E-COMMERCE

RECOMMENDATION
SYSTEM
BACKGROUND

• In the modern digital marketplace, the vast number of products available

online, customers often face difficulties in finding items that match their
preferences and needs.
• This challenge has driven the need for personalized shopping experiences,
leading to the development of recommendation systems.
• It makes the shopping process more efficient and enjoyable for customers.
 STATEMENT OF
PROBLEM
• Customers can feel overwhelmed by the sheer number of available products,
making it difficult for them to find items that match their preferences.
• Decreased likelihood of conversion, leading to missed sales opportunities.
• Companies falling behind their rivals, due to lack of adequate recommendation on
online markets.
• General aim is to enhance user experience, increase customer satisfaction, and
boost sales, provide a competitive edge and drive business growth.
OBJECTIVE OF THE
PROJECT
GENERAL PURPOSE OBJECTIVE
• Build a robust and advanced e-
SPECIFIC PURPOSE OBJECTIVE
• Collecting and examining data to
commerce recommendation understand what users like and
system need
• increase customer satisfaction, • filtering and content-based
boost sales, and provide a filtering algorithms in
competitive edge combination with hybrid
algorithm to create a
sophisticated recommendation
engine.
• Evaluate the performance and
effectiveness of the
recommendation system,
adjusting when necessary to
ensure it meets our goals
FEASIBLITY
Technical feasibility TECHNICAL FEASIBLITY
• we made sure that the • IMPLEMENTING THE
technological requirements are RECOMMENDATION SYSTEM
available and are at hand. REQUIRES MINIMAL CHANGES TO
• considering level of experience THE CORE OPERATIONS OF AN E-
we’ve gained through the years COMMERCE PLATFORM.
development of the system is • THE SYSTEM CAN BE SEAMLESSLY
possible. INTEGRATED INTO THE EXISTING
WEBSITE
 SCOPE OF THE
PROJECT
• The scope of this recommendation system includes the design, testing,
and implementation of a complete E-commerce recommender system
will be covered.
SIGNIFICANCE OF THE
PROJECT
• The significance of implementing an e-commerce
recommendation system is multifaceted
• Improving Customer Experience
• Increase sales
• Providing a Competitive Edge
TARGET BENEFICIARIES
• Customers: Personalized suggestions based on their interests give
customers a better shopping experience. E-commerce platforms
can recommend goods that suit personal preferences by utilizing
data.
• E-commerce Enterprise: tailored recommendations boosts sales
and improves client retention for e-commerce companies.
• Payment processors: third party payment processors that
facilitate the online transactions on the E-commerce website.
• Team Marketing: Analyzing customized recommendations gives
the marketing team important information about consumer
preferences and behavior. It aids in understanding customer needs.
METHODOLOGY
Data source Fact finding technique
• We have used dataset from kaggle and • Observation: We will use observation to
different sources which includes the get the right information and understand
results gathered from a customer survey the by observing how the current system
for an E-commerce platform. works.
• 80% of the data is used for training the • Document Analysis: Reviewing existing
model and the remainder 20% data will that are related to what we want to build.
be used for testing the model.
 LIMITATIONS OF THE
PROJECT
• Cold start problem: in case of new users, the system does not have
information about their preferences in order to make recommendations.
• Sparsity: this refers to the presence of a large number of empty or
zero-valued data points in a dataset.
 INTRODUCTION OF EXISTING
SYSTEM
• Electronic Commerce or e-commerce is business transactions that take
place by communication
• Online shopping is a form of electronic commerce where the buyer is
directly online to the seller's computer usually via the internet.
PLAYERS IN THE E-COMMERCE
SYSTEM
• Customers
• Merchants (Sellers or Vendors)
• Payment Gateways
• Logistics and Delivery Partners
• Technology Providers
• Advertisers and Marketers
• Regulators and Legal Authorities
• Third-Party Service Providers
MAJOR FUNCTIONS/ACTIVITIES IN
THE EXISTING SYSTEM
• Product Catalog Management-From adding new products to updating existing
listings, this function ensures a seamless display of offerings

• Secure Online Transactions-These systems integrate payment gateways and

implement security measures to protect customer information during the
purchase process.

• Order Processing and Management-Efficient order processing and management

streamline the post-purchase workflow
• Customer Relationship Management (CRM)
• Integrated Analytics and Reporting
BUSINESS RULES & REPORT
GENERATED IN THE EXISTING SYSTEM
Business Rules Reports
• GDPR (User Data Privacy and Usage) • E-commerce overview dashboard
• Forming a Business Entity • Sales reports
• Paying Taxes • Sales by product
• Choosing a Payment Gateway • Profit margin
• Using Trademarks, Patents, and • Finances summary report
Copyrights Forms and other documents
• Understanding Shipping Restrictions: • User Registration Form
• Understanding Age Restrictions • Order Form
• Following PCI Compliance • Payment Form
• Customer Satisfaction Survey
PROPOSED SOLUTION

• The proposed work aims to develop a modern recommendation system for e-

commerce platforms, centered around collaborative filtering techniques content
based filtering and hybrid mechnaism.

