Fine-Grained Sentiment Analysis of Arabic COVID-19 Tweets Using BERT-Based Transformers and Dynamically Weighted Loss Function

Alturayeif, Nora; Luqman, Hamzah

doi:10.3390/app112210694

Cited by 12 publications

(12 citation statements)

References 49 publications

(51 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Finally, the Logistic Regression (LR), Multinomial Naïve Bayes (MNB), Support Vector Machine (SVM) and K Nearest Neighbor (KNN) were examined, resulting in having a model with maximum accuracy to analyze the people perception regard coronavirus of 90% using Support Vector Machine classifier. In the same vien, Alturayeif and Luqman ( 2021 ) employed two transformer-based models for sentiment detection of the Arabic tweets, with the assumption that many emotion can exist in the same tweet. Resulting in F1-Micro score of 0.72 with the ability to precisely predict the educational-based tweets.…”

Section: Literature Reviewmentioning

confidence: 99%

Sentiment analysis and emotion detection of post-COVID educational Tweets: Jordan case

Qaqish

Aranki

Etaiwi

2023

Soc. Netw. Anal. Min.

View full text Add to dashboard Cite

Education evolved dramatically under Covid-19, and owing to the conditions, distant learning became mandatory. However, this has opened new realities to the educational business under the label of “Hybrid-Learning,” where educational institutions are still using online learning in addition to face-to-face learning, which has changed people’s lives and split their opinions and emotions. As a result, this study investigated the Jordanian community’s perspectives and feelings on the transition from pure face-to-face education to blended education by examining related tweets in the post-COVID era. Specifically, using NLP Emotion detection and Sentiment Analysis approaches, as well as deep learning models. As a result of analyzing the collected tweets, 18.75% of studied Jordanian’s community sample are dissatisfied (Anger and Hate), 21.25% are negative (Sad), 13% are Happy, and 24.50 percent are Neutral about it.

show abstract

Section: Literature Reviewmentioning

confidence: 99%

Sentiment analysis and emotion detection of post-COVID educational Tweets: Jordan case

Qaqish

Aranki

Etaiwi

2023

Soc. Netw. Anal. Min.

View full text Add to dashboard Cite

show abstract

“…Alturayeif and Luqman [2] used two transformer-based models, namely, AraBERT [14] and MARBERT [15], with a loss function that is weighted dynamically (DWLF) to analyze the sentiment of Arabic tweets. They evaluated their proposed method using SenWave and SenAIT datasets [16].…”

Section: Related Workmentioning

confidence: 99%

“…Since its emergence in the end of 2019, the pandemic affected people's lives in different fields, such as social life, psychological, learning and teaching, healthcare, and finance [1] [47]. During this phase, people used social media platforms, such as Twitter and Facebook, to express their feelings and opinions about the current situation, thereby encouraging researchers to investigate people's feelings among these social platforms through sentiment analysis studies [2]. However, Twitter is considered as one of the most popular social platforms, because of its availability and ease of knowledge exchange [3].…”

Section: Introductionmentioning

confidence: 99%

“…It is considered a complex process with five steps, which begins with collecting data and continues with preprocessing text, detecting sentiment, and classifying text into positive, negative, and other categories. The final step is to present the output [2], [5]. D'Andrea et al classified the techniques applied for sentiment analysis [5] into three categories, namely, lexicon-based, machine learning (ML), and hybrid approaches that incorporate both ML and lexicon-based approaches.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Sentiment Analysis for People’s Opinions about COVID-19 Using LSTM and CNN Models

Al-Khazaleh

Alian

Biltawi³

et al. 2023

Int. J. Onl. Eng.

View full text Add to dashboard Cite

The emergence of social media platforms, which contributed in activating the patterns of connection between individuals, leads to the availability of a huge amount of content such as text, images, and videos. Twitter is one of the most popular platforms of social media that encourage researchers to investigate people’s feelings and opinions among through sentiment analysis studies that elicited the interest of researchers in natural language processing field. Many techniques related to machine learning and deep learning models could be used to improve the efficiency and performance of sentiment analysis, especially in complex classification problems. In this paper, different models of long short-term memory recurrent neural network are used for the sentiment classification task. The input text was represented as vectors using Arabic pre-trained word embedding (Aravec). Experiments were conducted using different dimensions of Aravec on 15779 tweets about COVID-19 collected and labeled as positive and negative. The experimental results show an accuracy value of 98%.

show abstract

“…The research suggested a multi-label emotion classifier with emojis replacement based on AraBERT and MARBERT. They also presented a DWLF technique to give the loss function more weight in minority class data [10]. The hybrid network outperforms other models for various word embeddings, and its accuracy is higher than other models.…”

Section: Related Workmentioning

confidence: 99%

Spam and Sentiment Detection in Arabic Tweets Using MARBERT Model

Alotaibi¹,

Rahman²,

Alhaza³

et al. 2022

MMEP

View full text Add to dashboard Cite

Saudi Telecom Company (STC) is among the most popular companies in Saudi Arabia, with many customers. Yet, there is still a big room for improvement in users' satisfaction. Social media is the most robust platform to gauge users' satisfaction and determine their sentiments and critics. Twitter is among the most popular social media platform in this regard. STC customers prefer to use Twitter to write their feedback because it's a fast way to get responses due to the STC customer services account. One way to achieve customer demands and improve customer service is using the Sentiment Analysis tool. Sentiment Analysis on Twitter is highly used because of the significant number of tweets and the different opinions. Likewise, Deep learning is the best existing Sentiment Analysis method, and it has diverse models. Bidirectional Encoder Representations from Transformers (BERT) model is one of the deep learning models which have achieved excellent results in Sentiment Analysis for Natural Language Processing (NLP). NLP is mainly investigated in the English language. However, for Arabic, there is a significant gap to be filled. This study trained the proposed model using MARBERT and measured the performance using f1-score, precision, and recall metrics. We trained the model with an Arabic dataset of 24,513 tweets, including 1,437 positive, 13,828 negative, 5,694 neutral, 1,221 sarcasm, and 2,297 indeterminate tweets. The main goal is to analyze the tweets and get the sentiment to improve STC customer service. The proposed scheme is promising in terms of accuracy in contrast to existing techniques in the literature.

show abstract

Fine-Grained Sentiment Analysis of Arabic COVID-19 Tweets Using BERT-Based Transformers and Dynamically Weighted Loss Function

Cited by 12 publications

References 49 publications

Sentiment analysis and emotion detection of post-COVID educational Tweets: Jordan case

Sentiment analysis and emotion detection of post-COVID educational Tweets: Jordan case

Sentiment Analysis for People’s Opinions about COVID-19 Using LSTM and CNN Models

Spam and Sentiment Detection in Arabic Tweets Using MARBERT Model

Contact Info

Product

Resources

About