SDP END-TERM EVALUATION

FAKE NEWS DETECTION

Supervised By: Dr. Prativa Das

Group No.:D13 Department of Computer Sc. and Engineering

Ashutosh Kumar (2041011113) Faculty of Engineering & Technology (ITER)
Aditi Rath (2041018064) Siksha ‘O’ Anusandhan (Deemed to be) University
Ashutosh Sarangi (2041019145) Bhubaneswar, Odisha
Indrajit Das (2041004164)

1
Presentation Outline
▪ Introduction
▪ Overview
▪ Motivations
▪ Uniqueness of the work
▪ Literature Survey
▪ Existing System
▪ Problem Identification
▪ Model Diagram
▪ Libraries and Algorithms used
▪ Experimentation and Results
▪ System Specifications
▪ Datasets Description
▪ Parameters used
▪ Experimental outcomes
▪ Result Analysis and Validation
▪ Conclusion and Future Scope
▪ Bibliography 2
 Introduction
❑ Overview
• Fake news refers to false or misleading information presented as news, often designed
to deceive and manipulate public perception.
• It impacts society by eroding trust in legitimate media, polarizing communities,
influencing public opinion and behavior, undermining democracy, inciting violence,
and causing economic harm.
• Fake news detection involves using techniques to identify and flag news articles or
other content that contain false or misleading information.
• Leveraging machine learning algorithms can significantly enhance our ability to detect
and combat fake news in today’s digital landscape.
• Predicting the accuracy of the news from various algorithm can help us to effectively
combat fake news in today’s digital landscape.
3
Introduction contd..
❑ Motivations
• Enhancing Media Credibility.
• Combatting misinformation and protecting consumers.
• Countering Malicious Activities like scams and fraudulent schemes.
• Ensuring accurate representation of Democracy.

4
Introduction contd..
❑ Uniqueness of the work
• Various algorithms are used by different fake news detection systems to recognize and

categorize fake news, and the algorithm selected has a big influence on the system's
accuracy.

• Numerous data sources, such as social media sites, news websites, and fact-checking

databases, are accessible to these systems.

• These algorithms employ a variety of features, including linguistic ones like syntax and

grammar as well as semantic ones like word choice and context, to detect fake news.

5
 Literature Survey
❑ Existing System

• Fact-Checking Website: These are website that specialize in verifying the accuracy of
news articles and other content . Examples include FactCheck.org and Snopes.com .

• Browser extensions: Such as News Guard and Fake News Detector, that can help users
identify fake news by flagging articles that have been identified as fake or misleading.

• Media literacy education: Some organizations are focused on promoting media literacy
education to help individuals recognize and avoid fake news .

6
Literature Survey

7
Literature Survey contd..
YEAR PUBLISHER TITLE DATASET ALGORITHM USED RESULT

Fake news detection using machine LIAR Random forest, SVM, Accuracy more than 75%
2020 IOP learning approaches by Khanam et al. Decision Tree and Naive
Bayes, XGBOOST

Fake news detection using machine Dateset containing real Naive bayes, SVM, SVM hasaccuracy of
2021 IEEE learning by Rupali Patil et al. and fake news Passive aggressive 95.05%
classifier

Fake news using machine learning ISOT fake news dataset Logistic regression(LR), RF algorithm with accuracy
2020 Hindawi ensemble methods by Ahmad et al. and Kaggle KNN, LSVM, RF of 99%

Fake news detection using machine publicly available data Naive Bayes , RF, LR LR with accuracy of 75%
2021 IJERT learning algorithms by Uma Sharma sets algorithms
et al.

Fake news Detection on Social Twitter, Facebook Machine Learning, Deep Analyzed different
2021 Springer Media: A Review by Kumar et al. Learning, Natural approaches to fake news
Language detection and suggested
future research directions.
Processing(NLP)

8
Literature Survey
YEAR PUBLISHER TITLE DATASET ALGORITHM USED RESULT

Fake news detection on social Twitter, Facebook Hybrid Deep Learning Model Achieved 93.5% accuracy
2018 ACM media : A Data Mining Perspective
by Wang et al.

Fake news detection: A deep Kaggle fake news Convolutional neural Achieved 93.84% accuracy
2019 IEEE learning approach by Khan et al. challenges network(CNN)

A survey on Automated Fake news N/A N/A Analyzed different approaches and
2020 Elsevier detection: Trends and Challenges challenges to automated fake news
by Thakur et al. detections

Fake news detection using machine LIAR XGBoost, RF, AdaBoost, Bagging Classifier and AdaBoost
2019 SSRN learning by Goswami et al. ExtraTrees, Bagging attain accuracy 70%

Fake news detection using machine kaggle Support Vector Machine Recognition rate of 100
2021 IEEE learning by Baarir et al.

9
Literature Survey Contd..
❑Problem Identification
• Lack of Standardized Dataset.

• Explainability of Models.

• High Variability in News Sources.

• Evolving Nature of Fake News.

• User Behavior Analysis.

• Ethical Considerations.

