Skip to content

francheska-vicente/data102-fake-news

Repository files navigation

Tsimis Yarn?: Identifying Fake News Among the News Articles Online

In the age of digital media and information, false information, especially, fake news can now be easily disseminated across the globe. The potential damage that could be inflicted by fake news can be serious as news could travel more quickly now through social media platform, especially sensational stories that disinformation producers often manipulate to sell fake news (Siar, 2021). The consequences of dissemination of fake news could go far and wide as disseminating misleading or fabricated news articles can have significant impact on public opinion, social discourse, and even democratic processes.

A new Pulse Asia survey revealed that about 9 out of 10 Filipino adults perceive the widespread dissemination of “fake news” as a problem within the country (Lalu, 2022). According to another poll, 58% of Filipinos view social media influencers, bloggers and vloggers as peddlers of fake news about government and politics, followed by journalists at 40%, national politicians at 37% and local politicians at 30% (Gregorio, 2022). This underscores the continued threat to Philippine press freedom posed by disinformation and fake news that persistently spread and mislead the people, especially Filipinos (Casayuran, 2023).

Now, this project is rooted from the urgent need to address the widespread dissemination of fake news in today’s society, especially in the Philippines. As there is evidently a need for effective tools and techniques to identify and combat fake news, this project will focus on creating a model for fake news detection. In this project, classification and clustering algorithms would be applied for building the detection model. Through these algorithms, differentiating fake news from real news will be performed by the model and utilize this as its knowledge to identify fake news in online texts.

By developing and implementing reliable fake news detection algorithms, we can empower individuals, social media platforms, press media freedom, and news organizations to identify and combat misinformation effectively.

How to set up and run the project locally through JupyterNotebook or JupyterLab

  1. Extract the folder from the zipped file that you can download through this DownGit link.
  2. Launch Jupyter notebook or JupyterLab.
  3. Navigate to the project folder containing main.ipynb.
  4. Open main.ipynb. This contains the data pre-processing and cleaning, and the Exploratory Data Analysis.
  5. To see the model training and tuning, open ModelingPT1_Base.ipynb, ModelingPT2_5000.ipynb, ModelingPT2_Base.ipynb, ModelingPT3_Undersampling5000.ipynb and ModelingPT3_UndersamplingBase.ipynb. It is recommended not to run these files as they need resources.

Authors

  • Banzon, Beatrice Elaine B.
  • Buitre, Cameron
  • Marcelo, Andrea Jean C.
  • Navarro, Alyssa Riantha R.
  • Vicente, Francheska Josefa

About

In this project, a dataset about fake news is collected and combined with pre-existing datasets. In addition, a model that can detect if an input text is a piece of fake news is created.

Topics

Resources

Stars

Watchers

Forks