5. Build Collaborative Filtering model.

6. Evaluate both the models. ( Once the model is trained on the training data, it can be used to compute the error (RMSE) on predictions made on the test data.

7. Get top - K ( K = 5) recommendations. Since our goal is to recommend new products for each user based on his/her habits, we will recommend 5 new products.

8. Summarise your insights.