FAKE NEWS DETECTION SYSTEM
1
Arnab Chatterjee
1 21MCA10023
Team Members Shubham Kumar
2 21MCA10064
Team Ayan Kumar Dey
Members 3 21MCA10069
Survi Banke
Under the guidance of
Dr. Rajeev Goyal
4 21MCA10156
2
Introductio
1 n
Proposed work
2 and
methodology
Hardware and
3 Software
requirements
Table
Architecture
of 4 Diagram
conten 5
Complete
module split-
t ups
6 Future scope
3
INTRODUCTI
Want ON
to know what’s real ?
Fake news detection system is a platform where you
can verify bits and bites of everything you hear.
Takes the user input from the client and predicts the
accuracy of the model.
4
Proposed work and
methodology
Project Scope: Working of
Project Present
Definition: Now a days social System:
media platform
Fake news have become a This website fetches
detection system prime source of data from the user
is a platform rumors. So anyone and this input data
where we can can verify the is then passed to
verify any kind of news or the article the models and gets
news that we before blindly trained and tested
here in our daily trusting and accordingly the
life. it and spreading accuracy of the
further. news is showen
5
Drawbacks of present
system: Methodology:
We have used the incremental
software development life cycle
Unable to work on current or model for the development of
real time updated data. modules, as we are performing
planning, designing, development
and testing in the increments.
6
7
8
INPUT DATASET
DATA PROCESSING
DATA TOKENIZATION
DATA SEGMENTATION
LEVEL- 0 OUTPUT
NEWS ARTICLE INDEX
LEVEL- 1
CALCULATE MISMATCH
RATIO MODEL
YES NO
IS MISMATCH
RATIO > THRESHHOLD
RUMER NOT RUMER
( FAKE ) ( REAL)
Module
Split-ups
11
User Interface 0
1
News Detection Training the
0 model
Components according to
2 the user input
Predecting the
accuracy and
displaying the
0
correctness of the 3
news
12
User Interface
• This is the data given by the user to
test the news.
• User given data is received through
the text box.
• After receiving the user data through
the text box it is passed to the model
for training and testing.
13
Model Training
• Dividing the data set into training and
testing set in the ratio of 70% and 30%
respectively.
• Using logistic regression model to train
the set.
14
Prediction
System
Creating the predictive system to analyze
the news
15
Displaying the
result
A pop up is generated on the UI
displaying the correctness of the
news.
16
FUTURE SCOPE
Adding a Working with
personalized real time Launching on
dashboard for Android
data
We will add real time
the user Make it more user accessible by
Providing a login and data so that latest
launching it in the Android
signup page for news can be
predicted with more platform
individual clients.
accuracy.
17
18
