Semi-supervised Sentiment Annotation of Large Corpora

Brum, Henrico Bertini; Nunes, Maria das Graças Volpe

doi:10.1007/978-3-319-99722-3_39

Cited by 7 publications

(3 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…After, we applied fine-tuning in the SA, AE, HS, and ID tasks. And we test in datasets TweetSentBR (Brum and das Grac ¸as Volpe Nunes, 2018), ABSAPT 2022(da Silva et al, 2022, ToldBR (Leite et al, 2020), and IDPT 2021 (Corrêa et al, 2021). Finally, we analyzed the results obtained in each task.…”

Section: Methodsmentioning

confidence: 99%

BERTimbau in Action: An Investigation of its Abilities in Sentiment Analysis, Aspect Extraction, Hate Speech Detection, and Irony Detection.

Junqueira

Silva

Costa

et al. 2023

FLAIRS

View full text Add to dashboard Cite

Social Media has revolutionized how individuals, groups, and communities interact. This immense quantity of unstructured data holds valuable information expressed in informal language. However, automatically extracting this information using Natural Language Processing requires adaptations of traditional methods or the development of new strategies capable of extracting information tackling web-prone language. BERT, a Deep Learning methodology proposed by Google in 2018, brought transfer learning to Natural Language Processing. In this work, we used a BERT model for the Portuguese language called BERTimbau to create models for Sentiment Analysis, Aspect Extraction, Hate Speech Detection, and Irony Detection. We experimented with the two BERTimbau models, base and large. Finally, we compared the results obtained in each task. Experiments with BERTimbau based models obtained improved results, F-Measure of 0.88 and 0.89 in Sentiment Analysis and Hate Speech Detection tasks, respectively, compared to classical Machine Learning approaches.

show abstract

Section: Methodsmentioning

confidence: 99%

BERTimbau in Action: An Investigation of its Abilities in Sentiment Analysis, Aspect Extraction, Hate Speech Detection, and Irony Detection.

Junqueira

Silva

Costa

et al. 2023

FLAIRS

View full text Add to dashboard Cite

show abstract

“…The semi-supervised annotation approach has been applied for other languages, such as the Brazilian Portuguese language. In [23], the authors extended a small sentiment corpus, which was annotated manually, to annotate a large unlabeled corpus. They used only one classifier to predict the classes of the unlabeled documents and added those documents which have a confidence value above a predefined threshold.…”

Section: Sentiment Annotationmentioning

confidence: 99%

AraSenCorpus: A Semi-Supervised Approach for Sentiment Annotation of a Large Arabic Text Corpus

et al. 2021

View full text Add to dashboard Cite

At a time when research in the field of sentiment analysis tends to study advanced topics in languages, such as English, other languages such as Arabic still suffer from basic problems and challenges, most notably the availability of large corpora. Furthermore, manual annotation is time-consuming and difficult when the corpus is too large. This paper presents a semi-supervised self-learning technique, to extend an Arabic sentiment annotated corpus with unlabeled data, named AraSenCorpus. We use a neural network to train a set of models on a manually labeled dataset containing 15,000 tweets. We used these models to extend the corpus to a large Arabic sentiment corpus called “AraSenCorpus”. AraSenCorpus contains 4.5 million tweets and covers both modern standard Arabic and some of the Arabic dialects. The long-short term memory (LSTM) deep learning classifier is used to train and test the final corpus. We evaluate our proposed framework on two external benchmark datasets to ensure the improvement of the Arabic sentiment classification. The experimental results show that our corpus outperforms the existing state-of-the-art systems.

show abstract

“…The best results were obtained with the Naïve Bayes classifier, so we consider these to be the baseline results. Brum and Nunes [9], in a subsequent work, used Naïve Bayes, SVM, Logistic Regression (LR), Multi-Layer Perceptron (MLP), Decision Trees, and Random Forest classifiers. In this case, Multi-Layer Perceptron was the bestperforming classifier, as listed in the table.…”

Section: Fine-tuning Resultsmentioning

confidence: 99%

Employing Transformers and emoji to perform sentiment classification of social media texts

Martinho de Barros

View full text Add to dashboard Cite

First and foremost, I thank God for everything.Then, I would like to thank my parents, Maria Terezinha and Genivaldo, for they have done their best to support me throughout my life. Among the many invaluable lessons I was given, perhaps the most important one is to be a good human being. I sincerely hope to live up to their expectations. Also, I extend my gratitude to my family, that regularly encourages me in my search for knowledge.This work would not have been possible without the precious help of my supervisors Prof. Dr. Hélio Pedrini and Prof. Dr. Zanoni Dias, who constantly shared their hardearned knowledge to solve the problems that cropped up during this research. I feel that I ended up trying their patience on some occasions, and for that I apologize.I also would like to thank my colleagues from the Visual Informatics Laboratory (LIV) of the Institute of Computing (IC), especially Gabriel Bianchin, Leodécio Braz, Vinicius Teixeira, Daiane Mendes, and Marianna Severo, who helped me with many things and with whom I shared many laughs and good moments.I should thank all my friends as well because they make my life substantially better and happier. I feel blessed to have them in my life, and I hope this continues for many more decades so we can have many more memorable moments and more stories to tell our grandchildren, should they come.Finally, I express my gratitude to Ji-chan -my beloved bicycle -who has been with me for about 25 years. Not only she makes my life easier but also more enjoyable and adventurous. She has been a superb partner and I am grateful for that.Last but not least, thank you, dear reader, for the interest in this modest work. I hope you find it easy to understand and maybe learn a thing or two. ResumoAvanços recentes na área de Processamento de Linguagem Natural trouxeram melhores soluções para uma série de tarefas interessantes como Aceitabilidade Linguística, Respostas a Perguntas, Compreensão de Leitura, Inferência de Linguagem Natural e Análise de Sentimento. Neste trabalho, focamos em Análise de Sentimento, que é um campo de pesquisa voltado ao estudo computacional de sentimentos. A Análise de Sentimento possui muitas aplicações práticas como sistemas de recomendação, monitoramento de satisfação de usuários e previsão do resultado de eleições.As tarefas mencionadas são importantes para o avanço da Inteligência Artificial, pois são desafiadoras e podem ser aplicadas em vários problemas. A abordagem tradicional é construir um classificador específico para cada tarefa, entretanto, com a popularização do conceito de pré-treinamento seguido de ajuste fino, tornou-se muito comum a utilização de uma mesma arquitetura em diferentes problemas, por meio de ajuste fino com dados da tarefa em questão.Métodos como ULMFiT, ELMo, BERT e seus derivados obtiveram sucesso substancial em muitas tarefas de Processamento de Linguagem Natural, no entanto, eles compartilham uma desvantagem: para pré-treinar esses modelos do zero, quantidades substanciais de dados e recursos computacionais são necessári...

show abstract

Semi-supervised Sentiment Annotation of Large Corpora

Cited by 7 publications

References 11 publications

BERTimbau in Action: An Investigation of its Abilities in Sentiment Analysis, Aspect Extraction, Hate Speech Detection, and Irony Detection.

BERTimbau in Action: An Investigation of its Abilities in Sentiment Analysis, Aspect Extraction, Hate Speech Detection, and Irony Detection.

AraSenCorpus: A Semi-Supervised Approach for Sentiment Annotation of a Large Arabic Text Corpus

Employing Transformers and emoji to perform sentiment classification of social media texts

Contact Info

Product

Resources

About