A technocognitive approach to detecting fallacies in climate misinformation.

Area of Science:

• Interdisciplinary research at the intersection of technology and psychology.
• Application of computational methods to address societal challenges in climate change communication.

Background:

• Climate change misinformation is a complex problem requiring novel solutions.
• Existing interventions often lack automated methods for detecting misleading techniques and logical fallacies.
• A 'technocognitive' approach synthesizes computer science and psychology to combat misinformation.

Purpose of the Study:

• To develop a dataset mapping climate misinformation to reasoning fallacies.
• To train a computational model for detecting fallacies in climate misinformation.
• To improve the accuracy and efficiency of automated misinformation detection.

Main Methods:

• Utilized a critical thinking methodology to deconstruct climate misinformation.
• Created a labeled dataset of climate misinformation examples and associated reasoning fallacies.
• Trained a machine learning model to identify fallacies within climate change narratives.
• Evaluated model performance using the $$ score for precision and recall.

Main Results:

• Achieved $$ scores 2.5-3.5 times better than prior research.
• Demonstrated higher accuracy in detecting fallacies like fake experts and anecdotal arguments.
• Identified challenges in detecting fallacies requiring background knowledge, such as oversimplification and slothful induction.

Conclusions:

• The developed model shows significant promise for automated detection of climate misinformation fallacies.
• This research provides a foundation for creating automated systems to counter misinformation with technique-based corrections.
• Future work can focus on improving detection of more complex fallacies and integrating generative correction mechanisms.