• A Collaborative filtering approach is essential for e-commerce recommendation

systems because of its ability to provide personalized recommendations by
leveraging user behavior patterns and preferences.

• A Collaborative filtering approach is essential for e-commerce recommendation

systems because of its ability to provide personalized recommendations by
leveraging user behavior patterns and preferences.
FUNCTIONAL REQUIREMENTS

User Personalization:
Recommend products based on user behavior, browsing history, and purchase history.

Product Suggestions:
Display "Similar Products," "Frequently Bought Together," and "Trending Items."

Search-Based Recommendations:
Show relevant suggestions based on search keywords.

Feedback Mechanism:
Allow users to like or dislike recommendations to improve accuracy.
 CONT...
Input-Related Requirements

• The system accepts structured and validated inputs, including user-provided data
such as ratings, reviews, and preferences, as well as product metadata like
categories, tags, descriptions, and pricing, and interaction logs such as clickstream
data and purchase history.

Output-Related Requirements
• This system displays recommendations in an intuitive format within user dashboards
and product pages, such as sections labeled “Recommended for You” or “Similar
Products.”
• It offers various types of recommendations, including top-rated products in specific
categories, frequently bought together suggestions, and trending products based on
user interactions.
 CONT...

Storage-Related Requirements

• This system uses a robust database to store user information, such as IDs and
interaction logs, product metadata, including reviews and descriptions, and
precomputed recommendation scores for fast retrieval, ensuring scalability and high
performance.

• It ensures that all data is appropriately indexed to improve query performance and
response times and implements a caching mechanism to reduce the database load
for frequently accessed data.
NON-FUNCTIONAL REQUIREMENTS

Performance

• The system guarantees 99.9% uptime to ensure availability, even during high-traffic
events such as flash sales or seasonal promotions.
• Recommendations should load within 2 seconds to avoid impacting user experience
User Interface (UI)

• On all platforms, including PCs, tablets, and smartphones, the user interface (UI) is
developed to be responsive and intuitive

• It incorporated features such as filters to sort recommendations by criteria such as

price, brand, or user ratings, along with visual cues like badges (“Top Rated” or
“Trending Now”)
NON-FUNCTIONAL REQUIREMENTS

Security and Access Permissions

• It implements encryption protocols to protect sensitive user data both in transit and
at rest.
• Regularly audit the system for vulnerabilities and apply security patches to mitigate
risks

Backup and Recovery

• Critical data, such as user profiles, interaction logs, product details, and metadata,
are backed up by the system every day. It ensures strong disaster recovery
capabilities by maintaining redundant backups using geographically dispersed data
centers.
SYSTEM ANALYSIS

• System analysis is the process of developing abstract models of a system, with each
model presenting a different view or perspective of that system.
• System analysis is about representing a system using some kind of graphical
notation. Usually Unified Modelling Language(UML).
• For this presentation we will focus on these three diagrams
⚬ Use case diagram
⚬ Sequence diagram and
⚬ Class diagram
USE CASE DIAGRAM
ACTORS IN THE USE CASE

• Customer/Users: users are the main actors of our system who interact with it in a
daily basis to buy good for use.
• Retailer/Merchant: merchants are the people/companies that regularly sell stuff on
the E commerce site.
• Site Administrator: site administrator manages the E-commerce platform's
operations, including user accounts, product listings, order processing, and site
maintenance. They ensure smooth functionality, compliance, and enhanced
customer experience through effective oversight and support.
• Payment processors: third party companies that facilitate the online transactions
on the E-commerce website. Eg: santimPay,arif pay, chappa
SEQUENCE DIAGRAM

• the sequence of interactions among objects and used to represent or model the flow
of messages, events and actions between the objects or components of a system.

• Sequence diagrams are also used primarily to design, document and validate the
architecture and interface of the system by describing the sequence of actions that
need to be performed to complete a task.
THE NEXT SEQUENCE DIAGRAM SHOWS THE SEQUENCE IN WHICH DIFFERENT ACTORS
INTERACT WHILE USING THE E-COMMERCE RECOMMENDATION SYSTEM.
THIS SEQUENCE DIAGRAM SHOWS THE SEQUENCE OF A CUSTOMER BUYING A PRODUCT
CLASS DIAGRAM

• Class Diagram is a high-level representation of the system's structure during the

analysis phase of software development.
• It focuses on understanding the problem domain.
• It defines the key concepts, relationships, and attributes needed to solve the
problem, without diving into implementation details.
ANALYSIS LEVEL CLASS DIAGRAM
THANK
YOU