Skip to content

GyaneshSamanta/CANIS-Data-Analysis

Repository files navigation

Top Insights

  • In general, Even if a news content is false, it has become alarmingly similar to true news.
  • Some of the key difference as per this study:
    • true news is presented in highly positive tones, whereas false news is generally more negative.
    • Stop words or vulgarity is more common in false news compared to true news.
    • capital letters are more thrown around in false news compared to true news.
    • top features which are present in true and false news according to this data set mostly deals with politics; some of the key highlights being Trump, Hilary, Clinton
    • If we study the top features that distinguished the true and false news, it was some factual entity as days of the week, quoting a statement by someone or Twitter.

Legend.ipynb

- So I believe this study showcased more similarities between true and false news rather than differences. 
- It is a NECESSITY of the hour to have a deep understanding of what might distinguish two news pieces which are similar to each other but one is true of the two.
- This reminds me of a game called one truth and two lies. 
- So if we could identify the key features which distinguish the lie and the truth, in our daily lives we could discard misinformation better. 
- sadly, this is not a simple model. 

About

Official submission by Gyanesh Samanta to Canis Data Analysis Hacakthon

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published