MisRoBÆRTa: Transformers versus Misinformation
نویسندگان
چکیده
Misinformation is considered a threat to our democratic values and principles. The spread of such content on social media polarizes society undermines public discourse by distorting perceptions generating unrest while lacking the rigor traditional journalism. Transformers transfer learning proved be state-of-the-art methods for multiple well-known natural language processing tasks. In this paper, we propose MisRoBÆRTa, novel transformer-based deep neural ensemble architecture misinformation detection. MisRoBÆRTa takes advantage two state-of-the art transformers, i.e., BART RoBERTa, improve performance discriminating between real news different types fake news. We also benchmarked evaluated performances transformers task For training testing, used large real-world articles dataset (i.e., 100,000 records) labeled with 10 classes, thus addressing shortcomings in current research: (1) increasing size from small large, (2) moving focus detection binary classification multi-class classification. dataset, manually verified ensure that they were correctly labeled. experimental results show accuracy problem was significantly influenced method employed learn context, size, vocabulary dimension. observe empirically best among models use only one transformer obtained BART, DistilRoBERTa obtains least amount time required fine-tuning training. However, proposed outperforms other To arrive at conclusion, performed ample ablation sensitivity testing datasets.
منابع مشابه
Scientific Facts Versus Media Misinformation
There are two intentional difference between this version of the article and the original published version. First, the original published version contains an error in Figure 2 which was inadvertently introduced during the editing process. Specifically, the sign of the condom use/HIV correlation was reversed. This electronic version of the article contains the correct version of Figure 2. Secon...
متن کاملMonad Transformers as Monoid Transformers
The incremental approach to modular monadic semantics constructs complex monads by using monad transformers to add computational features to a preexisting monad. A complication of this approach is that the operations associated to the pre-existing monad need to be lifted to the new monad. In a companion paper by Jaskelioff, the lifting problem has been addressed in the setting of system Fω. Her...
متن کاملThe spread of misinformation
How does information behave? A recent controversial approach is to use ideas in evolutionary biology to study the spread of information. However, in order for the information to ‘evolve’, there has to be some mechanism for variation. This paper would explore the ideas of memes which use the evolutionary biology approach as well as the pitfalls in using this picture. Finally, we will focus our a...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
ژورنال
عنوان ژورنال: Mathematics
سال: 2022
ISSN: ['2227-7390']
DOI: https://doi.org/10.3390/math10040569