Scribd
Upload
15 views15 pages

HackShield (SIH Final)

Uploaded by

22ec3025
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
15 views15 pages

HackShield (SIH Final)

Uploaded by

22ec3025
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 15

Fake Profile Detection on

Social Sites by Using Machine


Learning Algorithm
Basic Details of the Team and
Problem Statement
Ministry/Organization Name/Student Innovation: Government of Gujrat

PS ID:1364

Problem Statement Title : Fake Social Media Profile Detection and Reporting

Team Name: HackShield

Team Leader Name: Aman Jangir

Institute Code (AISHE):

Institute Name: Rajiv Gandhi Institute of Petroleum Technology

Domain : Blockchain and Cybersecurity


Team Member Details
Team rdLeader Name: Aman Jangir
CSE(3 year)
Team rdMember 1 Name: Abhishek Gupta
CSE(3 year)
Team rdMember 2 Name: Gourav Pandey
CSE(3 year)
Teamnd
Member 3 Name: Astitva Pathak
IT(2 year)
Team ndMember 4 Name: Md. Faizan
ECE(2 year)
Team Member 5 Name: Devisha Bhargava
CSD(1st year)
OBJECTIVE
Detecting the presence of fakes and spammers
represents a significant challenge across all social
media platforms, necessitating machine learning
techniques for effective identification and prevention..
The objective is to-
● To develop a robust solution for spammer/fake
accounts, after carefully examining each instance.
● To leverage machine learning so that high level of
accuracy can be achieved.
● Proper Data Insights and Patterns can be identified
FLOWCHART
Dataset Handling of
Exploration Outliers

Neural Network Data Insights


Development

Testing and
Evaluation
Inference
Roadmap

• Initial Data Exploration:


• Data Wrangling
• Data Insights
• Core Neural Network Development
• Evaluation
• Testing and Inference
TechStack
Application Frontend : HTML ,CSS ,JavaScript , ReactJs
Backend: Nodejs for routing
Web Frameworks: ExpressJS
Database: NO SQL
Authentication: NodeJS
TechStack used in ML model

• Python
• Pandas
• Numpy
• TensorFlow or PyTorch
• Scikit-Learn
• Matplotlib and Seaborn
• Jupyter Notebook
• CSV file Handling
• Text editor or IDE
• Conda or Virtual Environments
• Data Processing Tools
DataSet Used:
The ML model was meticulously trained using a carefully curated dataset of
Instagram profiles randomly selected from four distinct countries, ensuring
diversity and inclusivity. The dataset also thoughtfully encompassed a wide
range of fake profiles for comprehensive coverage. In its initial phase, over
900 sample profiles were meticulously extracted and utilized for model
training.

To enhance the accuracy of the ML model, an ambitious expansion of the


dataset is underway, with a target of accumulating more than 8000 profiles.
This expansion aims to bolster the model's performance and raise its
accuracy to an impressive 99% or higher.

This comprehensive approach to dataset collection and expansion reflects a


commitment to ensuring the model's robustness and reliability, while also
addressing the real-world challenges posed by Instagram profiles across
various regions and the prevalence of fake accounts.
Current Accuracy of the Prototype:
Interactive Website:
Our user-friendly website has been designed to provide easy access to our
model. Users can simply paste a link to any profile on the website, initiating a
profile analysis process that extracts valuable machine learning (ML)
features from the profile. These extracted features are then fed into our ML
model, and the website displays the model's prediction, accompanied by a
confidence percentage. This percentage indicates whether the profile is likely
to be genuine or fake, offering users a reliable assessment tool.

Moreover, our website boasts a convenient login page that keeps track of
users' past predictions and activities. This feature ensures a seamless
working environment for users, allowing them to access their previous
interactions with the model effortlessly. It provides a valuable reference point
for users and enhances their overall experience by streamlining their
workflow. Whether you're a casual user or a frequent visitor, our website is
designed to cater to your needs efficiently and effectively.
Websites Login Page:
Websites Work Envoirment
Security Threats:
SQL Injection:
● Threat: Attackers inject malicious SQL code into input fields,
exploiting vulnerabilities to access or manipulate your database.
● Protection: Use parameterized queries and prepared statements to
validate and sanitize user inputs. Regularly update your website's
software and employ a Web Application Firewall (WAF) to block
malicious requests.
Cross-Site Scripting (XSS):
● Threat: Attackers inject malicious scripts into web pages viewed by
other users, potentially stealing user data or hijacking sessions.
● Protection: Input validation and output encoding are crucial.
Implement Content Security Policy (CSP) headers to restrict script
execution. Use security libraries and frameworks that offer XSS
protection.
Brute Force Attacks:
● Threat: Attackers try to gain unauthorized access to your website by
repeatedly trying different username/password combinations.
● Protection: Enforce strong password policies, implement account
lockout mechanisms, and use CAPTCHA or reCAPTCHA to protect
against automated login attempts. Consider implementing
multi-factor authentication (MFA) for added security

You might also like

pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy