Building a Twitter Sentiment Analysis System with Recurrent Neural Networks

This paper presents a sentiment analysis solution on tweets using Recurrent Neural Networks (RNNs). The method is can classifying tweets with an 80.74% accuracy rate, considering a binary task, after experimenting with 20 different design approaches. The solution integrates an attention mechanism ai...

Full description

Bibliographic Details
Main Authors: Sergiu Cosmin Nistor, Mircea Moca, Darie Moldovan, Delia Beatrice Oprean, Răzvan Liviu Nistor
Format: Article
Language:English
Published: MDPI AG 2021-03-01
Series:Sensors
Subjects:
Online Access:https://www.mdpi.com/1424-8220/21/7/2266