Fake News Detection with PAC and XGBoost

Problem Statement

In this project, it has been tried to guess whether the news is fake or real based on title and content.

Dataset

Dataset has 4 columns and 7795 rows without the header.

Column	Details
title	title of the news
text	news
label	Is the news fake or real?

Methodology

In this project, as stated in the title, results were obtained through PAC and XGBoost.

Analysis

#	Column	Non-Null Count	Dtype
0	Unnamed: 0	6335 non-null	int32
1	title	6335 non-null	object
2	text	6335 non-null	object
3	label	6335 non-null	object

dtypes: int64(1), object(3)

Numbers of News in Overall

• REAL 3171

• FAKE 3164

Confusion Matrix of PAC (Passive Aggressive Classifier)

592	46
43	586

Accuracy score: 0.9297553275453828

---

XGBoost

Confusion Matrix of XGBoost

576	50
62	579

Accuracy score(XGBoost): 0.9116022099447514

Process took 14.131280899047852 seconds.

How to Run Code

Before running the code make sure that you have these libraries:

• pandas
• time
• sklearn
• seaborn
• xgboost
• matplotlib

Contact Me

If you have something to say to me please contact me:

• Twitter: Doguilmak
• Mail address: doguilmak@gmail.com