An Effective BERT-Based Pipeline for Twitter Sentiment Analysis: A Case Study in Italian

Over the last decade industrial and academic communities have increased their focus on sentiment analysis techniques, especially applied to tweets. State-of-the-art results have been recently achieved using language models trained from scratch on corpora made up exclusively of tweets, in order to be...

Full description

Bibliographic Details
Main Authors: Marco Pota, Mirko Ventura, Rosario Catelli  and Massimo Catelli
Format: Article
Language:English
Published: MDPI AG 2021-12-01
Series:Sensors
Subjects:
n/a
Online Access:https://www.mdpi.com/1424-8220/21/1/133