File size: 3,613 Bytes
f88d641 dda8a83 ab9f8c3 6846430 dda8a83 6846430 f88d641 dda8a83 3275c8c 5f5a651 6846430 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 |
---
license: mit
language:
- en
widget:
- text: >-
There is little doubt that some players in the climate game not a lot, but
enough to have severely damaged the reputation of climate scientists in
general have stepped across the boundary into postmodern science.
example_title: Ad homienm
- text: >-
Another famous place is the Tuvalu Islands, which are supposed to soon
disappear. There we have a tide gauge record, a variograph record, from
1978, so it's 30 years. And again - absolutely no trend, no rise.
example_title: Cherry picking
- text: >-
So do petitions signed by more than 30,000 scientists that have challenged
IPCC's 1995 procedures and report representations.
example_title: Fake experts
- text: >-
Fourth, if industrial civilization is dangerously altering global climate,
can any treaty stop it? The Kyoto accord in itself would do nothing to
mitigate climate change, since it exempts the developing countries, which
will be the major emissions source in the next century.
example_title: Impossible expectations
tags:
- climate
pipeline_tag: text-classification
---
# Model Card
## Model Details
### Model Description
<!-- Provide a longer summary of what this model is. -->
This model identifies 12 distinct climate change denial strategies or fallacies to classify and analyse texts that express skepticism or opposition to climate change scientific findings.
These 12 distinct labels come from the FLICC taxonomy created by John Cook and his colleagues.
The FLICC taxonomy divides denial strategies into five primary categories: fake experts, logical fallacies, impossible expectations, cherry-picking, and conspiracy theories.
- **Developed by:** Francisco Zanartu, John Cook, Julian Garcia and Markus Wagner
- **Model type:** DeBERTa (Decoding-enhanced BERT with disentangled attention) is a Transformer-based neural language model
- **Language(s) (NLP):** Finetuned and evaluated on a English dataset.
- **License:** [More Information Needed]
- **Finetuned from model:** microsoft/deberta-v2-xlarge
| Fallacy Type | Definition |
|----------------------|-------------------------------------|
| Ad hominem | Attacking a person/group instead of addressing their arguments |
| Anecdote | Using personal experience or isolated examples instead of sound arguments or compelling evidence |
| Cherry Picking | Selecting data that appear to confirm one position while ignoring other data that contradicts that position |
| Conspiracy theory | Proposing that a secret plan exists to implement a nefarious scheme such as hiding a truth |
| Fake experts | Presenting an unqualified person or institution as a source of credible information. |
| False choice | Presenting two options as the only possibilities, when other possibilities exist |
| False equivalence | Incorrectly claiming that two things are equivalent, despite the fact that there are notable differences between them |
| Impossible expectations | Demanding unrealistic standards of certainty before acting on the science |
| Misrepresentation | Misrepresenting a situation or an opponent’s position in such a way as to distort understanding |
| Oversimplification | Simplifying a situation in such a way as to distort understanding, leading to erroneous conclusions |
| Single cause | Assuming a single cause or reason when there might be multiple causes or reasons |
| Slothful induction | Ignoring relevant evidence when coming to a conclusion | |