10
Model Diagram

Fig 1: Model diagram 11

Libraries and Algorithms used
❑ Libraries
• NumPy
• Matplotlib
• Scikit-learn
• Pandas
• NLTK
• Word Cloud

❑ Algorithm Used
• Decision Tree
• Random Forest
• SVM
• Naïve Bayes
• Logistic Regression

12
Algorithms used

• Root Node: The first decision is made based on a

critical feature.
• Decision Nodes: Internal Nodes that split the data,
each of them represents a question on attribute.
• Leaf Nodes: Terminal Nodes that provide the
final classification.
• Sub-Tree: Useful for breaking down complex
decisions into simpler parts.

Fig 2: Decision Tree

13
Experimentation and Results
❑ System Specifications

• Processor: Intel i5 11th Gen

• RAM: 8GB
• Storage: 512GB SSD
• Graphics Card: NVIDIA GEFORCE GTX 1650
• Operating System: Windows (x64 bit)
• Language: Python
• Platform: VS Code

14
Experimentation and Results Contd..
❑ Datasets Description
NEWS
• Dataset are from Kaggle repository.
• It’s consists 2 Sets :
Real
o Fake News 48%
o Real News Fake
• No. of rows : 52%
o 23539- false
o 21418- true Fake
Real

Fig 3: News Percentage 15

Experimentation and Results Contd..
❑ Datasets Description
• Government News
• Middle-east
• News
• US_News
• left-news
• Politics News
• World news

16
Experimentation and Results Contd..
❑ Parameters used
➢ Count Vectorizer Parameters:
• Lowercase , removal of stop words
➢ Future Scope Parameters:
• Tf-idf Transformer Parameters
o nngram_range: (1,1)
o max_df: 1.0
o min_df: 1
• Decision Tree Classifier Parameters
o max_depth: None
o min_samples_leaf: 1
o max_features: None
• train_test_split Parameters
o test_size: 0.25
o train_size: None
17
Experimentation and Results Contd..
❑ Experimental outcomes
Classifier Accuracy Precision Recall F1 Score
Naïve Bayes 93.65% 94% 94% 94%

Random Forest 98.75% 99% 99% 99%

Logistic Regression 98.78% 99% 99% 99%

Support Vector Machine 99.33% 99% 99% 99%

Decision Tree 99.67% 99.7% 99.52% 99.61%

18
Experimentation and Results Contd..
❑ Result Analysis and Validation

Fig 4: Confusion matrix without normalisation 19

Experimentation and Results Contd..
❑ Result Analysis and Validation (User Interface) Contd..

Fig 5: Prediction Fake News 20

Conclusion and Future Scope
❑ Decision-tree algorithm has a promising future scope in the detection of fake-news. It’s a well-
liked ML model that is simple to use and comprehend.
❑ Decision trees can be modified to function in an online learning environment, where the system
can rapidly update itself with the addition of new data.
❑ Ensemble approaches can be used to integrate decision-trees with other ML methods. This method
can improve accuracy in fake-news-detection.
❑ Fake-news can be expanded through many medium such as images, text, and videos. Decision
trees can be extended to handle multimodal data from different modalities to enhance the accuracy.
❑ For the purpose of real-time fake news detection, decision-trees can be improvised. The decision-
tree’s size and complexity can be decreased to enable quick identification and prediction .

21
Bibliography
[1] Albahr, A., & Albahar, M. (2020). An empirical comparison of fake news detection using different machine learning algorithms.
International Journal of Advanced Computer Science and Applications, 11(9).
[2] Khan, A. I., Shahzad, F., & Ali, S. (2019). Fake news detection: a deep learning approach using CNN. IEEE Access, 7, 44112-44121.
doi: 10.1109/ACCESS.2019.2901590.
[3] Thakur, P., Shah, R. R., & Rana, N. P. (2020). A survey on automated fake news detection: Trends and challenges. Information
Processing & Management, 57(2), 102026. doi: 10.1016/j.ipm.2019.102026.
[4] Kumar, R., Singh, R. K., & Roy, P. P. (2021). Fake news detection on social media: A review. Artificial Intelligence Review, 54(4),
2997-3030. doi: 10.1007/s10462-020-09981-4
[5] Allcott, H., & Gentzkow, M. (2017). Social Media and Fake News in the 2016 Election. Journal of Economic Perspectives, 31(2), 211-
236. https://doi.org/10.1257/jep.31.2.211
[6] Conroy, N. J., Rubin, V. L., & Chen, Y. (2015). Automatic deception detection: Methods for finding fake news. Proceedings of the
Association for Information Science and Technology, 52(1), 1-4. https://doi.org/10.1002/pra2.2015.145052010082
[7] Shu, K., Sliva, A., Wang, S., Tang, J., & Liu, H. (2017). Fake News Detection on Social Media: A Data Mining Perspective. ACM
SIGKDD Explorations Newsletter, 19(1), 22-36. https://doi.org/10.1145/3137597.3137600
[8] Rashkin, H., Choi, E., Jang, J. Y., Volkova, S., & Choi, Y. (2017). Truth of Varying Shades: Analyzing Language in Fake News and
Political Fact-Checking. Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing (EMNLP), 2931-2937.
https://doi.org/10.18653/v1/D17-1317
[9] https://www.kaggle.com/datasets/clmentbisaillon/fake-and-real-news-dataset
22
23