Logistic regression + TFIDF Vectorization has shown to be a good baseline method for many classifications tasks, such as hate speech detection, text classification. Random Forest + TFIDF Vectorization (RF-TFIDF) We use the Random Forest (RF) classifier with TfidfVectorizer word embedding...
Nevertheless, LLM based ones like BERT based ones and, even, GPT-based prompted systems are raising. Regarding performances, even if some disorders, like eating related ones, have been found to allow for automatic detection with good performances, others, like depression, are still difficult to ...
In recent years, numerous measures have been introduced to address hate speech and offensive language. On 25 February 2019, United Nations Secretary-General Guterres unveiled the “United Nations Strategy and Plan of Action on Hate Speech” to tackle the escalating global issues of xenophobia, racism...
analysis in autonomous driving, where sensing riders’ feelings can provide feedback to enhance the journey. Our study focused on textual inputs; an expanded work can include written, facial, or audible expressions. Moreover, we plan to compare BERT with other transformer like GPT-3 and XLNet....