Returns any logical fallacies in a news article and any related news articles for better perspective on the story.
Logical fallacies are flawed reasoning or false arguments that can undermine the validity of a news source. Examples include circular reasoning, false dichotomies, ad hominem attacks, etc - full list is below. News Stories online are optimized to perform well on specific metrics like clicks or time spent, however optimizing for metrics alone does not guarantee the story was written with logically sound reasoning.
Comparing news articles specifically for logical flaws requires reconciliation of reasoning about arguments rather than clicking on a headline, and will leave you with a greater awareness of the quality of your news. Overall, allowing logical flaws to persist makes news stories less safe and ethical.
Run in streamlit: clone repo, then 'streamlit run newsvalidation.py`
Eliminating fallacies ensures news sources remain logically valid and aligned with human reasoning. This builds user trust and could lead to greater long-term retention.
For this app to run, you'll need API keys from OpenAI (get it here) and Serper (get it here).
List of logical fallacies below (there are 19 from this scholastic paper that reconciled new categories of logical fallacies since Aristotle's identification of 13). https://arxiv.org/pdf/2212.07425.pdf
To deploy on Railway using a one-click template, click the button below.
