TY  - JOUR
T1  - Artificial Intelligence-Based Sentiment Analysis
AU - Mesbah, Saleh AU - Samir, Aymen AU - Madbouly, Magda 
JO  - Journal of Engineering and Applied Sciences
VL  - 16
IS  - 1
SP  - 18
EP  - 22
PY  - 2021
DA  - 2001/08/19
SN  - 1816-949x
DO  - jeasci.2021.18.22
UR  - https://makhillpublications.co/view-article.php?doi=jeasci.2021.18.22
KW  - BERT
KW  -latest technology
KW  -neural network
KW  -accuracy
AB  - Bidirectional Encoder Representations from
Transformers (BERT) represents the latest technology of
pre-trained language models which have recently
advanced a wide range of natural language processing
tasks. This study aims to investigate how BERT can be
usefully applied in sentiment analysis tasks with fully
connected neural network. The proposed model is
developed using simple tips preventing it from
over-fitting and enabling it to be fine-tuned easily on such
down stream task. BERT performs much better as a
Strong text embedding Model. Using such procedures
successfully provide a better accuracy than the expensive
machine learning procedures.
ER  -