In this final year project we are going to develop a recommender system for an E-Commerce Store which will recommend the products to users which they may like. There are millions of products on E-Commerce stores and it is very difficult for a user to select a product. Therefore our recommender system will help customers by giving them the recommendations about the products of their interest. It will become easy for customers to buy products and of course he will like it and come back again. So by using a recommender system the business of E-Commerce store will increase.
Recommender Systems are basically systems which give us recommendations for items and products which might meet user’s interest. Facebook is a real world example in which users get recommendations for friends in form of people you may know. Also on youtube users get recommendation videos. Similarly our recommender system will recommend products to users of E-Commerce store.
Importance of recommendation system:
Now a day’s people are so busy in their lives and daily routines that they cannot get time for even eating food. In this busy era people like to shop online. This trend is increasing day by day in Pakistan now. So if people go on online stores and they have millions of products in front them which will make them confuse therefore they could not shop properly. Our recommender system will help customers a lot to save their time by recommending them the relevant products that they may need or like. Recommender systems have a great role in most frequently used and exceedingly rated Internet sites like Amazon.com, Youtube.com, Yahoo, Trip Advisor, last.fm and IMDb.
Applications of recommender system:
Now a days recommender systems are being used in following domains:
E-Commerce: Here system recommend the products according to user’s interest and need.
Service: System recommend services like Hotel and Travel services.
Entertainment: This domain belongs to websites where system recommend movies and music etc.
Content: Recommend reading material like newspapers and books.
Social Networks: All types of systems that recommend something on social networking websites like facebook does by giving users friend suggestions.
Medical Informatics: Recommend treatment, also help in diagnosing the medical issues with patient like detecting tumors.
Advertising: Recommend ads to interested users of the products.
1.Intelligent Tourism Recommender System:
In 2014 1 author make a recommender system for tourists to customize their tours. It will recommend places based on their taste and restrictions using personalization techniques. This system automatically learned user’s preferences through their feedbacks. This system used automated planner to schedule the recommendations within a route which can span several days of the tourists. Clustering algorithms are used to classify tourists having same tastes. Reasoning procedure are used to deduced user’s preferences. The system also tell about the opening and closing timing of attractions.
2.The Youtube video recommender system:
In 2010 2 the author built a recommender system for Youtube videos recommendations. According to a survey in one minute users upload more than 24 hours of videos. Therefore there is huge need of recommendations. This system recommend videos to only signed in users based on their previous history. There will be limited recommendations for signed out users. Authors face some problems like quality of video is is very poor. And mostly on Youtube the duration of video is not so long so interaction is short and noisy. Authors convert data in two broad classes:1) content data. 2) user activity data on based of which the system did recommendations.
Motivation and Scope:
The field of data science has appeared in response to the amplified data richness in industry, science, and engineering and when it comes to NLP, it uses machine learning techniques for the predictions and recommendations.
The motivation to develop this Recommendation System is only to save the important time of people so that they can use it at some other task which is more important than shopping. Also due to such recommender systems people will prefer online shopping and the online business will increase.
Goals and Objectives:
Our goal is to built such recommender system which will recommend costumers
Tools and Technologies:
Following tools and technologies will be used for our FYP project:
scikit-learn (aka sk-learn)
RiVal (An open source toolkit for recommender system evaluation)
Surprise (A Python scikit for building, and analyzing (collaborative-filtering)
Recommender systems. Various algorithms are built-in, with a focus on rating